I am going to break one of my unspoken cardinal rules: Only write about real problems and measurement that is actually possible in the real world.
I am going to break the second part of the rule.
I am going to define a way for you to think about measuring social media, and you can't actually easily measure what I am going to recommend. [Update: Please see update #2 below, you can now easily measure what's recommended in this post.]
So why break the rule?
Social media is evolving at an incredible pace. Most of us have no idea how to participate optimally in this unique channel – we are doing TV on Twitter (breaks my heart). The impact on the data side of the ecosystem is that massive amounts of data is being generated and much of what goes for measurement in "social media tools" is profoundly sub optimal (I'm being polite). We have IT-minded people engaging in massive data puking (one report with 30 metrics anyone?) and Marketing-minded people who are using lousy measures of success ("I got 158,632 Fans! Hurray!").
I want to propose a framework you can use to measure success using metrics that matter for one simple reason: They actually measure if you are participating in the channel in an optimal fashion.
Isn't that revolutionary? Use data to incentivise our companies to do the right thing by measuring what matters, what makes this channel so unique.
No more embarrassing your brand on Twitter, Facebook, Google Plus, YouTube. And we build out a loyal cadre of followers / friends / subscribers to boot!
So what actually matters in Social Media?
Not the number of Friends / Followers / Subscribers. Not the number of posts / tweets. Not the ridiculous Followers to Following ratio. Not the… well there are so many horrible ones to choose from.
What matters is everything that happens after you post / tweet / participate!
Did you grab attention? Did you deliver delight? Did you cause people to want to share? Did you initiate a discussion? Did you cause people to take an action? Did your participation deliver economic value?
The "so what? " matters!
Oh, I totally forgot to say this…. the advice in this blog post is only for businesses and brands that participate in social media. Businesses as in Red Bull and T-Mobile and Johnson & Johnson. Brands (all of the aforementioned plus…) as in Mitch Joel and Stephen Colbert and Nancy Pelosi. If you don't fall into those two categories then this social media measurement framework might not apply to you.
I'm proposing four distinct social media metrics we should measure, (and this is so cool) independent of the social channel you participate in.
Excited? Let's go….
1. Conversation Rate.
When I say most brands do TV on social media what I mean is that we do the same uninformed shouting and pimping on social media that we do on TV.
We know little about who is on the other end of the TV set and the medium places limits to what we can do. So to make our marketing more efficient we shout more loudly, more frequently!
We don't have to do that. We can get a very good sense for who is following / friending / subscribing to us. We can measure if what we are saying connects to them (in near real time!). And unlike all others, this channel has the word social in it! Social as in talk and listen and discuss.
So why not measure that?
Conversation Rate = # of Audience Comments (or Replies) Per Post
One beautiful thing… you can measure this on every social channel on the planet. Blog. Twitter. Facebook. Google Plus. YouTube.
What to do with it?
A high conversation rate requires a deeper understanding of who your audience is, what your brand attributes are, what you are good at, what value you can add to your followers and the ecosystem you participate in.
That is why I love this metric. It forces you to do the right thing right away. And it is a lot of work.
So aim for a higher Conversation Rate. Build your own watering hole in the digital universe. Have meaningful conversations with your audience. That's Marketing money just can't buy.
You can always be provocative, say silly things and get a high Conversation Rate. Pick Sarah Palin for your topic. :) But that would not be accretive for your brand equity, would it?
Remember we do not measure to manipulate the metrics, we measure to know if we are adding business value.
How to measure it?
Individually this is not that hard to measure. But across channels there does not seem to be an option.
This is where I need your help. Do you know of a tool that measures conversation Rate easily as defined above across the main social media channels? Please share via comments and I'll add it here. Thanks!
Up next, our second delightful metric…
2. Amplification Rate.
Every channel has inherent limitations, often exhibited by the number of ads you can buy. On Google (paid search), on Facebook (display ads), on Radio (audio ads), and all other channels you can think of.
But social media has a profound advantage you can tap into.
Not only do you have a network, but every node in your network has a network of its own! If you follow my advice and post something "incredible, relevant, of value" to your audience then they can allow you to break free of the limitations of your network and spread your word around to a more massive audience!
Take me as an example. I have, as of today, 57k followers on Twitter and around 12k on Google Plus. That's the limit. Even if every single person who follows me reads every single thing I write, I can at most reach 57k people on Twitter.
But the size of my second level network (the unique people who follow the people who follow me) is 6.3 mil. My real "reach" it turns out is not 57k, it is 6.3 mil!
So measure Amplification, the rate at which your followers take your content and share it through their network.
On Twitter:
Amplification = # of Retweets Per Tweet
On Facebook, Google Plus:
Amplification = # of Shares Per Post
On a blog, YouTube:
Amplification = # of Share Clicks Per Post (or Video)
(Share clicks as in number of times your social media buttons were used to spread the content.)
What to do with it?
As you post and tweet and you rock and you roll… measure what pieces of content (type) cause amplification (allow your social contributions to spread to your 2nd, or even 3rd, level network). Understand times and geo locations and topics and things.
Then do more of the type that increase amplification. You'll get more sharing and spreading of your content. But this is very, very important: You'll be giving your audience content they consider to be of such incredible value that they want to share it (and hence you'll know what your audience wants / loves).
Oh, oh, oh and…. over time your 2nd level network becomes your 1st level network… because they discover that you rock!
Marketing, relationships and a reach that money, honestly, can't buy.
How to measure it?
I don't quite know how to do it easily across all the channels. Individually you can, see image above, pull out Excel and make magic!
Do you know of a tool that precisely measures Amplification across all channels as defined above? Please let me know via comments, and I'll add it here.
Now on to a metric that had us at "hello"…
3. Applause Rate.
I'm sure you've noticed my secret evil plan to force you to understand your audience (and not just pimp your agenda in Social Media).
One powerful, more immediate way, to understand them is to measure Applause.
One Twitter:
Applause Rate = # of Favorite Clicks Per Post
On Facebook:
Applause Rate = # of Likes Per Post
On Google Plus:
Applause Rate = # of +1s Per Post
On a Blog, YouTube:
Applause Rate = # of +1s and Likes Per Post (or video)
What to do with it?
Simple… you want to know what the audience likes (to use the Facebook terminology) and what they don't. You get a much deeper understanding of what your audience likes so much that it will +1 your content (or contribution) and allow for that to be then shown to others in their social graph.
And consider this…
If you +1 this blog post, you'll not help me understand its relative quality, but when someone in our extended social graph does a search on Google for Social Media Metrics your endorsement of this content will show up in the search results. That's reassuring to your social graph, and it is great for me because your endorsement makes this post stand out over others and I get a relevant visitor/customer.
Sweet, right? Your selfless social media contribution comes back to assist you in driving valuable business outcomes.
That's why you measure Applause. It matters in ways you can't imagine!
How to measure it?
Individually the numbers are available in most tools. Easy to find in Google+ (see example in the end). For Facebook the number is included in Facebook Insights, though it is not available as easily in a simple way (at least not as expansively as outlined above). For Twitter, sadly I could not find it anywhere (inside Twitter or other tools).
So help me. Do you use a tool that will allow us to measure Applause Rate? Please share via comments.
Finally the metric any company leader will adore…
4. Economic Value.
I am smiling. I know that the long time readers of my blog would know that I would never let you get away without measuring hard business bottom-line impact of any digital effort!
It is foolish to believe that just Conversation Rate, Amplification Rate, Applause Rate will get you the eternal love and gratification (and perhaps budget!) of your company's leadership. Yes they care a little bit about this "social media thing." But if you want their adoration (and let me repeat: budget!) you are going to have to quantify the economic value created via social media.
You don't participate in social media to only drive business outcomes. I cannot stress that enough. If that is your primary objective you are going to suck at it (and the above metrics will reflect very efficiently how much you suck).
But.
A small percent of the people in your company / brand's social graph will come to your main digital outpost (usually your company website) and choose to do business with you. Some of them will buy something, others will sign up for your email marketing list, others still will order a catalog or write reviews for products on your site or sign up as an affiliate or create wish lists or marriage registries or phone your call center to order something or… stay with me…. buy your products or services in your supermarket / store / real world thing.
And you know what all of those things are? Macro and Micro Conversions!
And you know what you can do with macro and micro conversions? You can measure Economic Value!
On all social media channels:
Economic Value = Sum of Short and Long Term Revenue and Cost Savings
Social media participation, done right, adds value to the company's bottom-line. Some of it can't be computed. That is okay. But some of it can be and it is your job, nay duty (!), to quantify that.
It is not very hard to do. Read the two posts immediately above. They share very specific guidance for businesses of different types (B2B, B2C, A2K) about how to identify the macro and micro conversions and then compute economic value.
What to do with it?
Those of you who have been at one of my recent keynotes have seen this slide:
Your job is to identify that blue arrow, and the orange box (what it stands for and what the amount is). It is not very hard, just takes a little patience and imagination.
And here is the incredible, amazing, magical thing. Once you have your highest level segmented view of the acquisition strategy, above, you can in two seconds segment down to individual channels you participate in.
Your view will look something like the one below, from Google Analytics:
I can focus on the Per Visit Goal Value (economic value delivered by visitors from social media channels across my macro and micro conversions – note the 0% in the macro conversions column, ouch!) for each channel. StumbleUpon rocks ($1.43), Twitter takes the next spot (around $0.60) and then comes Facebook ($0.26, clearly not a winner for me).
Now, not only can I tell my CEO what the small amount of direct value added to the business is, I can also report to her/him exactly which channels are contributing how much. You can't be in every social channel that pops up. The above data can give you guidance on where to be.
You do Economic Value and you will never, ever have to beg for investment in Social Media, and your career will get on the fast track. I promise.
And just to repeat one more time. A vast majority of value your business / brand gets from social media will be owning your message, building the watering hole I've mentioned, having a direct relationship with your customers and so much more. But showing some direct economic value will get you permission to do more of that. Without it you are just another "smarty pants" promising "vague outcomes" via "the next hip thing."
How to measure it?
Use Google Analytics, Omniture, WebTrends, CoreIBMInsights, etc.
Takes less than five minutes to set up. Provides a lifetime of joy.
Four Metrics That Rock.
Conversation Rate. Amplification Rate. Applause Rate. Economic Value. Four simple measures that get you to focus on the right thing from a social media participation perspective, help you understand how well you are doing at it, and quantify the business impact.
The challenge is that thus far it is hard to pull them all together in one place. As I had mentioned earlier, Excel is your bff for now. My hope is that vendors will stop creating tools in silos (just do Twitter or Facebook or Google Plus or YouTube or…) and start to think of real world needs of Brands and Businesses and pull together metrics we need into one place (from all social channels).
There are small signs of hope.
Crowdbooster has a very interesting view of twitter:
You can see Retweets (x-axis) and Replies (size of the circle) overall and individual tweet perspective. So both Conversation and Amplification. The other two metrics are missing, but it is a start.
All My + is a early prototype of data from Google+ and provides three of the four metrics recommended in this blog post:
It is missing Economic Value. But you can get that out of Google Analytics or Site Catalyst in five minutes.
If you are a tool vendor… I would love for you to adopt the aforementioned metrics, and definitions, into your tool. All I ask for is a donation of one million US dollars to Doctors Without Borders. Doable?
What About Social Media Advertising?
If you are engaging in brand advertising on social media channels then the metrics you should solve for should be the first three. If you do a Promoted Tweet or Facebook Like campaign or whatever Google+ decides to come up with then you want to measure resulting Conversation, Amplification and Applause (of course only if you did not stink at your campaign).
If you are engaging in direct response advertising on social media channels then the fourth metric, Economic Value delivered, comes into play from a strategic perspective. It covers both the immediate value (revenue via macro conversions) and the longer term value (economic value via micro conversions).
For tactical reporting of your direct response social media campaigns, the metrics you'll use will be the ones I've recommended for all other advertising channels (paid search, display, affiliate, whatever).
Here's that picture, applied to SM DR campaigns:
Value Per Acquisition. Shoot for that.
It will be hard. The enchanting temptresses that are Clicks and Impressions and Avg. CPC will try to lead you astray. Resist their charms. Trust me.
Go for Ninja-hood.
Closing Thoughts.
Social media presents and incredible opportunity to rethink what it means to connect with and influence customers. You need to forget what has worked in the past (and that is why this is so incredibly hard to do. The biggest brands in the world embarrass themselves every day on social media). You'll have to rewire your brain.
In presenting new metrics for you to measure, what I'm really trying to do is provide a very small assistance in helping you think differently.
I hope it works.
Update: Bonus:
Erik Ohlen was inspired by this blog post to create a very simple, and effective, dashboard where you can track the four recommended social media metrics.
As I had stressed above, currently if you want to report these metrics exactly as defined above and from ALL the social channels mentioned then you have to do so manually. Small price to pay for communicating the actual impact of social media to your management right?
Sheet 1 is the dashboard itself, with instructions. Sheet 2 is where you type in the raw data. Could not be simpler.
Download: Social Media Metrics Dashboard. Adapt it to your business. Rock a lot!
Update2: (Apr 26)
When I'd written this blog post it was not possible to measure the metrics that I'd created here. But now there is a very good, still in alpha, tool that allows you to measure the metrics recommended in this post. It's called TrueSocialMetrics.
Here's a sample of what the report looks like…
Play with it, and get just the data you need to make smarter decisions when it comes to social media.
As always it is your turn now.
How do you measure the success of your social media efforts today? Got a favorite super lame or super awesome social media metric? Does one of the four (or all four!) metrics above resonate with you? What did I miss about social media? Is there a benefit / outcome / facet that I missed in my measurement framework?
Please share your feedback, critique, suggestions, and cool tools to measure these four metrics, via comments.
Thanks.
PS: A few helpful links for you:
A couple of my older posts with thoughts on social media measurement:
~ Social Media Analytics: Twitter: Quantitative & Qualitative Metrics
~ Viral, Social, Sentiment, Mobile: 4 Delightful Web Analytics Solutions
A post how to segment your social media data in Analytics (includes a downloadable advanced segment):
A comprehensive post on the most important gift you can give your business, compute economic value of your digital efforts:
Via 
Nice post, Avinash! It highlights the need to have clear business objectives and a logical framework of how the activity in the specific social media channel will roll up to something that matters to the business.
It seems like "economic value" as you define it understates the potential for social media, as it relies on campaign-tracked and referral traffic tied back to on-site conversions. While there isn't a ton of data (Foresee did a study during the holiday season last year, which is all that I'm aware of), it seems reasonable that there is a lot of engagement in social media relative to a brand that *does* contribute economic value to the brand but that does *not* ever include a clickthrough from the social media channel to the site. Rather, the conversation in social media influenced an in-store purchase or influenced a subsequent visit to the site that wasn't initiated by a social media clickthrough. That makes it a real bugaboo to measure, I realize. So, if the "referral/tracked" measures demonstrate sufficient economic value to drive continued investment, great! If it doesn't, that doesn't mean the value isn't there — it just didn't show up easily in GA, Sitecatalyst, etc.
Tim: I could not agree more with you.
Economic Value is just one of the four things I recommend people measure. Conversation, Amplification and Applause help measure the other bits of impact (which translate into, if nothing else, lowered marketing / acquisition / customer retention costs).
But without Economic Value most Social Marketers / Social Media Gurus will get laughed out of any decent sized company (as Jonghee highlighted in a later comment). So we show value, get permission/investment to do the right thing correctly, rinse and repeat. :)
Wonderful people in the ecosystem such as yourself I know are working very very hard to tie all other bits of Social Media value into a quantified number. And that work does not stop.
Avinash.
So totally true, Avinash. All your posts are great…and this one rocks again. I wish there was that one comprehensive measurement solution. Someday.
Stephen
There is already a solution – klout.com … check it out if you have not already.
"Your selfless social media contribution comes back to assist you in driving valuable business outcomes."
It is simply a paradigm shift. The firm is set up to make money not to care about customers (regardless of what each and every firm claims – sometimes in silly videos about how much they care about the mission/vision/whatever).
Actually caring about customers (like you would yourself and stop giving them shallow unpalatable – I'm being nice – content). This is the paradigm shift.
Thankfully this article may help people choose the right metrics but without the right mindset it won't really matter and customers will just be $ signs. I for one thank you! This means I can truly differentiate my own business and I can take a deep sigh of relief because I don't want to give junk to my customers that I wouldn't want. I can be customer-centric (truly not just hyperbole!).
Thanks firm!
Wonderful article Avinash, as always. Among the four metrics, the most important thing would be economic value.
My real-life challenge on finding economic value is that most of the case eventually it will be based on certain assumptions and making reasonable assumptions is a key to the right economic value. It's not easy but even trying it would be valuable since we can learn a lot during the process.
Tim mentioned the value of non-trackable social media influence such as in-store conversion. I agree that we should not neglect that portion of influence, but in real corporate life, at first, I'd better use something we can measure and quantify to persuade management. If we do it correctly, we can justify the value of social media investment most of the case even not touching intangibles.
Jonghee
Isn't there substantial conceptual overlap between the first 3?
It seems like commenting, sharing, and liking are all manifestations of the same larger, underlying thing, no? (call it what you will: say, affinity for the content?)
Scott: The commonality between Conversion, Amplification and Applause, if there is any, is that they creating a set of permission based influence engagements with your 1st level, and through them the 2nd level, network.
Other than that I personally believe they measure different things and, more importantly, different outcomes. Thus provide different insights into how well you are doing.
Avinash.
If the underlying construct is something like 'how much the content resonates with me' or 'I feel an affinity for this content,' then 1, 2, and 3 are measuring essentially the same thing. They differ regarding the actions we take as a result of that affinity – and thus perhaps a level of commitment to the content (e.g., we're more committed if we share than if we just click) – but the state of mind for the actions is the same, no?
Hi Scott, Avinash,
I agree that the state of mind for the actions from the audience may be the same, but there are several differences in the magnitude and the expected follow-up response from these actions that should trigger different tactics on how the marketer should respond, which I believe is what really matters.
For example, I think one big difference that lies between Conversation Rate vs. Amplification Rate, is how much the marketer himself is involved after he published the initial post. If you publish a really viral post, it may get shared many times, spread across several network levels, building strong brand awareness for you. However, without your direct involvement in a conversation about the content, the awareness itself may not have enough content to turn into brand trust that may have been built with more direct engagement. How that actually affects the final economic value of course, depends on each business or situation, but it sets up potentially different tactics on how marketers can take action.
Avinash
Thanks for sharing this post.
Since you ask, we are measuring conversation for weblogs so we can benchmark them.
However, what you call conversation (see above) we call engagement (kind of pre-step to conversation). We then distinguish between author replies, reader replies, and so forth.
For me having a conversation needs a kind of: write – read – write/reply – read – write-reply.
At such a level we can then measure things like reader who commented about a reply to another reader’s comment
So in turn My.ComMetrics.com offers the blogger a :
– raw engagement score
– reader versus author engagement score
– reader conversation score
– author conversation score
and so forth.
I hope this is useful.
Urs: Engagement seems to mean nothing and everything so I'm personally biased in terms of how best to communicate what is being measured effectively.
I concur with you on the metrics you mention in your comment. My good friend Joost de valk had created plugin for WordPress that measures those metrics (based on post I had written – talk about circular references!).
http://yoast.com/wordpress/blog-metrics/
Avinash.
Avanish
Thanks for this information. I did not know Yoast had a plugin …. checked it, it does some of the things we do and very nicely indeed.
Thanks so much for this pointer
Urs
I agree that we need to stop pushing the one way communication marketers are used to with TV and radio. It's time for a conversation!
And I'm glad you've come up with solid ways to measure the results of social media marketing!
Wow, Avinash, we think along the same lines on the subject!
As we've been quite disappointed by current analytics tools on this very specific subject, we've been working on our (wisemetrics.com), which provide similar metrics to yours.
And a bit more… ;-)
For example, the "Community Engagement Rate" which measure members' bonds with each other (=average # of comments/post published by member). This way you can see if members are talking to each other (stronger value) or if they only talk to you.
Once again, thanks Avinash. It was bold and refreshing of you to be willing to deviate from your norm.
That being said, and far be it for me to pick a fight with The Master, I'm not so sure I agree with the Applause metric. My (personal?) problem with the Like & Company is it has no context. That is, maybe even more dis-Liked something. IMHO, the Like in its current form only shows: of the people who saw *and read* the update, how many of those people moved the cursor across the page, hovered over a link and then clicked that link. While not mindless, it's pretty close, yes?
Therefore, in other words, it boils down to: The update made a number of impressions; some of those actually paused to read my update; some of those people might be pro-Like and bothered to click a link labeled Like; and it is always true that no one clicked the link labeled dis-Like.
I suppose it's better than nothing. We can have that discussion some other time :) None the less I think it is very important the true context of that metric be understood. The definition of impression is pretty loose and you also might not have a pro-like crowd and/or material that triggers a Like in what is often a playground-esque setting (i.e., people don't go on FB to do work, do they?)
Also, at the risk of repeating myself and at this point sounding like a broken record – if that metaphor is still valid :) – my belief is link tagging (a la Google URL Builder) is a useful tool that is too often overlooked. Tagging when paired with a URL shortener that provides analytics (e.g., bit.ly) can help to gather real data. For example, in using the URL Builder format I treat each update as a unique "campaign". That is, in a way, the link in each update is not really that different from a banner (for which is URL Builder's original intent) that drives traffic back into my site.
Please note: Aside from tagging the links back to your own site, you can also tag links to the other sites/pages that you share. The URL shortener's analytics will let you watch the activity on those links too. And while I don't yet do so myself I'm beginning to warm up to the idea of also logging the actual message posted. That is, trying to understand (for example) if a particular hash tag on Twitter seem to draw more clicks.
As you noted, today's social media measurement tools might not be idea but much like anything in this wild world wide web west, it's a work in progress. Getting involved sooner means figuring it out sooner. Getting involved later will only delay the development of the muscles necessary to do the necessary heavy lifting.
Mark: Think of it in layers. Applause > Amplification > Conversation. Each adds an additional amount of "pain" to the user to do, and hence provides more context and detail.
But in a world where pressing a button is sufficient to communicate "awesome" (like/+1) and "sucky" (dislike/-1, hopefully buttons coming soon :)), I believe that that action, small as it might be, provides insights into what people are connecting with and what they are not. First layer of insights.
And don't forget the impact what these simple button clicks then cause. SEO. Social Ads. Social Graph activities. So much more. It is not just a button click any more.
But think of it in layers of value.
Thank you so much for your detailed comment, and dissent. I appreciate that more than you can imagine.
-Avinash.
Hi Avinash, a nice thought provoking blog post.
It's good to see practical measures that can be calculated through traditional web analytics systems and you've covered those clearly and persuasively.
An observation is that your measures are towards the bottom of the funnel and don't look at top of funnel reach or conversations independent of website/blog posts. I know these aren't available in traditional analytics tools, but with social media monitoring tools it's common to use these to report on:
1. Conversation volume e.g. brand mentions or other market relevant keyword mentions tracked across media. This is similar to your "conversation rate" except you reference a specific post – the keyword mentions can happen independent of this.
You asked how to measure this conversation rate – all of the main vendors would claim to do this – this was our compilation:
smartinsights.com/blog/online-pr-social-media/online-reputation-management-software/
2. Conversation share – volume compared to competitors for target keywords – I think this is the most important type of measure you are missing here – you could have an efficient lower funnel but the brand is not discussed so much relative to competitors
3. Conversation sentiment or polarity – automated and/or human assessment of brand mentions which are positive or negative and need to be followed up.
I hope that adds to your framework – have you written a post/have a deck on the future intersection of site-based analytics with social media monitoring tools and social media platform based analytics e.g. Facebook Page Insights?
Dave
Dave: Three of the four metrics I recommend can't be measured using Web Analytics tools. :)
I'm unsure as to what stage of the traditional sales/marketing funnel they fall into. Mostly because I believe that social media is unique in that it allows businesses and brands to influence across the entire funnel (highlight, persuade, acquire, retain). Most great brands are doing that now.
Measuring "mentions" is something I've personally not valued, because that word/metric lacks any context. Hence I've gravitated towards conversation, amplification where specific context around "what" and "how" is present, and we can figure out "why".
The half serious joke I tell is that we are pretty soon going to reminisce about Mentions and Fans and Followers like we do about Hits (how idiots track success – to use KD Paine's lingo!). :)
I could not agree more with you on competitive analysis and gaining valuable context from that data. It is priceless.
-Avinash.
Oh my word…hits. That cracked me up. It does boggle my mind how many people, even the media, still report hits as being a reliable metric.
Stephen
Ahhhh but aren't "fans" just the equivalent of those who clicked "like" (at least in Facebook terms)?
Heather: That is true, but some Likes on Facebook are in context of comments and other Likes are in context of a page and others still are in context of a brand.
In our analysis sadly we will have to ensure that we accommodate the context in which we want to report the data, and then choose accordingly.
Avinash.
I loved this beefy post, thanks for the deep-dive!
I did want to just note one thing–regarding tracking of Twitter favorites—you can get an email notification for those, and then have the notifications sent to a specific folder. It's a hacky way to get the number, but it will work if you want to track favorites.
You've presented quite a lot of useful and interesting material here, but you've left out the celebrity factor. Simply put, the most engaging tweet or post from an unknown will not elicit the number of responses that any reasonably coherent tweet or post from a prominent individual will inspire. People enjoy the idea that they can interact with important people.
Social media has a kind of intimacy that is not really intimate, allowing people to offer apparent personal access while keeping the audience at arm's length. So, it may not be true that someone's media efforts "stink," but only the case that they have not managed to break into the rarified atmosphere of those people deemed by millions to be important enough to warrent a following.
Brenda: This might be an error on my part but I don't care about the celebrity factor. I deeply believe that any person (or brand) has an opportunity to create a hyper relevant network on social media and succeed with it.
Yes, the size of the network might be 100 people or 1000 people (vs 1 million). But if you do it well and you are able to create a social graph with the right 100 people then what does it matter that Ashton has 15 trillion?
My Conversation Rate is 13 on Google+ and around 40 on this blog. Ashton's numbers are probably more like 154,128 on either channel. But we are both solving different problems and, dare I say, we are both winners. :) The data helps us understand how to do what we do better.
Even if you have 20 people liking your Facebook existence, figure out how to solve for these metrics.
Avinash.
Avinash,
Thanks for your thoughtful reply. I was concerned that your metrics approach would be too discouraging for the little guy. I'm the little guy, and my clients are mostly even littler guys. Admittedly, Ashton has more of a following, but to us, you are also one of those people who enjoy benefits from the celebrity factor. However, your point that even small numbers can work is encouraging.
Brenda
Avinash – Great metrics and blog post as always. I get conversation, applause, amplification and economic impact for actions that occur in session.
But what about long sales cycles and repeat visitors. For those interested in tying social media to multi-channel funnels to see the impact of social across multiple visits, do you have any advice on reports/metrics to use to show how well social is assisting your other campaigns?
Josh: With analytics of all types (from all vendors) becoming more Visitor centric I personally believe that the challenge you mention is becoming less of a challenge over time.
Google Analytics's Multi Channel Funnel report (and equivalent reports in the tool you are using, if it has it) provide insights into this pan session behavior. Which in turn should help you communicate how Social impacts future behavior (especially if it leads to a Macro or Micro Conversion).
Avinash.
As a social media student, I found this article both interesting and timely.
I have to submit a discussion piece for tomorrow's class and your post has given me much food for thought.
Thanks, you've got my applause and my +1.
We've created a tool called Measured Voice that tracks FB likes and comments, Twitter retweets, and bitly link clicks on social media messages. Reports can be exported in csv.
We've been in beta for a while and are overhauling our architecture at the end of the year to make it faster and allow for gathering of impressions (amplification), Twitter favorites, and the addition of support for other platforms (we'll do Tumblr long before we bother with Google+).
It's really gratifying to read this post because it validates some of our philosophies: specifically that we need to be tracking metrics on a message by message basis – not generalized follower growth stats which do nothing but stoke egos. Thank you for writing it!
Jed,
Does/will your app present some of the metrics Avinash outlined?
Currently:
* "applause" via Facebook likes.
* "amplification" via retweets
* "conversions" via clicks on bitly links and Facebook comments (although I'm not sure if I agree that a Facebook comment always counts as a conversion)
In the future, among other things, we will add tracking for Twitter favorites, Twitter replies, and Facebook and Twitter impressions.
Thanks for asking!
Thanks for this post!
I wonder if there's not some more work to be done on simplifying or differentiating the first 3 metrics. Social media being the unique beast that it is, a +1 (applause) would also seem to fall into the conversation (react) and amplification categories.
I think there's value in all 3 independently, but we should see that each metric would plot roughly the same on a graph for a given message, just with different axis values. So, I'm wondering if there's not a (more) brilliant formula hiding in here somewhere for a combined metric. "Social Value" perhaps.
The 3 independent metrics should probably remain separate as you start to analyze a single message impact across different networks.
Thanks for making us think!
Justyn: Please see my response to Mark regarding relationships between the three metrics. There is a lose tie, but they are different enough.
It is tempting to think of a single metric (why make our life harder! :)). I am sure there is a magical "Social Value" metric out there. It would not shock me one bit that some are already selling it as the secret sauce.
I've personally found that Compound Metrics (/Calculated Metrics) have a regressive impact on a company's ability to 1. find insights 2. know when disaster strikes 3. figure out what action to take. Hence I'm biased on that account.
If you would like a bit more context please see #4 in this blog post:
~ Actively Avoid Insights: 4 Useful KPI Measurement Techniques
Thanks so much!
Avinash.
Avinash,
This is terrific, thought provoking stuff!!
Thanks for this thought provoking post.
You are really telling us to be more "social" and less "media" by measuring a 2-way interactions with our readers, instead of just a 1-way broadcasting.
Interesting there's such a bias towards believing there is value in social beyond what can be properly measured. Is it possible the value of social is exactly what can been measured, and no more?
Just wondering.
Jim: You never fail to bring a smile to my face. :)
There is an astonishing bias towards believe that there is an overwhelming amount of value in Social (beyond what can be properly measured).
Having engaged in the social web revolution with gusto (both as a creator and as a recipient) I've come to believe that there is value (beyond immediate economic), perhaps not as much as the SM Gurus might indicate but there is value.
I have a much more positive brand affinity towards Red Bull purely because of their incredible social media participation. I am not going to switch away from Comcast (inspite of their horrible DVR) because of how amazing their support is via Twitter. So on and so forth.
But short of running clean test and control experiments (there are such huge data gaps here, it is hard to do them) it is impossible to quantify value in a clean way.
Hence the stress on measuring Economic Value. If social media is done well by any brand that Economic Value will easily surpass any people, process, technology investments the company is making. That should tide us over until we figure out the value of the other stuff. And then we will rule the world! :)
Avinash.
I have no doubt there is some value in social beyond what can be measured, as this has been the case for all marketing since it began ;) The problem is this value is often situational, not too mention not properly measured using an incremental basis (as you point out).
For example, to small local businesses who do no other form of advertising, there is a huge amount of relative value to using social media, versus no advertising at all. Some advertising is much better than none, and since it's free, the incremental value created by (properly) using social is huge.
On the other hand, I wonder why social analysis seems to forget that people have to be aware of you to "Like" you in the first place. Further, it seems unlikely a person would "Like" a brand or product if they have not already experienced it, and are already a fan. If this is not true, if people "Like" a company even thought they do not (paid to Like?), then the problems with social go way beyond analysis…
But if true, , the number of "Likes" doesn't have as much to do with awareness as it does with size of customer base, and is much more aligned with tracking customer issues (retention, loyalty) than anything to do with awareness / acquisition.
Add the fact many companies are running lots of advertising designed to create awareness, and the incremental value of social as a "media" may be close to zero, or at least less than the cost to analyze the true value of it.
And this last, really, is the core of the issue. It's simply not possible to measure "all" the value created by any kind of marketing, and there are hugely diminishing returns as you try to capture the last bits. I think it's quite possible the optimism for "value beyond what can be measured" is less than the cost of measuring it *if* people keep looking in the awareness / acquisition field.
Folks who want to find this "missing" social value should start doing customer analysis, and look in the "retention / loyalty" area, where the whole idea of social is a natural, rather than a forced, fit.
Jim & Avanish – I feel like I have this conversation regularly at my company.
Sales: We need 100k likes on facebook. Our comptitors have it.
Analytics: Why? We're so much better off buying users through SEM or Outbrain than getting facebook fans, in terms of revenue and user loyalty.
What makes this type of debate so much more difficult is that Social is very high level of effort in a way that SEO is not (today). Partially because the tools aren't up to snuff, partially because there is a lot left to learn and because external factors keep changing.
But, I do think you are overlooking one crucial element to the value of social media, which its synergy with search. We continue to see examples which prove it to be a very strong relationship. For instance, a user shares a page to their friends on Facebook. Because the link was shared through our on-site tools, a custom tracking code is appended. 15 of this user's friends come back. Then, suddenly, we see a huge spike (30k visits) coming from Bing with that particular URL, including the sharing tracking code. Huge economic value!
But, what about the instances we can't track? When users grab the URL out of the toolbar and paste it into facebook and twitter etc…? This relationship to search is EXTREMELY valuable for a company who garners the bulk of organic traffic through search. But, the lack of transparency in this relationship make it very difficult to craft a bullet-proof argument for investing in Social.
Also, we just signed on with an awesome little start up called Argyle Social (argylesocial.com). They provide for link tracking, publishing, on-site conversion tracking, account or platform level analysis. It seems like they could track #1, #2, Part of #3 (FB not Twitter) and #4.
Thanks for the great post!
Jessica
One thing that is going to kick this framework into high gear is sentiment analysis. Sadly, no one has nailed it yet – automatic processing of language is a ridiculously difficult problem to solve.
Imagine being able to see something like "John Doe commented positively to your tweet, and 300 of his followers saw it, resulting in sales of $40". That would open up a crazy new world of measurement.
Thanks for the post!
David: It is a hard problem to solve and data today, not matter how good, is just not there. Human processing of the context of the words still wins for now.
But computers will get better.
For example here is a very good visualization created after applying sentiment analysis to the bible:
http://www.openbible.info/blog/2011/10/applying-sentiment-analysis-to-the-bible/
It looks pretty and has some very interesting trends. But even a cursory glance will identify imprecise classifications.
-Avinash.
Lots of great stuff to think about here, Avinash. Thank you for bringing this conversation to the table and for framing the topic so nicely.
I really like the idea of measuring the individual tweets for the type of content that causes better amplification or better applause.
I like your google analytics report for measuring economic value. My mind immediately wanted to see bounce rate in there to get a glimpse at missed opportunities.
Problem is originality of the posts are decreasing in blogging world!
Even though some re-post a blog, they change some text, but dont even bother to mention the source. The original writer doesnt get noticed, the person copying the post does. Lot of "duplicate" post blogs have spoilt the real blog world, which makes it worse day by day!
Great post as always!
On measuring applause on Twitter, what about favstar.fm? That allows you to see how many people favourited a specific Tweet (and who did the favouriting – is that a word?).
On the subject of secondary networks, have you looked at tools like Followerwonk? Quick way to better understand the make up of secondary networks – as well as identify which people you might consider following engaging with.
Great article as always Avinash, and the topic of social media metrics and measurement strategies is something I am quite interested in and am spending quite a bit of time doing ongoing research upon.
One question I wanted to get your thoughts on, since you mentioned micro/macro conversions on the landing-page /website as an integral part of the whole-story for social media measurement. When you defined macro vs. micro conversions in the past, you had defined it in the context of the website only, which can be used to help answer "why does my website exist"? I'm wondering, would you consider improving Conversation, Amplification, Applause potentially as micro-conversions as well, since it helps answer, "why does my online marketing efforts, including social media, exist?"
Peter: Very good question. I've defined Micro Conversions as things that add direct short or long term value to the business.
Conversation, Amplification, Applause absolutely add value to the business. But for now, while the ecosystem evolves, it is very hard to understand how the direct part of the definition is met. Hence for now I would say these metrics (and success they identify) adds indirect short and long term value to the business.
So I would not consider them to be "micro conversions." But that does not mean that we can't set goals for them or that we should not segment and report like Analysis Ninjas and understand how to improve them.
Hope this helps clarify the issue.
Avinash.
Avinash,
I like the three metrics you propose and share your aversion for aggregate metrics. However, I'm not sure whether the fourth metric isn't falling a bit short.
You seem to imply that you use social media to drive traffic to your site which will then create economic value. My brand manager colleagues would argue that the value is created by their beautiful facebook pages themselves (i.e. positive impact on brand perception and purchase intent (which we can't really measure…)). They don't even expect consumers to go to web pages. Any idea how we can get management's attention for social if we don't use Google Analytics goals?
On another note: what do you think of Facebook's new "people talking about" metric? Where does this fit in the framework?
Thanks,
Daniel
Daniel: Beautiful Facebook pages are not the end. They are, if they deliver relevance, the starting point. Then you measure Conversation, Amplification and Applause to understand if your beautiful Facebook pages contain beauty that is just skin deep. :)
Those three metrics are squarely focused on judging if your Agency and your Company are "doing Facebook right."
I've stressed a couple of times that Economic Value will capture only a fraction of the value delivered, but it is a real value to the business. If you can't show your management that your social strategy is somehow tied to business success then it would make life difficult.
It is also possible that your websites don't deliver any value to the business. In as much having a social strategy that provides a path for outcomes for some of your audience via your websites might not be an option. That would be quite sub optimal.
Regarding "People Talking About"… it is a compound metric that mushes a few different things together. I am a little bit biased against compound metrics, please see the links in the above comment.
-Avinash.
PS: Please also see Jim Novo's comments on measuring "social value" all by itself.
Great, thought provoking post. Thank you.
Just to add insult to injury, I copied this url into an email msg that I shared with about 10 folks who are important to projects I'm leading. They will likely share my email with others using their fwd button. Who's gonna measure that? ;-)
Really found this on target. Will follow you to learn more. – Ben Waxman
Ben: Pure copy/paste sharing of the url can't be tracked easily. If those people use Outlook then they show up in Direct Traffic. But if they use modern email programs like Gmail then they will show up just fine as email referrers (and hence trackable! :)).
Also if you use the "Email to a friend" link on most sites then they contain tracking parameters. So trackable.
Finally if you just copy and paste the blog text then that is also trackable, please see this post:
https://www.kaushik.net/avinash/viral-social-sentiment-mobile-data-web-analytics-tools/
But I've disabled it because I personally find that annoying and don't want to foist it on my readers something I don't like myself.
So you are right that there are limits to what can be tracked. But we can track more than you can imagine. :)
-Avinash.
PS: I do realize that you were probably joking a little bit in your comment.
yep, just a bit.
Provocative and attracktive ideas, Avinash and Ben (Hi, Ben. It's been a while! Too long!). For managers who are non-averse to aggregates, just think of the possibilities: only 4 more metrics away from a tool you could call the "8-Track Ply-er". That'll grab 'em!
But for me — who lives on the other side of the tracks — who is not hugely knowledgable about Marketing, it seems logical NOT to listen to all those managers who want everything to square. There's room for intuition in business, no? (Mr. Apple had said as much.) Some businesses do social media because the people there like it — I mean the old fashioned way of liking something: it's what they do. Or to borrow a line from a P-Funk track, "We do this and this is what we do. We Do do this and what we do to this, we do to you." So, go ahead and do social media if that's what you do, but don't expect an accounting panacea.
My grandfather used to say "Figures always lie and liars always figure." That may be a bit too far for me. I believe you should study your stuff and shoot for quantitative and qualitative measures. The absence of self-study is an abysmal tunnel.) But ask those managers who stare at the bottom line to try a social(logical) experiment: Write down and share with their colleagues how all your products, projects, plans and ideas fail to meet their intended goal. Then compare that with your quarterly reporting. The truth lies/lay somewhere in the middle.
Fantastic post, Avinash!
I really like this as a way to measure the overall impact of an organization's social media initiatives.
On a tactical level, I also like to segment by overall topics, or themes, as well, so that I can see what's resonating with the followers/fans/friends.
Thanks for sharing Avinash.
Really like Jim comment; there will always be a brand awareness dimension in any broad marketing campaign (display, social media, TV). Bottom line, you always need a way to measure effectiveness of a campaign (money rule – here comes KPI #4).
For the tool part, though I'm not a social expert (trying to focus on WA), I have to think about Social Media from a reporting POV. So far, best tool closely able to measure KPIs #1-#2-#3 is SimplyMeasured (though it doesn't support Twitter favorite count yet).
Cheers!
Very informative article Avinash!
You might check out a company from Durham, NC called Argyle Social. In addition to various other tools, they provide a very nice interface you to track/monitor the four metrics outlined in the post!
Thanks!
Great post as always Avinash!
I've been drumming the point home at my company that we need to move from a model of a race to the highest number of fans to quality engagements.
The first step is to find a tool which will deliver simple reporting on key metrics similar to what you discussed on our most used social platform (for now) Facebook.
I've found a great tool at socialbakers.com which amalgamates the following metrics into a single page 'score':
– Fan Growth
– Content (frequency, diversity of content posted)
– Interactions (number of likes and comments)
– Quality (length of post)
I can also see the 'Key Influencers' which help me identify 'Super Fans' and brand advocates.
It doesn't yet provide sharing data but I expect this will come.
The best part is that I can add a number of brands pages into the platform and compare based on the above scores to see how well is brand is really doing.
It is the classic case where brand A may have 100,000 fans and a poor score while brand B may have 20,000 fans but fans who engage, comment, like and share actively on the page.
The holy grail as you say is to have "one tool to unite them all" and it is badly needed.
Until then we're going to use Social Bakers in combination with the new insights metric – 'people talking about this' and marry this with our Google Analytics report which reports which (other) social channels are delivering against our key business metrics which we've assigned a score to.
We're busy with setup right now but I'm really looking forward to being able to point to the numbers and divert marketing spend accordingly.
I can see Google buying out a combination of these vendors or perhaps one vendor who has cracked the problem for the most used social platforms and building this into the next GA.
As ever, inspiring stuff!
Michael
Michael: Thanks for sharing the tool, I'll be sure to check it out.
The challenge with Facebook only tools is that they fall into the trap of just doing a data puke of Facebook data rather than focusing on what happens after a brand posts on Facebook (the four metrics outlined in this post).
Any tool that looks beyond that would be most delightful.
Avinash.
This is a great post!
I would just to bring up two things:
– I agree amplification = shares, but also Liking and Commenting can amplify the message.
– I love the idea of the applause rate, but It will start being accurate once Facebook, Twitter, Google+ incorporate the Unlike, -1, etc. Youtube incorporated this so time ago and I think is much better to really find out the tone of the engagement, don't you thing?
Passion makes the difference. juanmarketing.com/social-media-buzz-viral-marketing-consultor
Me, Stephen Colbert and Nancy Pelosi… well, that's a first :) Thanks, Avinash. Big LOL from me :)
Thank you for yet another well thought out and insightful post. I particularly enjoyed this one because you’re opening people’s eyes to the idea of measuring social media impact across channels instead of evaluating individual silos. I believe that this concept is at the core of true social media analytics and should be the direction that is taken by any organization who truly wants to understand the impact of their social media efforts.
The concept of cross-channel metrics is a key component of a new social media analytics tool called Social Snap that my company is releasing next week at eMetrics NYC.
While the metrics that we have currently incorporated into the tool aren’t the exact calculations that you have outlined, they are based on the same concepts but with slightly different calculations.
Social Snap currently incorporates metrics from social channels like Facebook, Twiiter and YouTube, as well as Google Analytics data, social media monitoring data, and custom data sources, such as CRM data.
Social Snap has also been built with the idea that each organization has unique business objectives that are tied to their social media efforts. Because of this, the tool is completely customizable, so that users can create a set of custom reports and dashboards which allow them to truly understand how their efforts are impacting their business.
Cross-channel analysis is just one piece of Social Snap’s functionality. If you’re going to be at eMetrics NYC next week, it would be great to show you the tools full capability. We’ll have a booth in the exhibit hall as part of the newcomers pavilion. If you won’t be there, hopefully I can show you a demo at some point and get your feedback.
Thanks again for the very insightful post.
Nathan
Nathan: Thanks for recommending your tool. Social Snap seems to be doing a lot, there is a ton of data from the social channels that one gets from many other tools (perhaps too much data, and variations on metrics that add to the too much data!).
The integration between some activity on the site is wonderful, as is the categorization into positive, negative and neutral (always such a hard thing to get right).
But other than "Retweets by me" all the metrics mentioned in this post don't seem to be in Social Snap. Perhaps you are evolving in that direction.
The challenge for any vendor in this space is to figure out how to bring everything in and yet stop it from becoming such data puke that the users don't know what to do. Worse the handful of metrics that matter don't exist or are effectively hidden. *All* vendors face this problem, and will have to solve it.
I wish you all the best.
Avinash.
Thanks for this, Avinash. I have been struggling with "monetizing" social media traffic. I used quotes because it's about more than just financial business outcomes.
A lot of the times, I'm pleased with social media traffic that has more than one pageview on their first visit. Or better yet… RETURNING visitors with social media as the point of first referral.
Working effective measurement of social traffic into your multi-channel funnel is good stuff. I spend a lot of time thinking about that; moreso than looking at the numbers independently.
Hey Avinash,
Once again increasing my daily data insight intake from kaushik.net :)
Some thoughts to share on how one can frame their approach:
– Customers whom "Know" about your BRAND
– Individuals whom "Know but still not convinced" about your BRAND
– Individuals whom "Don't know" about your BRAND
– Individuals whom "Doesn't care at all" about your BRAND
Once the above framework is run through the organization and start to understand the "whats", the "whys" and "hows" of their subset of customers and individuals,and start to plug in the four points you kindly share with us namely Conversation Rate, Amplification Rate, Applause Rate and Economic Value into how could a BRAND seek to approach or engage them based on their subset group.
And in case you or your reader missed out, Assistly (http://www.assistly.com) is awesome tool to put together the above and have that be actually implemented into an organization to walk the talk and not merely paying lip service to social media and the next "Fill in the blanks ____" silver bullet. Which goes back to having a BRAND to think not about "Social Media" but focus 101% about their "No.1 Boss – the Customers" instead.
Rock on,
Deric
Deric: I agree with you completely on the value of segmenting the Big 4 Social Metrics for the customer personas you mention in your comment. Sometimes on social channels it is hard to figure out such information about people who engage with us. But that should not stop us!
I would also expand your segments into other customer engagements of the more traditional ilk. Prospects and Customers. Or attract, acquire, retain (serving existing customers is such a great use of SM).
Thanks so much for the delightful comment.
Avinash.
Deric,
Couldn't agree more. Thanks for posting your reply. It's a nice supplement to what Avinash wrote. I find that many organizations don't ask enough questions of themselves before charging out into the market with their shiny ideas. Deric, you're comment nailed it.
Cheers,
Brendon
Hi Avinash, I want to correct the record about Social Snap’s coverage of the metrics discussed in your post but first I need to thank you for pointing out something we needed to be reminded of: Our web site is woefully behind in conveying what the tool does. Our attention has been focused on development and we are behind on marketing. So, thanks to you, an update to the features pages on our site has just been added to the top of the marketing priority list.
I also want to point out that Snap does indeed collect hundreds of metrics from social channels, however, we are very conscious of the data puking issue. Social Snap goes way beyond the individual metrics collected. Snap collects these metrics in order to combine them into a robust set of derived metrics similar to the concepts you’ve discussed in this post. What's more, the user can select any of the single or derived metrics available from all the various data sources in the system (Web Analytics, social channels, buzz data, custom sources) and create the reports that mean something to them.
Regarding the specific metrics you define in your post:
1.) Conversation Rate: We provide a very similar metric in Snap. We collect individual metrics in each channel that captures how much the "audience" is interacting with the content posted and combine those into a cross channel input/output ratio metric. We also provide a total audience engagement metric across channels for the purpose of tracking engagement growth.
2.) Economic Value
Snap provides reporting on soft and hard conversions and goal value very much the way you have described it here. This data is reported on in several different places in Snap. In the campaigns section, it is combined with campaign buzz and retweets to provide a more comprehensive view of what happened before the visitor arrived at the site.
3.) Amplification: We are collecting the metrics you describe on a per channel basis (not Google Plus yet) but we have not yet combined those into a metric like Amplification rate. We like the idea!
4.) Applause Rate: This one is not currently available in Snap although again, we are collecting some of the individual metrics you describe. Great idea and we will consider rolling up into a derived metric.
Like you, we believe that calculating derived metrics (metric combinations) across channels is absolutely necessary. We've created many derived metrics in Snap and will continue working on new ones. This is a task that will never be "done" as the data available for collection changes, new thinking like this post emerges, and the channels and web analytics tools evolve. We’re very happy to see you writing about this topic and welcome the opportunity to give you a demo of the tool.
Does anyone have experience using Simply Measured? It's a social tool, but I've only viewed the demo video–seems to have some of the data Avinash discussed in this post.
Hugely insightful post, Avinash! What I found very interesting, in relation to my own experience, is that all your metrics are things I've observed QUALITATIVELY as a solo blogger who sells an eBook online and is highly active on Facebook.
My friend Marsha and I started a group on Facebook called "Strong is the New Skinny" or "SINS" and have seen it grow into a highly active community of over 27,000 in about 1 year.
Some of the things I've observed as one of the two administrators of SINS:
Raving, evangelist fans
Our followers INITIATING posts that either praise the community, the people or REQUEST a purchase link to my eBook or Marsha's Spreadshirt store. (Now THERE'S a metric! Number of customers REQUESTING you to help them buy and initiating a buying conversation!)
Sometimes massive responses to the links we share to blog posts (or just status updates) and many shares of them from there
This past weekend, I had a vendor booth at a fitness event and had a girl drive up from about 3 hours away to hang out with me and my girlfriend because she "knew" us from Facebook. She brought us each a T-shirt from her gym, bought an autographed hardcopy of my book and hung out with me at my booth for several hours on Sunday.
This is something that's definitely worth a measure for solo bloggers/personalities – how many relationships that begin online are moving offline or blending between the two? As another example – I found YOUR work online and now have a strong desire to attend a conference that you're speaking at and have you sign one of your books for me – It's a transfer of online connection to offline…
Thanks for the great post!
Adam
Avinash I love what you have posted here. I know of a lot of sites that measure one or two of these none that measure all.
It has been my opinion for some time that there are many times more lurkers in social than there are engagers and by measuring the interaction with the content you get a much more accurate picture of influence than than measuring 1 to 1 engagement with the person.
The question I think is how do you enventually gain a level of influence with your audience to the point a large number act when you request it? Our testing seems to indicate it comes from 1) gaining attention through sharing of good relevant content to the given audience 2) Creating insightful orginal content that the audience can identify with and 3) Good old clear messageing and call to action.
The kicker seems to be audience fragmentation. For a specific niche like Hubspot it is easy for them to know their audience and what content to produce and share, however for a retail store like say a car dealership this becomes a bit more difficult. Only a small part of the audience is interested in automotive related things at any given time and those interest times can be far between. The question we are working on right now on is what interest topics do you share & write content on if your audience is fragmented and generally only interested for short periods of time?
Thoughts?
Larry: Your three part advice is very prudent, and my own experience bears that out. I would just add one slight nuance. You can't just "pimp." IE have all your social participation be about you, your products or always with a call to action pointing back to you. I've developed a best practice for myself, 1 aggressive pimp to 9 giving back. Sounds reasonable?
That last part of your comment is the core reason why number of followers etc are horrible metrics. It is not about the size of the audience, rather the relevancy of the audience. My personal path has been to collect an audience (AND participate in an ecosystem) where there is a large overlap between what I can contribute (what I am good at) and what the audience/ecosystem will value. 1,000 hyper relevant followers are a million times better than 100,000 non relevant ones.
-Avinash.
Avinash couldn't agree with you more on 1000 raving fans there was actually a good post on that some time back but I cant find the link now.
Actually step one in my process above starts with sharing content that has nothing to do with you just of interest to your audience, that gets their attention. Only after you get their attention can you then layer in your own content to build authority that leads to eventual action.
In that sense you 9 to 1 ratio is very prudent we just push more up front.
LB
Once you determine a number using the metrics above, how do you know, on a scale from 1-10, what is a good number?
Is there a scale somewhere?
Barbara: I'm afraid there are no benchmarks on any of this (mostly because we don't even have tools to measure half the stuff we want to! :)).
But, and we can't say this for all the work we do, more is better. More Conversation, more Amplification, more Applause and, lord knows this is true, more Economic Value.
So benchmark against yourself and aim to get better over time. (I think you have my book Web Analytics 2.0, there is more in Chapter 11 about this.)
In some cases you can snag your competitors numbers easily. For example the Google Plus tool above also measures your competitors (just type in their id instead of yours) and you can start to benchmark against them.
Overall given where we are, early in the cycle, just aim to get better each day.
-Avinash.
Wow great article Avinash,
I came across this post from another interesting post you had about CI tools. I love how you have links from your other topics to new topics that even more interesting. Great work!
Now to my question….
I tried to find tools and API's to get me these social media metrics and my competitors. Are there any tools out there that will give me this level of detail at the post level? I am using Supermetrics but I could not get the level of granualarity on my competitors.
I ended of having to build an web scraper via excel VBA that collects these metrics at the post level. It also lets me scrape for my competitors social pages to see what they are doing so I can set benchmarks.
You mentioned if there were any tools that captures these for you and I would love to share this tool I created for everyone but wanted to know your thoughts on web scraping these social media channels.
Thanks for the article!
Levett: You can use this one, it measures the metrics mentioned in this post:
http://www.truesocialmetrics.com
Avinash.
Thanks Avinash,
Signed up for it and testing it out now.
I have 2 questions:
1) Why are the conversion, amp and applause rates calculate with total posts and not the total fans/followers etc for each channel? If we use these calculations wouldn't a site with a huge fan base always get higher conversions, amplification and applause rates?
I read on Facebook that they calculate engagement rate by sum(likes,comments,share,clicks)/ total reach
Simplymeasured calculated engagement rate as sum(likes,comments,shares)/ total fans.
http://simplymeasured.com/blog/facebook-metrics-defined-engagement-rate/#sm.1uhpvkpckddl11uzh9ceralzg
2) I am playing around with the free version and it doesn't seem like this tool gives me post level data? I want to know which of my competitiors post had the most engagement this year. Any way to achieve that?
Thank you for this blog as I always get inspired when reading your posts.
Thanks,
Levett
Semrush's new social media monitoring tool gives likes, comments and share by each different social media channel which is helpful for calculating competitor metrics.
But I have to agree with Levett – a huge fan base that hardly posts and gets crappy engagement for the size of the fan base will show relatively high engagement. E.g. Petsmart has 2.3M followers on Facebook a but posted 7 times in the last week and had 800 likes, shares comments whereas Chewy has 380K followers on Facebook and posted 17 times on Facebook and had 2,800 likes, shares, comments.
Clearly Chewy doing better but the metrics are only marginally better for Chewy and better for PetSmart across Instagram and Twitter because of sheer volume of fans even of only a tiny fraction engage.
Shouldn't there be a metric for % of active fans metric as well?
As always, loved the detail and the clear road to usefulness.
BTW – Your blog is a perfect example of social media done right. It's the interaction and the conversation that you create – the writing style, the responses to comments – it's all social and creates a connection.
This is what most companies should embrace and understand. I don't see this kind of execution normally and thus the results are not as impactful.
Thanks again for the great post.
Wonderful article Avinash,
In the competitive marketing world, however, there’s a data-driven way to determine and understand extensively more data than is possible with traditional advertising. We have to do different type of measurement for make the better performance of our website. You have discussed the various great things like Amplification Rate, Applause Rate which are quite new for me and I love this concept.
Social media is great platform but how to enhance with it GA is definitely a tough job. You explain in a very simple manner. I am using Radian6 for tracking my social media updates and it works well for me. I have just read your interview in SEOmoz and it is simply awesome for all Mozzer including me.
Currently I am working on Facebook page with GA workaround. Someone also prefers to use Facebook insights for this. I am bit confuse where to start. It’s really good if you drop your feedback on this.
PS: Do not forget to measure brand impact over time along with direct response. We have to see the engagement level on timely manner.
Avinash,
Very pleased to have discovered you thanks to the SEOmoz newsletter with the interview from I Love SEO. Looking forward to reading more and learning from you.
Chip
Thanks, from Portugal.
This is great Avinash! I'm using something similar but I will take inspiration from this post and update my current score card.
I've two things to add though:
1) I find Spredfast to be an excellent tool to aggregate and report on the raw metrics from each platform. It doesn't do the calculations though, you still have to do that in excel. However, I'm sure some tech savvy person could figure out how to pull that API-data into excel automatically.
2) Measuring the economic impact from referrals means you need to track each link. When this link travels through twitter, facebook and linkedin shares, is it getting all the attribution it should? I'd love just to be able to look at referrals and assign some rules to the referral path but unfortunately Facebook won't give you the full referral.
Just my two cents!
Br, Joakim
Joakim: If you add campaign tracking parameters to the links you share via social media then you should be able to track all the visits through those channels cleanly and completely. Across platforms and devices.
This makes it ever more easy to do things like measure conversion rates and economic value. Additionally you can also do multi visit analysis using solutions like Multi-Channel Funnels.
-Avinash.
Me again, I've been through your post one more time and I'm not sure about the "Applause Rate" and how you've picked metrics from each network. I think this is a difficult one. What you want to measure here is really if people show that they appreciate your content or not. Each network has their own function for this, are we using it that way though?
I consider myself pretty social savvy but just recently discovered how the favourite-button could be used (ok, it's a very opinion based comment of mine this one). I' suggest to look at "Engagement" instead, this would be all interactions with your content such as likes, comments, ratings and clicks.
Joakim: The common thread through all the "actions" chosen for Applause Rate is that they are single click "thumbs up" across the different channels.
I would look less for that each are called and more at the action that is being measures. If it fits then group it, if it does not then not. :)
-Avinash.
I'm so sorry for the spamming here :)
Thinking out really loud here and came up with 5 KPI's to benchmark:
– Content relevance rate (Activities/Engagement) – This is close to your Conversation Rate but includes all Engagement metrics i.e. any interaction with content such as a re-tweet, comment, or like.
– Avg. Reach per Activity (Divide your total potential reach with number of Activities, reach is things like newsfeed impressions on facebook, mention audience on twitter etc)
– Network activity rate (network size / engagement) Viava lots of fans but how engaged is the network?
– Tot. Referrals to site
– Est. Tot Attributed Value
Kudos for your KPIs here.
In particular Network Activity Rate. If we follow metrics, such as the proposed conversation rate, over time, we do need to think about the growth in our follower base (in particular on Facebook).
If the conversation rate is growing, but at a much slower rate than our fan base, we should rethink our content sharing strategy.
Hi Anne-Signe,
My experience is that growth of Network Size is pretty linear and not affected very much by Conversation Rate or Network Activity. However, sales and traffic to your website are affected by the two first i.e. you're maximizing the value from your network rather than just growing it.
Br, Joakim
Incredible post.
It's the nitty-gritty of social media in web marketing. Maybe a bit overwhelming when trying to explain this to clients whom we want to set up with a social network and have someone in their business dedicated to engaging in that social media. But I think it's worth the effort.
Hi Avinash,
Once again, this is a great post. It definitely help social media marketers pull a brilliant scorecard that highlights the underlying value of their campaign this quarter!
While Conversion(engagement), Amplification and Applause are all measured through physical interaction, have you thought of the capabilities of measuring VIEW-THROUGH Conversion on Social Media? For example, friends in my social network viewed my post regarding, say "Dapuri White Coffee" but did not reply, like, or +1. However, 2 days later, they went to Dapuri.com and bought some products.
If we can measure at this level of granularity for this type of "interaction", it would be great!
Thank you
Alvin
Alvin: If you look at the crowdbooster screenshot included in the post, you'll see that they do try to show something like what you were looking for when they show "People Reached."
Kinda sorta view through. Perhaps "exposure" is a better term. Overall though in Social Media is is hard to measure that someone read something. We use proxies (like you see in this post) that are better in some ways.
I have to admit I am not the world's biggest fan of the View-thru metric. The challenge is that we are trying measure something where we have no clarity in terms of what happened. We are assuming that something was seen, we have no idea that they did.
One good way to validate that hypothesis is not use controlled experiments to figure out if something moves the dial. So for example in the context of Display Advertising pick some test and control sites / geos/ people / types of ads, run the experiment, measure if there is a material difference in ultimate outcomes, use that to decide. Much much better than using view thru.
-Avinash.
Hi Avinash,
First of all congratulations for this great great post :-)
I have one doubt, it's about the dashboard social media metrics.
How do you calculate the economic value per source of visitor? Do you obtain it from GA?
Is it just the figure you give to the goal when you are setting it in GA?
I think this is how you get it but I'm not 100% sure.
Thanks,
Avinash, we are using Campalyst – one of the only tools that sync Facebook and Google Analytics and provide a dashboard view of how every post converts and completes the conversion goals. Great stuff and worth looking at.
Katerina
Hello,
I have just started working with a client who has more than 80 twitter accounts and about 20 facebook accounts, the dashboard looks great but would I just add on all the twitter and facebook accounts?
It all seems set-up for just one account when what I am facing is very different!
Kind Regards
Carol
Carol: This is a common, and complex, challenge. Thanks for bringing it up.
I am personally not for aggregating all the data up into one rolled up view. All the good and the bad get mushed up into a unrecognizable goo and it becomes hard to understand which accounts are doing a great job and which ones massively stink. That's my broad point of view. Collect all 80, but don't do 80 dashboards, do them for the top five and bottom five. Call the rest "middles that are ok but should do better to show up in the top five." :)
If you do aggregate across 80 accounts then still make sure you call out the good ones and bad ones separately so you have some context. In the end you do want to to drive some action, not just report the data.
Hope this helps.
-Avinash.
Nice post Avinash. Although I believe the economic value component of your matrix needs to be taken very carefully (or perhaps even with a few grains of salt…) the conversation component (which is engagement) might be the holy grail since we are trying to measure the connexions between people, their ideas and networking relations. The only tool that I know of that tries to measure the conversation rate is from a company called Nexalogy : nexalogy .com
Thanks !
Patrice Leroux
Patrice: I appreciate your point of view. But I have to admit that while you are right that Economic Value should be computed carefully (see links below on how to), I do not believe it should be taken with a grain of salt. There is some art involved, it is mostly science. And it is a solid number. And if Social Media efforts don't show some Economic Value then in the long term they are doomed. No matter how much "engagement" happens.
As the post outlines… I am totally in agreement with you that the conversation, participation, applause, amplification add a tremendous amount of value. We can't fully comprehend or measure it for now. But there is value. As a brand that actively participates in Social Media I see it, and feel it :), every single day!
-Avinash.
Avinash – you should check out the metrics available in ThinkupApp. It's a cool application that captures and measures your social activity on multiple platforms, and has just the sort of data you mention. I couldn't find a hosted version of Thinkup that's open for anyone can use, but it's pretty easy to install yourself on the free version of AWS.
A redesign of the dashboard is coming soon, some details are here – expertlabs.org/2011/09/designing-thinkup-20.html
Kevin: Thanks for pointing out the tool. I think it looks interesting.
From the limited view of the screenshot though it looks like one of many tools that just go grab the data and then a bunch of it is represented in that image. Some of it full of promise (perhaps this is what you mean when you say redesign). But much of it suffers from the problem that all tools (from Google or Radian6 or whomever) suffer: Just puke a bunch of data out. :)
My hope is that vendors will understand a lot more about what makes the experience unique, what's actually important to measure, and then primarily guide the user to that. That would be adding great value.
I concur with you that the fact that they have access to all the data gives them a leg up on computing Conversation Rate, Amplification, Applause and Economic Value.
Avinash.
Thank you for the post. Simply brilliant!!
Nice post Avinash!
So I have a question; There are differences at the moment of establishing metric between different sectors? For example, between public institutions and health there would be differences or ultimately, we all measure the same thing?
Thanks for your post and job
Gemma: In this case I've defined these metrics based on what's unique about the medium and how people behave in it. Hence the metrics won't change for different types of companies (we all participate in the social sphere in the way).
There might be some change to the last one, Economic Value. Clearly a government website will have a different macro and micro conversions and how they compute economic value. That will be different again for a newspaper or a pharmaceutical company etc.
Here's a blog post on rules for choosing the best metrics: Web Metrics Demystified
Avinash.
Hi Avinash
Wow, this is so timely. I'm working in UK local government as a Senior eCommunications Officer and I just managed to argue for a social media metric to be included in our performance measures. The rate at which people engage with us, whether they're happy or sad, is a measure of their satisfaction with the experience of engaging with us. In other words, if we don't actively participate in the conversations that spring up and respond in a satisfactory manner, they'll quickly lose interest or even hope.
So lack of chatter = bad engagement, lots of chatter = good engagement. And like I always say, always embrace the bad comments, how else will we learn to improve.
Jon: Congratulations on making the UK government a bit more data driven! All governments in the world need that type of evangelism.
In my post my overwhelming emphasis was on measuring what happens, in your case for the government's contribution to social sphere, after you participate. Amplification, Applause, Conversation and Economic Value.
"Chatter" is important (Conversation and Amplification get to that). But I've had less favorable experience calling the existence of chatter as "engagement" or that a lot of it is good. We certainly want to measure conversation rate, and we want to know the tone and texture of it so that we can adapt.
Thank you,
Avinash.
PS: Two posts you might find to be of value:
~ Web Analytics Success Measurement For Government Websites
~ "Engagement" Is Not A Metric, It's An Excuse
Avinash,
Saw this when you originally posted it, which was incredibly timely as we were finalising our own social media performance KPI for brands.
We've now launched this (sociagility.com/print) and, I'm pleased to say, were able to incorporate your suggested metrics into our model and methodology (we're not quite able to make the donation though – yet!). The only exception is economic value, but that's because our approach is comparative and economic value is specific to each organization. It can easily be applied by each brand once they know it, which as you say, isn't that difficult.
However, we have been able to link our measurement of social media performance to recognised measures of brand value and growth (which many organizations take very seriously), which should add further power to the social media investment argument.
I'd be more than happy to explain our model and methodology further if you're interested.
Niall Cook, Advisor and co-founder, Sociagility
Excellent Article, Thank you very much :)
Great post, as always !
With Conversation Rate, have you considered posts initiated by the customers. These can be customer service, product info, feed back, sales enquiry, nevertheless real conversations and the brand engage in conversation.
How would you measure Conversation rate in Twitter – any thoughts ?
Andrew: Conversation rate is # of replies per tweet. It is a bit difficult to measure for Twitter but just today I was trying to use Twtrland and it shows a pie of overall replies:
http://twtrland.com/profile/Avinash
Klout is also evolving, keep an eye out on that as well.
Avinash.
Hi Avinash,
Nice post, but I have 2 questions:
– Don't you believe that social media can generate marcro conversions as the arrow in the visual is only pointing to micro conversions?
– Wouldn't it be better to show total value as well in the dashboard of Erik (instead of value per visit), especially when you want to show it to your boss.
Bart: Yes of course it is entirely possible. But in my humble experience (across B2B and B2C) those macro conversions are tiny in number and hence often "prove" that Social Media is not worth it. But here is the wonderful thing, you can see what it is for your business and decide what you want to do with it!
The nature of the beast is such that it is people oriented so I like that it was showing value per visit (executives love that!), but it is prudent to also, as you suggest, show the total value.
Avinash.
As someone that has been in Social Media for some time now but has needed a fresh perspective on measuring all of the different "levels", this was the best article ive ever read on the matter.
Thank you!
Avinash- I think you've done a nice job at identifying the metrics for those properties that brands control. How does your approach toward conversation, amplification, and applause apply for those properties that a brand does not control but are just as relevant to the bottom line?
How do online communities play a role in the approach you describe above?
Page
Page: Owned property (on a company's own blog or Twitter channel or Google+ or self hosted online forum) metrics are clear. Use the four in the post.
For non-owned (others writing about the company on their own forums or websites or Facebook channel etc) I think things are a lot muddier. You (or I or P&G or Expedia) are not the center or attention or core purpose of those online forums or other people's blogs or entities. On those scenarios I think we can use "bleeding edge but as yet of uncertain quality" methods like measuring prevalent sentiment or (worse!) mentions or (if things really blow up) Conversion Rate and Amplification etc to keep an eye on things until they die down.
A possible example for "unowned properties" could be using something like Social Mention. Here's my report (keep an eye out on the stuff in the left nav): http://goo.gl/gO5oW or Omniture: http://goo.gl/JLwLy
The key difference is also the pro-activeness and re-activeness of actions we'll take on the data. That will strongly influence what we measure.
-Avinash.
Hi folks – in the process of impementing a similar dashboard have found 2 approaches to tracking Favorited Tweets.
1) Make sure you have set your Twitter notification settings to send you an e-mail when someone favorites one of your tweets. Filter and save those messages in a folder until you are ready to populate your scoring sheet.
2) Try this site favstar.fm it's designed to let you find the people that favorited and retweeted your posts but can also provide a way to find out which tweets were favorited after the fact.
Regarding Amplification on Twitter – I have a slightly extended approach.
If I share a post, and you retweet me then I consider that you are trying to influence your audience to read my post. When, your followers (i.e. my 2nd level of followers) retweet the same post, that is when I say "you" are an amplifier.
This is also important, when you see a pattern – few people will always retweet some posts. Does that mean, they are all important? It becomes important – when, the retweets indeed increase your reach. They can be just spammers, who have automated retweet based on name, or keyword. It starts making sense, when the same post gets retweeted by the 2nd, and 3rd level of followers.
So,
@akaushik: RT @amitapaul —> @akaushik is an Influencer
(I will be ecstatic, if I get retweeted by @akaushik, btw)
@2ndlevelfollower: RT @akaushik, @amitapaul –> Amplifier
It also suggests of two of your followers who RT your messages equally, one of them gets listened to more by his/her followers, while the other doesn't. For, example if I look at above, I will say ok, if @akaushik RTs my post, it increases the reach.
And, it is very easy to measure. I have built it in my SMM/Analytics product (ObjectiveMarketer), and is one of the very popular metrics used by my clients.
Amita: This might be a case of potato potata. : ) The spirit of the end goal we are trying to hit is the same, there is just a slight difference in perspective on the initial consideration.
For now, to take your example, I would consider it to be "first level amplifier" and "second level amplifier."
I've not thought about computing Influencer, because what does that really mean? And can quantitative data really measure something so qualitative? But I'm sure we will figure things out.
-Avinash.
I have been using qv source and qlikview to monitor and compare Facebook pages measuring likes posts and comments which then builds up sentiment for you and charts your competitors brands.
Same is done for twitter measuring average sentiment for individual tweets or #hashtags and identifying brand advocates and brand damages.
Very effective for tracking what people are saying about you or your competitors brands or service
Mostly people think that social media is about updating social profiles – such people have no target; hence, their business marketing using social media fails.
Another group of people do have social media marketing strategy but they don't know how to "use" it.
Your post clarifies both beautifully. It is essential to have both the purpose of intent and the knowledge to utilize the outcome of that intent.
Another great post Avinash!
And I completly agree – we tend to focus too much on up front numbers like # of Fans, Followers, etc., without digging deeper to see if we're a) actually engaging & connecting with our community; b) helping the business.
Thanks for another informative article!
Cheers!
Parin
I love the thinking in this, but I am inspired by a different post of yours. It suggested a measurement framekwork for web analytics but I think the structure lends itself to social media and fills in some of the gaps I see here.
You divided metrics into three buckets: acquisition, behaviour and conversion. What I see here is that you've identified metrics for behaviour and conversion, but you've ignored acquisition. I'm thinking that if acquisition relates to the total traffic to your website, and behaviour to the actions they take there, that in a social context, acquisition would be the total exposure to your content (number of fans/followers could be a useful proxy metric, though it should really be the true number of people who see your content (as measured by tools like tweetreach.com/).
Behaviour then covers interactions, clickthroughs and shares. Conversion is as you have outlined here: macro and micro conversions that are known to deliver business value.
What do you think?
Briana: I'm glad that you love the Acquisition, Behavior, Outcomes framework. It is central to ensuring that as Marketers and Analysts we are looking at what drives end to end success.
In this case you are absolutely right that my focus was profoundly on Behavior (what is happening in terms of us and our followers/likers/+1'ers) and Outcomes (on our social networks – increased reach et al – and our websites).
My lack of Acquisition focus was because I don't think, at least for now, that a quest to "acquire more likes / followers" is optimal. If you exhibit the optimal behavior / deliver optimal value in social channels then acquisition happens organically. I did not want to compute likes/followers/fans simply to not incentivise that behavior.
This might change over time. We'll see. :)
Thank you so much for bringing up this important question.
Avinash.
Avinash, hi,
I was wondering if these metrics wouldn't give a more accurate picture if calculated on the basis on number of users reached instead.
Indeed, if calculated based on number of posts, an increase in the number fans will automatically increase the number of likes, comments and shares.
But do you really create more conversations, applause and amplification ? On absolute numbers, yes. But to know if your fans are changing their behavior with your page, you'd better measure these metrics based on the number of users reached (could be tricky on twitter, agreed. Easier on Facebook)
Eager to know your opinion about this.
Steve
Steve: The goal for these metrics, per the post, is to not try and increase the number of fans or likes. We use both as diagnostic measures. It is of course entirely likely that using the metrics outlined in the post will increase fans/likes/etc.
Reach is a complex metric in Social Media. We don't know how many people see their newsfeed / Twitter stream / Google+ home view. We have no idea if they read anything we write.
We know we "reached" someone if they interacted with our social contribution in some way. So for now I prefer to keep the reach metric on the backburner.
Though please do note that metric #2, Amplification Rate, is structured to increase the reach of your social contributions by helping you understand what spreads.
-Avinash.
Thanks for your answer Avinash.
Understood, these metrics are activity-tracking metrics.
Facebook does provide the number of people who have seen your content in their news feed (or at least, the content has been shown on it).
My point was that by calculating based on a per user reached basis, we could compare the evolution of these 3 metrics without the bias created by the page's growth.
For example : My conversation rate was 2.7 comments per post in may. In June, it was 3.5.
It seems fans are increasing their engagement with my content, no ?
Not necessarily, if my fan page grew from 1000 fans to 10000 fans between May and June, and that my number of fans reached as changed from 500 to 10000, it may be the opposite. Users are engaging less with my content (on a per user reached base).
Wondering if you find any flaw in my thought…
Steve
Hi Avinash
Great post as ever. Only just come to it. Two perspectives I would add (which may already be in the comments – I haven't read them all I'm afraid!):
1. Different types of content will favour different metrics
I guess my point is that you understand what you are trying to achieve and craft your content strategy accordingly. Broadly speaking:
— News-y content is good for Amplification less good for Conversation or Applause
— Highly opinionated linkbait type content is good for Conversation less good for amplification or applause
— Tips/best practice type content is good for Applause and less good for amplification or conversation
So are you trying to get traffic? Links? Improve SEO? Value or volume? Most sites wants a combination. But it means your content strategy should probably consider all three metrics slightly differently.
2. *Who* is doing the conversing, applauding, amplifying?
Currently social media often relies on count values / popularity. Which isn't the same as credibility/authority/value. So in an ideal world all your metrics need to be viewed in the light of who is doing the conversing, amplifying, applauding. As a broad example if you look at the kinds of comments (conversing) you get these days on a Huffington Post, or a Mashable or Techcrunch even, then I'm afraid their lack of quality switches me off. So they score well on your metrics but, arguably, damage the brand. Whereas the comments on your site (or, dare I say it, Econsultancy's) may score less well for your metrics, I think they represent much better quality.
Ashley: It is always delightful to have your perspective, thanks so much for sharing it.
My post is very heavily influenced by what brands and people who are brands should consider an optimal approaching to doing the right things right.
You are absolutely right that linkbaiters and newsy content types might go for a slightly different emphasis for one metric over another. In the end though they have to figure out how to deliver on economic value, if they don't no amount of link baiting and amplification will save their business/site/brand.
Hence trying to solve for all four works optimally over the long run. In the techcrunch, huffpo cases (I'm with you on both) huge conversation, low economic value (hopefully!) and life will be fair. And Econsultancy will win big! :)
I'll keep the thought of credibility/authority/value in mind to see if I can evolve this framework over time.
-Avinash.
Great post Avinash! I just derive so much value from your approach. Thank you. A possible contender for measuring conversation across many different platforms is socialreport.com
It includes Twitter, Facebook, LinkedIn, Google Analytics, WordPress, and a whole slew of others. It gives overall measures for engagement, reach etc. but doesn't let you do much drilldown within the app, which is a pity. It does let you export data to CSV but that data doesn't have links or the text of individual content – so if you want to find out which tweet got RTed the most, you have to find the date in the CSV and then find it on twitter.com.
Apart from that (pretty significant) shortcoming, it's a good indication of overall health over several platforms. Good for time-challenged people who don't care so much about actionability of insights (e.g. someone who creates and responds in real time, intuitively, but needs some documentation to show a manager or client).
Great post Avinash, really helpful as we try and align digital metrics with organisational ones.
TrueSocialMetrics seems to be an amazing present, much better than a spreadsheet! Thanks
When calculating “applause rate” (likes/post), should we be counting every like, of every post?
For example, if you have ten posts and each of those posts gets liked once, then you would have a 100% applause rate, indicating that people like all of this content.
Now, say you have another ten posts, and nine of those don’t get liked, but one of them gets liked ten times. This also gives you a 100% applause rate, despite 90% of those items not really doing much for engagement; in this case would it be more accurate to count one of those ten posts, and say it had a 10% applause rate?
As I understand it, "applause rate" is the amount of likes (or favorites) compared to the total number of views of the post.
Sam: My summary definition of Applause Rate is:
# of “Positive Clicks” Per Social Contribution
The "positive clicks" will be Likes on Facebook, Favorites on Twitter, +1's on Google+ etc etc.
So we are trying to understand is how many positive clicks we can get on each social contribution (tweet, post, etc). You'll get an "average number of positive clicks per social contribution" which will give you essentially a benchmark you can use to judge which types of social contributions you should do more of (in your case the one with the 10 likes) and which types you should do less of (the ones with zero likes).
I worry less about the number, and more about how we'll use it. :)
I hope this helps.
-Avinash.
Conversation, amplification, applause, and how each social media channel contribute to the business bottom-line are insightful metrics. I learned about tools and ways to measure from this conversation.
I want to bring up softer side to this…
I have seen and read over and over again that end of the day quality of the content matters… Are you answering a question? Are you making prospect's/customer's life easier with the content? Is the content making a positive difference to the friends and followers? And it all makes sense to me.
My Q – please can you share good examples of content usage over social media channels for …
1. Manufacturing companies trying to promote their products?
2. Brand credibility initiatives – where objective is to build awareness?
Thanks,
-Deven
Deven: I'm glad you found the post to be of value.
You bring up a wonderful point about the quality of engagement. Amplification and Applause both try to get to the quality of engagement with the hypothesis that if you are not contributing something of value then neither metric will look great.
It would be wonderful to build on top of Conversation Rate and get a qualitative feel for the type of the conversation and not just the quantity of it. Something like what SocialMention does when it classifies things into positive, neutral and negative. For now that classification (in any tool) is simply not good enough. But it will get there.
To your questions….
#1. I would postulate that "promoting their products" can't be the goal of social media. That will most certainly not lead to a valuable social media presence. Flip to creating value for the world, promoting product is a by product that might happen if you make your social contacts life a little better.
#2. For all the flack it got I think BP did a great job with social media when the unfortunate gulf disaster struck. They used search and social to get their voice out there. On the social channels they tried to be as human as they could be (given legal and other constraints), and it as a good effort. Take a look at what small companies are doing, like Seventh Generation and Songza etc.
-Avinash.
Hi Avinash:
Thank you for your reply. I will look at BP, Seventh Generation and Songza.
Let me rephrase my Q # 1 – Can you share with me examples of products manufacturing companies that have done nice job in your opinion in using the social media channels – for answering questions, serving customers, explaining how to solve problems …etc?
-Deven
I meant paraphrase, and not rephrase… :)
Nice article Avinash!
I didn't knew about these things earlier and didn't have any knowledge how to measure the social media.
Thanks for the post.
Hi Avinash,
I know it's a while since you first published this, but in my earlier comments I was talking about how we used to measure activity, engagement, reach and network size. I've now developed a social media metrics dashboard free to download here: joakimnilsson.com/metrics-analytics/download-your-free-social-media-metrics-dashboard
It doesn't try to assign a monetary value but rather focus on producing "better" content.
Br, Joakim
Really like your concept of Conversation Rate.
I would add that an important piece of engaging in conversations are made up of a series of interactions over time. It can take a while for conversations to unfold. It can takes days and months, even years for conversations to develop.
Thanks for the post!
Wow – I have been learning from your blog over the past 4 months and feel like I have a solid analytic strategy primarily based on the knowledge you have shared through your blog on this topic.
You are moving this field forward in leaps and bounds.
So this True Social Metrics thing seems to be all you need for what Avinash Kaushik is talking about. Am I missing something?
Here is the link to the tool Mr. Kaushik provided: http://www.truesocialmetrics.com/
The Social Media world is still quite new to me and I realize that the more I learn, the more I have to learn. Having said that, I am embracing it all. In my current Social Media Metrics course at a University we were introduced to your measurement tool and it resonated with me.
Social Media is clearly not an exact science and keeps evolving, but it's clear that measurement tools need to evolve as well in order to make it a viable business proposition and in order to remain relevant. It's up to us to educate businesses, set expectations, and articulate / manipulate (in an honest and "reasonable" way) the tools that we have at our disposal to best meet the needs of the individual businesses. That is both the beauty and the challenge for social media.
I appreciate having a tool like this that can help me explain to businesses the value of social media. I haven't been quite certain up until now how to even remotely correlate numbers to actions. This is a good start. Thanks.
How can I measure economic value of a brand without e-commerce such as luxury brands Chanel or Louis Vuitton?
Jessica: First identify the macro and micro conversions.
Then identify the value of the easiest micro conversions. For Chanel this is the ecommerce outcomes, the value of the videos, offline referrals etc.
Now you have some basic numbers to go with.
Then identify the value of the harder micro conversions, and then you'll have the complete economic value.
Here's my blog post, with some examples like Victoria Secret, of how to compute economic value:
~ Excellent Analytics Tips #19: Identify Website Goal [Economic] Values
https://www.kaushik.net/avinash/web-analytics-tips-identify-website-goal-values/
-Avinash.
Fantastic.
Thanks a lot Avinash
At the end of the day, the economic value of a social media presence is as much about getting known and having a good solid reputation.
Interesting article.
Is there a way to compare brands with these metrics? I mean, if Brand A (with 20.000 FB fans) has a conversation rate of 2, and Brand B (with 1000 FB Fans) also has a conversation rate of 2 which brand is better? is there any kind of scale?
Thanks!
Alejandro: You can try to normalize the numbers by creating a percentage. "For xx,xxx fans what % engage with a post on average." That takes care of the size issue.
My only caveat is that I try to be careful with large sizes. So it is easy for me at 15k Likes to get 80% conversation rate but for a company with 15 million Likes they might only get 50% and that is ok.
If you want to play with a tool, checkout http://www.truesocialmetrics.com
Avinash.
Thanks! Rather than normalize, I'll just present the data within the "number of fans/followers" context, I think it'll be more easy for the clients to understand. Regarding your link, I appreciate that info, I've tested it and find it very useful.
That's a great comparison Avinash. Loved the article.
The nitty-gritty of the campaigns seems to indeed matter at the end !!
Hi Avinash,
I just tried out your metrics for Twitter and Facebook. I'm confused about what counts as a "comment" for Twitter. Is that a "mention?" Facebook is straightforward. Retweets and favorites are the most concrete to measure, but comments could be subjective. How should I proceed?
Thanks for your help!
-Kirk
Kirk: "Comments" on Twitter are simply replies to a particular Tweet.
Does that help?
If you want to play with a tool that measures my four metrics well, including Economic Value, please check out http://www.truesocialmetrics.com
-Avinash.
That does help. Thanks for your reply. I also found this helpful list of definitions over at Twitter FAQ: https://support.twitter.com/articles/119138-types-of-tweets-and-where-they-appear
Hi Avanish,
I appreciate there are a lot of comments before this that i have no read. I would however like to "applause" you, sorry couldn't resist what is already an over used pun.
Even though this blog post is two years old it is clear that the metrics you discuss are still highly relevant to understanding your business impact on social media and the amplification of these activities.
Terrific Post Avinash, but I would really appreciate if you could elaborate how you arrive at the figures for the Economic Value.
Thanks for this wonderful content.
Shiv: Here is a blog post that outlines how to go about computing economic value. Typically it comes from identifying key business goals and identify the value added to the business by each goal.
~ Excellent Analytics Tips #19: Identify Website Goal [Economic] Values
-Avinash.
Great post Avinash! Where can I learn how to set up the Google Analytics table measuring macro and micro conversions?
Thanks
– Chloe
Chloe: You can learn about how to do it in GA here:
Set up and edit Goals: https://support.google.com/analytics/answer/1032415?hl=en
Avinash.
Hey Avinash,
I was revisiting this post (always a go-to for when your social strategy needs to refocus) and I've got an update for #3. Twitter (finally) released their analytics last week, so now we've got a way to measure a tweet's amplification within the social service!
Great blog post!! All this fab info will help me out a lot, thanks for sharing!
Also, had a look at "True Social Metrics" – looks awesome and can't wait to try it out and let you know how it's worked for me.
Alex :)
Great measurement.
I am amazed how you have put the things in nice & precise way. And i definitely think it will help in re-designing the Marketing Strategy.
I am a marketing student & was advised by my professor to read this article .
Thanks
This post remains a great resource, thanks Avinash.
Some people might be interested to know that I’ve written some code in Google Apps Script (for use with a Google Docs Spreadsheet) to automate the process of fetching, calculating and displaying these metrics from Facebook, Twitter and Google Analytics.
It’s open sourced and available under the MIT License at: bitbucket.org/worldanimalprotectionuk/social-media-kpis
NB: As I’m primarily a marketer, not a programmer, I’m sure there’s plenty of scope for improvement to the code – hence the source being opened up to the community.
Hello Avinash Kaushik.
I created a spreadsheet with your stats for multiple social sites, across the board with grand totals. After reading your article, I noticed your invitation to submit our creations on combining everything.
Here it is:
Direct Download of File: http://www.socialmediavista.com/wp-content/uploads/2015/02/1nOriginal_Unlocked.xlsx (PLEASE download to your computer. This file is on my WordPress server)
Spreadsheet post on my blog: http://www.socialmediavista.com/social-media-metrics-spreadsheets/
(I made some of my own additions to the spreadsheet and asking for a cheap payment to buy them. To help my website/social media venture. Is that OK?)
Thank,
Glenn
Amazing site, an incredible study. I can stay in this article for hours
I feel key metric is engagement as a percentage of reach – I think where Facebook is concerned it is important to take into account your reach when measuring the 'success' of a post, as this gives you the true perceived value of a piece of content, regardless of how many people Facebook has allowed it to reach organically.
Hi,
I was reading this post and I have a doubt. How can I define the money value of a micro conversion? Should I just set a value?
I love your posts, thank your for the attention.
Marcus: Here's a blog post that will help you understand how to compute Economic Value (revenue and more) for your company's micro conversions:
~ Excellent Analytics Tips #19: Identify Website Goal [Economic] Values
All the best!
Avinash.
Nice post.
I just want to know if i create score card and i should measure my performance in below form – Very bad, bad, Avg, Good, Very good.
How should i categorized all 5 above points ?
Nitish: You could do that.
You could also simply benchmark your performance against your past performance or current performance of your competitors.
Or use True Social Metrics, where you can get industry standards.
Avinash.
Hey Avinash!
Well Really an informative article Man. I have a question: What is more effective Social Media Marketing Or Search Engine Marketing (SEO)?
As I read your posts and conversation I think you answered it well.
Thanks & Wish you A very Happy New Year.
Regards Leonard
Leonard: I'm afraid that is a little like asking if Oranges are better or Agave. But have their unique uses, it really depends on if you want to make orange juice or tequila. :)
Perhaps, it will be helpful for you to think about his using my See-Think-Do-Care Business Framework.
SEO is great for Think and Do. SMM is good at See.
If you company has a great Care content strategy, both can be pretty good at Care (though SMM a bit better than SEO).
Avinash.
Great Post! Very detailed and informative, thanks for writing such a good post.
Keep up the work!
Thank you for the wonderful post.
I am not sure if I missed this in the article/blog, but was just wondering how to evaluate your competitors effectively on Social Media. Which metric would you suggest is a good benchmark against your competitor.
Is it the amplification rate and audience growth or is there any other metric that we should watch out for.
Vaishnavi: I love Amplification Rate, because it is the hardest to earn – there is pain a person has to go through to amplify your content, and hence provides a great test if your content is good.
If you are looking or a source to benchmark, checkout our friends at https://www.truesocialmetrics.com/
Avinash.
To increase the strength of your content, you need to “amplify” it. What does that mean?
According to Avinash Kaushik, amplification is the rate at which your followers take your content and share it through their network.
Avinash,
Currently, I am writing a social media marketing plan. I also have to describe how the ROI can be calculated. Therefore I am very pleased with this method so I can apply it. I just do not really understand how you calculated the economic value. Or how do I have to get it.
It is also a company that is active on social media, only no interaction or engagement at all. Am I supposed to compare the results to previous performances, competitors or the standard? And why?
– Gien
Gien: You set up goals and goal values in Google Analytics, and True Social Metrics will pull that from there. As to how you identify what goals and and what values to assign to them, please see this blog post:
Excellent Analytics Tips #19: Identify Website Goal [Economic] Values
With regards to your second question, please see my reply to Karl as it has more details: How To Suck At Social Media: An Indispensable Guide For Businesses
Avinash.
Hi Avinash,
More and more people use their social media channels as a source of revenue by itself (e.g. http://www.harpersbazaar.com/fashion/trends/a10949/how-bloggers-make-money-on-instagram).
The first three KPI's in your framework work perfectly for measuring such kind of a business (what works well and don't for my audience).
However, the Economic Value seems more complicated here… You are not trying to send your followers towards an owned medium. Most of the times you're not even sending them anywhere. You just give exposure to a certain brand.
How would you suggest measuring that Economic Value in such business models?
– If you are referring to a third party website, you could agree with that party to use certain UTM tagging (or an affiliate link). That way, at least they can follow up what traffic and revenue you're driving with your posts. However, it doesn't give you any data to learn what types of content/posts/products work to create direct response reactions and which don't. Which could be a very valuable insight towards future advertising clients…
– When you can't use a link (which is the case on Instagram), you're completely in the dark.
– Another way to approach the Economic value could be to only consider your own income (not the ROI you're generating for advertisers) and calculate the net value of a new follower by dividing you're income by the total amount of followers. I suggest followers as advertising rates often depend on this (the old-school marketing way…).
Would love to hear your point of view on this Avinash!
Jente: My sincerest apologies for the delay in replying to your kind comment.
One simple approach you could take is to measure metrics like Brand Favourability Ratings, Likelihood to Recommend, or even the very obvious Brand Awareness and their ilk. These things as you know have value. You can use primary or secondary research methods to measure these metrics (if you do advertising on YouTube, ask your YouTube account manager how they can measure this for you).
If there are links involved, you can measure Assisted Conversions at least (ideally via item tagged links, if not via referring URLs where available).
Your last idea is also a good one. Takes a small amount of pain, but totally worth it.
Avinash.
PS: Here's a post on measure economic value: Identify Website Goal [Economic] Values
Wow – you blew my mind again.
I am already rewriting the social media material for my trainees and throw these KPIs in our meetings. This is awesome man!
Thanks!
It's interesting to see an article online for so long and to find that it is still relevant today and still has people commenting. Over 5 years of responses on this single page made for an interesting read. Great work Avinash.
Definitely!! I read it once a year, so I never forget where to focus!!
Avinash is my Gilgamesh!
Thanks for such a great post, it blew my mind.
I was not thinking about the points shared by you as well as by Avinash.
A sound conversation with a great knowledge. Thank you.
Dear Avinash,
You wrote above the following, a couple of years ago: "…..Oh, I totally forgot to say this…. the advice in this blog post is only for businesses and brands that participate in social media. Businesses as in Red Bull and T-Mobile and Johnson & Johnson. Brands (all of the aforementioned plus…) as in Mitch Joel and Stephen Colbert and Nancy Pelosi. If you don't fall into those two categories then this social media measurement framework might not apply to you…"
SO, Some post(s) you can recommend for us, marketers in small & medium enterprises selling B2B in niche markets ? I am per example marketing furniture and equipment for hospitals….
I thank you in advance…
Hi Avinash,
Really fantastic post, I have been using this framework in discussions about social performance for years. I really enjoy the True Social Metrics tool as a means for comparing a competitive set based on these metrics.
My management team has approached me asking if we can provide a formula for scoring social posts and as a monthly KPI that uses different weights based on the business value of likes, comments and shares.
My initial thought is that a calculation for this type of compound metric would not be ideal for getting the best insight per your article here: https://www.kaushik.net/avinash/insights-web-analytics-kpi-measurement-techniques/
Would you agree? Any other thoughts on this subject? Your input would be greatly appreciated!
Thank you,
Justin
Great insight to do better in the social networks.
Nice post, Avinash. Although the post has been there for quite sometime, most important it covered most important points here.
Social media is the direct line of communication for customers and their brand. Content is important piece if information to entertain the community, company news and updates, but without that support, marketing messages feel disingenuous.
"The biggest brands in the world embarrass themselves every day on social media"
I couldn't agree more, the market share for the biggest corporations are slowly dwindling because they fail to connect with their consumers on a more personal level.
YouTube, Facebook, Instagram is filling that void right now and they better smarten up. Personal brands are huge and making massive profits right now, thanks to mostly Youtube and Instagram. We are really crushing it for our clients with Instagram advertising right now.
Hi Avinash,
I am a big fan of your blog. I had read many of your great articles and learn great techniques and methods to do proper research.
This post was above and beyond in so many ways. The information is invaluable, and all case studies are very inspiring.
Thank you.
Hi Avinash
Probably you have got this q before but pls excuse me if I am asking again.
The q is that do you still think your measurement framework for FB is valid given the fact FB allows you to optimize for different purpose for e.g. reach & frequency, engagement, clicks to website etc
Faisal: Parts of the framework, especially those concerning Paid Ads, are still relevant.
For the rest, I've evolved my point of view based on how Facebook has changed its purpose and business strategy. See my latest detailed post on that topic, with a new recommendation for Facebook:
~ Stop All Social Media Activity (Organic) | Solve For A Profitable Reality
-Avinash.
This is Powerful Stuf! Timeless classic that will always be relevant in executing a robust digital marketing strategy.
Every SEO specialist and digital marketer can get so much out of this one post alone. I have printed copies of this hanging on my wall and come back to it from time to time!
Love the cute toddler to Adult ninja analogy as well!
Using data to make and facilitate good decision making that yields awesome results is the stuff of dreams for any business professional really -but more so for the digital marketers. This great educational piece provides the right context to do just that!
There really are too many choices out there where every social media is fighting for your dollar and many of them having minimal effect on the business at the end of the day. The "impressions" do not mean much if the conversion is uber low and the conversion value is in the single to low double digits.
Thanks again!
Usman
Great post Avinash, really helpful for a newbie as me.
I have shared your article in my Danish network.
Thanks for making things clear and thanks for sharing.
Which is the best amplification rate?
Gabriela: Since computation of Amplification rate is pretty from public data, my recommendation is to look at the accounts of your industry's leaders (who do social well) and use their number as a benchmark.
Obviously, you can also index against your past performance and create a goal for yourself.
Avinash.
Despite being crucial to success, many brands still fail to leverage social media analysis by using only native tools and analyzing each channel individually, A.K.A, the bare minimum.
This is a huge loss of key data points and other social media insights that brands could discover with the help of analytics.
Hello Avinash, Thank you for your descriptive guide. I have come across your blog and realized that i need to measure social media success as well and how much it is important.