Best Web Metrics / KPIs for a Small, Medium or Large Sized Business

Sunshine We have access to more data than God wants anyone to have. Thus it is not surprising that we feel overwhelmed, and rather than being data driven we just get paralyzed. Life does not have to be that scary. In fact a data driven life is sexiest digital life you can imagine.

In this blog post we are going to bring the sexyback. I am going to attempt to significantly simply your life by recommending the critical few metrics you should use to analyze performance of your digital marketing campaigns and website. You'll be able to quickly go from "omg what can I do!" to "omg what am I going to do with all the money and fame I'm earning!"

The approach I'm going to use is to 1. Use my Acquisition, Behavior and Outcomes framework to ensure an end-to-end view of important activity and 2. Recommend metrics / KPIs you can use based on the size of your company.

Each recommendation comes with hints on what analysis to perform once you have the data, and what changes you could make to your campaigns, content and overall digital strategy. [A summary in pictorial format is at the end of this post.]

Excited? Let's do this!

Best Metrics / KPIs for Small Business Websites

Small business websites are a very fragile ecosystem. People working hard to do the best they can on the smallest possible budgets. But not to worry. They have to start with just four simple metrics to start rocking!


Clicks? Visits? Backlinks? Impressions? No. We have something magnificent.

Cost Per Acquisition.

Obsess about this metric. You have very little money. You need to know, obsessively, what you get for it. This metric delivers that insight. Oh, and everything has a CPA (not just your paid search or display/banner ads). If you are doing SEO then you are likely paying for someone. That's the cost.

cost per acquisition

Kill things that don't have an optimum CPA. Invest more in ones that do. Simple enough, right?

Tip: Remember this is just cost, not profit. Remember to take the next step to profit, here's how to do it in Analytics: Excellent Analytics Tip #27: Chase Smart Calculated Metrics!

Where is it? Most likely in Excel. For Search it is in your Google Analytics or Omniture Site Catalyst reports. But for most other programs (Affiliate, Email, Social, Display) your Cost is likely sitting outside your web analytics tool. So extract the # of conversions, import into Excel, add a column for Cost, do the math, sing or weep (based on what the data says!:)).

If you are paying someone to do web analytics and this metric is not on top of the dashboard they've created for you, it might be time to say sayonara to them.


Page Views? Time on Site? No. You can do so much better!

Bounce Rate.

I continue to be a believer in trying to prompt love at first sight. Okay, okay, I'll settle for delivering relevance. :) Bounce Rate helps you identify campaigns where you might be targeting wrong people (who then come to your site and leave right away) or sending relevant traffic to irrelevant (and often flash-filled hideous) landing pages.

Bounce rate helps you find campaigns and landing pages that need to be killed / improved. Everyday.

Where is it? Standard metric in every web analytics tool worth anything. Look at your All Traffic Sources report and your Landing Pages report.

Checkout Abandonment Rate.

I find the fastest way to make money is to take it from the people who have already decided to give it to you. Obsess about checkout abandonment rate (the percentage of people who click Start Checkout to those who complete that process).

paditrack funnel setup

Focus on checkout steps with the highest abandonment. Tweak like crazy. A/B & Multivariate tests are a good option. But you are a small business… so just take away as many fields as you can, play with where to show shipping cost (I vote for way up front), reduce the number of checkout steps if you can, ask for account creation at the end of the process rather than at the start. Try, test, measure, be rich.

Where is it? In Excel. Or if you use Google Analytics: In Paditrack for free. (Google Analytics' native funnels are pretty sub optimal, ignore that entire feature.) For other tools: In KissMetrics. Create a funnel just for the checkout process (from clicking Start Checkout to Thanks for your Order) and both these tools will give you the metric automatically. They also allow you to segment the data! Make love to it.

[Bonus: What is abandonment rate?]


My favorite Economic Value? No. As a small business I recommend…

Macro Conversion Rate.

You are a small business. Obsess about conversion rates, and everything connected to improving them. What products are people buying? Every single day (okay week) look at the All Traffic Sources report and seek out the Conversion Rate metric. Ruthlessly punish sources that are not working well and reward the pretty babies. Be they Earned, Owned and Paid media – oh and have a marketing strategy that has each of those elements or as a small business owner you are not going to win a lot.

macro ecommerce conversion rate

I love creating an advanced segment with just the people who buy twice the average order size. I call them the Whales. Look at sources, locations, product bundles purchased, keywords and campaigns and all that to learn where/how you can find more Whales.

Where is it? Standard metric in all analytics tools. Remember to look at both the rate and the raw number of conversions for context. People make silly decisions when they don't do that.

That's it!

You are a small sized business and these four simple key performance indicators will literally rock your world as soon as you start measuring them. Cost Per Acquisition. Bounce Rate. Checkout Abandonment Rate. Macro Conversion Rate. Don't look at any other metric until you feel you've mastered them.

Tip: If you've hired the right analytics talent/consultant to help you, they'll be measuring these fabulous four.

Best Metrics / KPIs for Medium Sized Business Websites

What if you are a medium sized business? What key performance indicators are optimal for you?

First, you are going to measure the KPIs mentioned above. But because you are running a bigger and more complex business you'll also measure…



+ Click-through Rate

While CPA is a macro metric about your campaigns' bottom-line performance, Click-thru Rate (CTR) is a deeper dive into analyzing the creativity and relevance of your affiliate deals / search listing / blinky banner ads.

In the context of Search (Paid or Organic), the text in your ads, the number at which your listing is ranked, the match between the user query and your ad's intent all help you receive a higher CTR. And if someone comes to your site (and does not bounce!) then you get an opportunity to convince them of your product or service's glory.

click-through-rate custom report

Small tweaks to the subject line of your email campaigns can have dramatic improvement in CTR. Recency and Frequency capping of your display remarketing campaigns can have a huge impact. Changing demographic targeting options in your Facebook ads can work wonders. Etc., etc., etc.

Put another way… CTR helps you understand if you showed up at the right place for your first date. Are you dressed okay. And if you are smiling the right smile. Helpful to know, right?

Where is it?

Everywhere. Start at a campaign level. Drill down to individual creatives. Kill badness. Promote goodness. Rinse. Repeat.


Bounce Rate

Checkout Abandonment Rate

+ Page Depth

A very tiny percentage of visitors to your site will see more than a couple pages. That's the internet for you. As you improve the user experience, information architecture and relevancy of content on your site, it is important to keep an eye not on the rather useless metric of Average Page Views per Visit or Average Time on Site but rather on the distribution of page depth. Here's how that picture might look like (from a post I wrote in July 2006!)…

page depth analysis

From the deep detail reported by your web analytics tool you can choose to aggregate into buckets you most care about (like mine above). Categorizing the visits into Abandoners, Flirts, Browsers, One-off-Wonders, Loyalists will dramatically change your view of content consumption. Over time, as you move to deeper consumption, you'll see direct business rewards.

The above image emphasizes a sale/conversion at the end, but even if you are a content-only website improving Page depth helps you because more pages equal (at the very minimum) more ad impressions!

Where is it? The final table will be in Excel. If you use Google Analytics the data you need is here: Audience > Behavior > Engagement > Page Depth tab. If you use WebTrends, Yahoo! Analytics, Coremetrics please click around to find the data. They all have it.

+ Loyalty (Count of Visits)

If Page Depth helps you optimize for a single session experience, Loyalty helps you optimize pan session behavior. Put another way… how good are you at getting the same person to visit your website multiple times? For ecommerce or non-ecommerce websites, loyalty can mean the difference between life of survival and raking in profits like crazy.

First set a goal for the % of site Visits you would like for people who've visited more than x times. [Set a goal for x too. :)] For ecommerce websites use your Days to Conversion report (more on this metric below) to set your goal. For content sites perhaps mirror your content update schedule. If you are the New York Times and you update the website 24 times a day then should the average person be visiting the site at least 90 times per month?

Your BFF, as always, is analysis and not just reporting the metric. Create this simple segment in five seconds…

segmenting by visitor loyalty

Apply to your keywords and campaigns and referring sources reports and identify which sources drive loyal traffic. Apply it to your content reports and figure out which content drives Loyalty (Sports? Op Ed? International? Cat Stories?).

Where is it? In every web analytics tool on the planet. If you use Google Analytics the data you need is here: Audience > Behavior > Frequency & Recency.


Macro Conversion Rate.

+ Micro Conversion Rate

Pick your favorite benchmark and you'll notice that less than 2% of visitors convert. Focusing on just the Macro Conversion Rate means you don't care if you received any business value from the 98% that did not convert. I refuse to accept that uber-lameness.

Identify your Micro Conversions (/Goals) and obsess about the long and short term business value they deliver. You'll quickly realize the Economic Value they create for you is often far greater than the Revenue your Macro Conversion reports! And optimizing for that will ensure you win HUGE.

micro conversion rates

Where is it? In Google Analytics it is here: Conversions > Goals. Even if you are a content site the data is there. Details in the Goal URLs report. Setting up goals takes two minutes, setting goal values might take you a week (see measurement strategies here). If you use other tools, please check with your vendor.

+ Per Visit Goal Value

This Key Performance Indicator 1. helps you move beyond the obsession of focusing on the 2% (because it forces you to focus on Every Visit!) and 2. encourages you to create a business that uses the web to deliver multiple outcomes to your visitors.

per visit goal value

Every visitor will not convert, but every visitor will, hopefully, deliver some Economic Value. Looking at this metric helps you identify Goals that contribute higher value, and and understanding of simple things like where you should focus on. If Twitter delivers 87 cents of Per Visit Goal Value and Google delivers 97 cents then perhaps I want to keep focusing on my SEO strategies rather than following the advice of the Social Media Guru who's just informed me Search is dead.

Where is it? In pretty much every single report in every single web analytics tool. Click on the Goals tab.

That's it!

For a medium sized business we ended up with nine metrics. Seems about right if you are making more than five million dollars of economic value. They key difference from websites that are in the small business category is that we are going to shoot for multiple conversions, deeper site engagement and better analysis of acquisition efficiency.

Time now to deal with the big boys and girls… large websites!

Best Metrics / KPIs for Large Sized Business Websites



Click-through Rate

+ % New Visits

My choice of this metric perhaps betrays my refusal to rest on my laurels. There are clearly a finite number of people in the world relevant for any business. But staying hungry and staying foolish is a popular mantra for me. I use this metric to constantly calibrate my acquisition strategy to understand which inbound marketing efforts are bringing new "impression virgins" to the business.

If you look at your Earned, Owned and Paid media then this metric is especially important for your Paid media efforts. Except for your re-targeting / behavior targeting campaigns, you want your paid search, display, affiliate, and social efforts to bring new visitors to your franchise.

Where is it? It's like air, everywhere! Don't forget to segment for optimal analysis.


Bounce Rate

Checkout Abandonment Rate

Page Depth

Loyalty (Count of Visits)

+ Events / Visit

Every awesome large website delivers complex experiences (videos, demos, dynamic slideshows, configurators + + +) via sophisticated technologies (Flash, AJAX, Gadgets + + +). Almost all of the time we leave measuring their effectiveness on faith (or the HiPPO). I love event tracking because it helps us measure these often astonishingly, expensive initiatives.

events per visit metric

Of 110,842 visits to the site, 9,054 interacted with your delightful experiences and each of those visits had 2.24 Events per Visit. Is that good? Bad? Could be better? Are these 2.24 interactions delivering higher economic value to your business?

In the above case the answer was a big NO. In your your case you'll decide based on your strategy and goals. At the end of the analysis you'll make significantly smarter decisions about your content (especially because the Analysis Ninja that you are, you'll triangulate performance of this metric with first, Page Depth and, second, Loyalty).

Where is it? Most web analytics tools do some type of event tracking. Please check with your vendor (it might not be called event tracking in their lingo, just describe my first paragraph above). In Google Analytics the data is here: Content > Events.


Macro Conversion Rate.

Micro Conversion Rate

Per Visit Goal Value

+ Days to Conversion [or Time Lag for Content sites]

Another pan session metric I adore.

Life, no matter how hot you are, is not a series of one night stands. Yet because of how they analyze the data most companies end up optimizing their web marketing campaigns for one night stands. Come here and convert NOW! If yes: Oh, I love you. If no: Kill the campaign!

That approach is not just short-sighted; it is an insult to your visitors. Convert them at a pace they are most comfortable with. This metric helps you understand how quickly or slowly your visitors convert. You can, at the very minimum, change your campaign messaging and come hither calls to action and adjust your landing pages. If the Days to Conversion are much longer, then create a robust (slow dance) micro conversion strategy.

days to conversion time lag

If you have a non-ecommerce website then there is something delightful for you in the Google Analytics Multi-Channel Funnel reports. Checkout the Time Lag report . It is showing you exactly the same data as the Days to Transaction for Ecommerce sites. The metric you see immediately above is called Conversions. It is essentially your Goals (/micro conversions).

Optimize your "hello, nice to meet you, what would you like, here is what I have to offer, why don't you check with your spouse, come back and check it out again, multiple times, I'm still here, you ready to convert / deliver economic value, here's how… " process.

Where is it? Days to Conversion is in the Ecommerce section of your web analytics reports. It is a standard report. (Don't forget to segment your sources. Deep insights await.) Time Lag may or may not be a standard report in your tool. Please check with your vendor. In Google Analytics it is a standard report here: Conversions > Multi-Channel Funnels > Time Lag.

+ % Assisted Conversions

This is the newest metric I've made standard for all my clients / partners / BFFs. And it is a sweetie.

Assisted Conversions builds on the above mental model. It takes a while for a majority of your visitors to convert (macro and micro conversions), so why does almost all of web analytics focus on single channel analysis and optimizing that single channel in a silo? Just because the Affiliate click was the last one before conversion should it be optimized for that conversion? Especially if the Visitor originally came via Facebook (or Google or whatever)?

How many of your conversions had more than one ad / media / marketing touch prior to converting? Really smart Analysts at really successful companies understand that…

assist interaction analysis

…and then use that data to optimize the portfolio of channels rather than individual channels for the company.

Even if you don't do portfolio optimization (and desperately hope you do) you can easily see how the above data will cause you to execute a different marketing optimization and expectation strategy for Email (1.18 Assist / Last Interaction rate) vs. Organic Search (0.61).

I am being modest when I say that this metric and subsequent analysis will have a fantastic impact on your company.

Where is it? % Assist Conversions may or may not be in your web analytics tool. Please check with your vendor. In Google Analytics you'll find it here: Conversions > Multi-Channel Funnels > Assisted Conversions.

And we are done!

For large businesses we've identified 13 key metrics that would give a robust end-to-end view of business performance. The key difference vs. medium sized businesses is that we are really, really, really focused on pan-session (multiple visits) behavior. Put another way, we really care about people here and not just a single visit.

Here is a summary of the metrics I am recommending in this post…

best metrics small medium large business

I hope the picture above will quickly help diagnose where current gaps in your measurement strategy might be.

Additionally if you are a small business you'll know what else to measure when you start to become medium sized and if/when you cross that threshold you'll know the metrics that come with your large business status. :)

You'll notice that I'm not focusing on KPIs like AdSense Ads CTR or Page Load Time or Actions per Social Visit or Search Exits (I love this metric!) or Content Distribution vs. Content Consumption Rate or Conversation Rate (in case of a content site) etc. That's simply because these KPIs tend to be unique to the type of business you are running. My strategy above was to focus on just the KPIs that would be applicable across all types of businesses.

That brings me to a very important point.

While it is my hope that you'll find my recommendations above relevant and yummy… the most optimal way to identify that best key performance indicators for your company will come using the process and structure outlined in the Digital Marketing & Measurement Model.

I'll end with the thought I started this post with… we have more data than God wants anyone to have. But web analytics does not have to be scary or impenetrable. Use the roadmap above, focus on all three elements (acquisition, behavior, outcomes) and I promise you'll soon be on your way to being as happy as God wants everyone to be.

I wish you all the best!

Okay as always it's your turn now.

Does your business use the above recommended metrics / key performance indicators? Do you have an absolute favorite metric that's not mentioned above? Which metric above do you find most useful? Which one most useless? What is your strategy for identifying the most relevant metrics?

Please share your suggestions, critique, and helpful best practices via comments.

Thank you.

Couple other posts on metrics / KPIs you might find interesting:


  1. 1
    Web Analytics Europa says

    Very good and logic overview, covering every size of a company and especially every budget situation and human resources within those sizes.

    Remember: The smaller the company the less time they can spend on special disciplines (very general formulation).

  2. 2

    Great article – Thanks

    Splitting down to small / medium / large is especially helpful as it's tough for small firms to know what to focus. It takes so long to figure out which metrics are most useful there is little time to actually do something about it.

  3. 3

    Avinash – I love the idea of starting with CPA with each business, as opportunity cost is often something people ignore. People assume because there's a positive ROI, it's a good thing. I suppose technically there's a fair amount of truth to that, but we have finite budgets and time an a world where our campaigns could head in nearly infinite directions.

    Also, the idea of caring about more metrics as your organization grows is a great way to further capitalize on limited resources and time by focusing on the big stuff first. Great post!

  4. 4

    Great piece Avinash. Something in here for everyone.

    One side note on Cart Abandonment Rate. I’d agree its important to focus on it, and all the advice that you give about shortening, simplifying, removing obstacles is valid. But it’s important to note that there is a natural cart abandon rate which every business has, and no amount of optimization will ‘fix’ this.

    That’s because it is unfixable.

    The top two reasons why consumers abandon carts are Price and Timing (i.e. not yet ready to buy). Both these are hard to optimize on site. Our research shows that abandonment, rather than being a bad thing, is in fact a micro-conversion on the road to an eventual purchase. If a visitor abandons, your chances of getting a sale have just trebled compared with a new visitor.

    Conventional wisdom suggests that the majority of cart abandoners are time wasters, and this is not true: Our research shows that 75% of cart abandoners will return to the site again within 28 days, either to purchase or abandon again. If they abandon more than once (what we term a ‘Serial Abandoner’) then your probability of getting a sale goes up again, this time by an additional 266%.

    Our research looked at the online behaviour of more than 600,000 people to understand why some visitors buy, but the vast majority do not.

    The inevitable conclusion, which is very parallel to the overall thrust of your blog, is that the majority of customers require multiple visits before making a purchase. The more visits they make, and the deeper they go into the site, the stronger their intent. So abandonment, rather than being a bad thing, is a measure of intent.

    • 5

      Charles: I concur with you on the value of, and challenges with behavior of, the carting process. For that reason I'd picked Checkout Abandonment. Once people start the checkout process we should get 100% to convert because there is a much deeper intent from the visitor there. Of course we will not get 100% but the average abandonment seems to run 70% in the checkout process. That is a shame.

      For the cart itself… it is sad that most website owners have still not figured out how to provide shipping information, clear pricing, in-stock availability etc in a clean and clear way. Why in the name of Zeus should we have to add to cart to get that info? Hopefully this will change.

      Thank you so much for sharing your thoughts!


  5. 6

    One tip: if you can not convince your bosses about importance of web data measurement (or you have partial success with it), try send them that sexy real-time dashboard from GA (for example: our expensive banner campaign ended before two weeks and you say to me that there are just 3 people atm in our freshly (expensively) redesigned website even if media agency said to us that there will be lot of delayed conversions due to kind of our product, so where they are?) , so as eye-opener to your bosses i vote for real-time :)

    Ofcourse as analyst ninja, you always know, that real time doesnt have big value for bottom line!

    • 7

      Eelke: Good strategy!

      I'm not a big fan of real time data, if we don't have the capacity to take real time action. But your recommendation is wonderful. Let's entice them with "pretty stuff" and get them hooked on the "powerful stuff." :)


      • 8

        I like this:

        I'm not a big fan of real time data, if we don't have the capacity to take real time action

        This is quite some quotable.

        I was impressed by the new GA feature for showing real-time data, but the most efficient way to analyze traffic is still the "dead listing" approach of the data. It's still a really nice and catchy toy though :D

  6. 9

    Such a useful post!!! (This post could almost be as popular as the "Custom GA Segments and Reporting" Post!!!)

    I digress.

    Adding to the + Loyalty KPI – the great thing about segmenting by keyword, marketing campaigns, and referring sources is that over time, you'll be able to develop proverbial leavers and pulleys that will tell you where to allocate customer acquisition resources more efficiently.

    This also segues into understanding the life time value of a customer as a whole when you start to benchmark your customers to whatever time period your cookie is set at – which is SUPER VALUABLE to analysis ninjas.

  7. 10

    Yes not everyone converts on the first visit and having the option to track a long run conversion is great.

    The data has to make sense and be of use to the end viewer. The more you break it up to a point that it is understandable at a glance the better it is to show end viewers. This is a great way to be able to make a point.

    The trick of course is to be able to simplify it without losing the impact of what you are trying to say.

  8. 11

    You said, "I continue to be a believer in trying to prompt love at first sight." But shouldn't it be love at first "site"? (Ba dum bump goes the drummer.)

    As usual, great stuff. Two thoughts about bounce rate that I'd like to share.

    1) Since I'm finally becoming more comfortable with Event Tracking I'd like to recommend that anyone who is going to obsess about bounce rate make sure they use Event Tracking (when appropriate) to help increase the accuracy of bounce rate counts.

    2) Moi? I'd be certain to also keep an eye on return visits. While it doesn't always apply to my personal surfing habits, quite often I'll visit a page and either leave it open in my browser (for days), bookmark it, or just go back to it from my browser's history. Perhaps I'm atypical?

    My point is (and sticking with your sex-appeal theme), not every visitor is looking for wham, bam, thank you mamma. I believe it's possible to becomes overly focused on quickie / one night stands (so to speak) such that you end up not appealing to the LTR type customers that is better for business.

    I appreciate that your recommendations always have caveats. I just wanted to add my 2 sexy cents.

    And fwiw, I think CTR should have been listed with small biz. There's no reason a small biz can't for example, do A/B testing on the subject line of an emailing and then roll out the one that does best a day or so later. If the difference is even slightly significant, the extra step would be well worth it. Let's also not discount how repeated "bad" subject lines is going to impact a receiver's willingness to even look at your subject line, let alone open it, in the future.

    Again, I realize you have to draw lines in the sand somewhere, but the marketer in me cringes when I get so many emails from the same source with awful email subject lines. I might not unsubscribe but you can be sure I'm turned off.

    • 12

      Mark: I wanted to keep recommendations for small businesses to four, or less. CTR, IMHO, would have less insights than CPA. And latter would force them into all kinds of analysis were CTR will pay a key role.

      And you are so right that many visitors don't want to buy. That makes a focus on multiple outcomes / micro conversions so incredibly valuable.

      Thank you for adding your comment, and adding the two caveats.


  9. 13

    As a reasonably experienced small to medium business owner who has made the leap to more of an online business owner I find this analysis very helpful.

    I'm used to analysing footfall in a more traditional way and making sense of all the online data available has been a little overwhelming.

    I appreciate the post and also the comments others have added, particularly Charles Nicholls and his comments on Cart Abandonment.

  10. 14
    Ander Jáuregui says

    Great post!

    Very useful!

    And if we make a daily, hourly conversion rate analysis, we can make a better campaign tuning to improve the cost per acquisition, among other things…

  11. 15
    Ozberk Olcer says

    Although CPA is a beautiful metric but it can be misleading as well. It is always a good start for small businesses to set some CPA goals for the purchases based on their margin, but what matters is what will that customer do after the initial purchase. Life Time Value is much more important.

    Same applies with memberships and leads as well. A marketing campaign can generate leads per 1$ spent but we cannot be sure if it is as effective as another campaign which has a CPA value of 10$ unless we measure the life time value.

    I see clients obsessed with psychological CPA values, setting lower and lower goals at the cost of quality leads or customers. To summarize without business insights and deep dive analysis CPA metric does not work well.

    • 16

      Ozberk: You are right that Customer Lifetime Value is a wonderful metric. Here's my blog post on it: Excellent Analytics Tip #17: Calculate Customer Lifetime Value

      The challenge is that it is incredibly difficult to measure even for the largest companies with the most sophisticated Analysts. You can only imagine the outsized hurdle a small business faces in trying to compute LTV. The most common outcome is that they do nothing. They even ignore simple things like Visitors or Conversion etc.

      By deliberately ignoring LTV my hope was to present a set of metrics that are simple to get to, yet metrics that provide complex enough insights to get the data driven ball rolling.

      Does that choice make sense?

      Thank you,


  12. 17

    Dear Avinash, thank you very much for this post. It's a delight to read your posts, especially as they are a fresh reminder of things I 've read in your Web Analytics 2.0 book. (yesterday I was reading how your "boss" – your lovely wife – needed outcomes to give you the go ahead

    I am an eBusiness consultant in Greece, where eBusiness is starting to boom (eCommerce at 2% of total retail with about $2B B2C on-line sales for 2011. Your posts and book will help all of us 'optimize' our efforts.

    I am sharing now your post and I hope your wife won't insist on "… quitting on of your 5 jobs". because you spend too much time helping us all. Thanks again for this post!

  13. 18

    Avinash – posting something here scares me. I feel like I am about to ask Einstein "that relativity thing is very interesting, but I am more interested in…"

    So I looked at your "large business" receommendations, and I feel that the offered metrics are very helpful if I were selling something directly or indirectly to consumers. Sales leads, sales conversions, etc. So here is my "but how about":

    – what if all I am "selling" is image attributes and brand KPI's?
    – what if I care most about what people will "think/feel" after being exposed, and less about "do"?
    – what if I wanted, in an ideal world, to link my brand attributes generated by (amongst others) brand exposure to sales and volume?

    Is the "behaviour" column the only thing I can look at? Be interested in your thoughts, which we can take off-line/real-world if you prefer.

    • 19

      Maarten: I've sprinkled metrics throughout the post that will work if you don't have an ecommerce site and only provide content of some sort. But to answer your questions more specifically…

      – what if all I am "selling" is image attributes and brand KPI's?

      If you are selling brand kpis then you would use those. Use the Digital Marketing and Measurement Model and that will help you focus and rock. Here's my perspective on brand metrics: Brand Measurement: Analytics & Metrics for Branding Campaigns

      – what if I care most about what people will "think/feel" after being exposed, and less about "do"?

      Checkout Outcome #4 in the above blog post. Likelihood to Recommend might be a great proxy for "did we manage to change your perception of our brand through our digital experience enough for you to recommend us to others." And Likelihood to take Offline Action might be a great proxy for "are you going to take some action offline as a result of your magnificent experience online."

      There has to be a bottom-line. In the case of a beer company might be: "Will you love Becks more than you did before?", "Are you going to be a brand ambassador for us?" or "Will you buy more beer because we made you this cool iPhone app / Facebook page?" If you have no bottom-line then of course you have no measurement. :)

      – what if I wanted, in an ideal world, to link my brand attributes generated by (amongst others) brand exposure to sales and volume?

      This is much harder. But in my efforts with large US & EU brands controlled experiments have helped identify value of brand marketing campaigns and their tie to sales/volume. Recently for a phone company there was a eighteen dollar lift in offline instore sales for every one dollar of online brand advertising. Not bad, right?

      Controlled experiments are complex but here my post outlining details of one such experiment: Measuring Incrementality: Controlled Experiments to the Rescue!

      Good luck!


  14. 20
    P.J. Labanowski says

    Your timing is impeccable!

  15. 21

    I'd like to hear more about "portfolio optimization" and a discussion on the relative value of "assists".

    Perhaps that can be a topic of a future post. :)

  16. 24
    Dave Rekuc says

    One thing I may add to obsessing over CPA is to understand your customer's lifetime value as well. In the same sense that we have to look beyond the 1 night stand in sessions, we need to do the same in orders to really understand what drives bottom line in our businesses. For larger businesses, we should be looking at this on a more segmented level to see lifetime value vary from campaign to campaign, in a smaller business it is perhaps just 1 value. The method that we collect this kind of data from small to large businesses may vary, but its value remains undeniable for any business size. I know this has been covered thoroughly in other posts, but its just a reminder I like to give when CPA comes up :)

    Regarding the assisted conversions portion, how do you determine how much value to attribute to an assisted conversion? Can this only be discovered by testing for incremental lift? My biggest obstacle in multi-touch attribution is defining this value, especially when complicated with more and more expensive touch points.

    Thank you for the post, Avinash. Thought provoking as always.

  17. 25
    Lindsay - 3seven9 says

    Thanks for your post Avinash! Some really good ideas, particularly for small businesses who may struggle more.

    I also like that you've encouraged companies to tweak and test when possible or if a strategy isn't working – I think many businesses have the tools to to measure but are unsure of how to make their results meaningful. We discuss how companies can use these techniques to their advantage

  18. 26

    Hi Avinash, great post as always!

    We have a large+ website but different segments, that is low-end, mid-end and high end.
    While low-enders have an AOV of $160, mid-enders of $760 and high-enders on $1450 (we have 10k orders as well). The average AOV is $420 but its a fake measure. :(

    How would you go about metricizing the reports to be able to measure success of each of the segments?



    • 27

      Joel: It is optimal to separate Metrics from Segments. For your large+ business the metrics that are perhaps a good starting point are those that are outlined in this post. You can add more complex or relevant metrics to the list if you would like to.

      What you are trying to do is then segment the data. In your case by average order value. That of course is a great strategy. If you use Google Analytics just create a segment for each bucket you are interested in. For example you can create a segment: Include Revenue Greater than 0 AND Include Revenue Less than 500. That will give you your low end. Repeat two more times to create your other two segments.

      Now apply those segments to your All Traffic Source report and your Content report and your Events reports etc to try and understand each batch of traffic better.

      The above segmentation is for GA, but you should be able to do this in any tool you are using.


      • 28


        That is helpful nut only at the outcome level. How about acqisition/behaviour?
        It is time consuming to create segments for product based, brand based or even keyword based searches/channels. Nevertheless, i'm unsure this will serve the purpose.

        I really appriciate your quick response, i feel we are closer than further…


  19. 29


    thank you for the few posts I have just read on your blog, not everything is clear to me. I am searching for an answer to a problem that might be an easy one for you. I need some advice.

    I analyze a big website that has a few categories ( those are also in Nav menu). There are many pages that belong to those categories. How to structure Google Analytics to treat those category pages as home pages, so I can have the same depth of information as I would have by analyzing the home page of this website? I want to know how much time the visitors spend on the whole section (the pages that belong to a category), the bounce rate of the whole section ( and not just a single page), to see the most visited pages in it, etc.

    Would you have some advice on this topic?

    many thanks

    • 30

      Martyna: There are many ways that you can do this with Google Analytics (or other web analytics tools). Depending on what we want to do you could use simple segmentation strategies (if your site url structure is cleanly laid out) or you could use event tracking or custom variables (for the latter if you set scope at a Visitor level you could even look at multiple visit behavior by the same person).

      In order to do the optimal implementation you'll likely hire a GACP (Google Analytics Certified Professional) who'll spend some time understanding your requirements and then make a recommendation. You'll find a list here:


      • 31


        Thank you for your answer. the solution is not clear, I will follow the terms mentioned by you. With the site, their url structure is not obvious, and the urls messy. Yet it is big firm that has specific goals and clean path to go.

        I looked on the list of GACPs, I will keep it in my mind.

        All the best


  20. 32
    Sergio Felix says

    Oh my God, this turned my head over into a headache haha must be because I'm not used to THIS level of analysis yet.

    The funny thing out of this, is that I was reading a blog just yesterday (One Spoon At a Time) on a guy that made around $97k on 2011 with a niche website and he wasn't tracking ANYTHING.

    I believe this is due because normally people who teach tracking, monitoring and analysis make it way too over complicated.

    Will definitely need to dig into this matter but from the beginning, this is a bit too advanced for me.


  21. 33

    Hi Avinash, great post as always!

    Could you please elaborate more in details about KPIs of AdSense Ads CTR please? Maybe on the future post

  22. 34
    Colin Clear says


    Thanks for a great post. I got your book a number of years ago and found it a great read. My main problem is implementing all the good suggestions you offer.

    You mentioned Paditrack in your post. I have just signed up for it and was wondering if you had any advice about setting up a funnel for the checkout process.

    Is there anything in particular you would recommend.

    As an aside I think the post is a must read for anyone interested in running a successful e-commerce business.


    • 35

      Colin: You should open PadiTrack's configurator (the screen that shows up after you log in) in one tab and open your website in another tab. In the tab with your website add a product to cart, then go to the Shopping Cart page. Then go to the Start Checkout page. And then go through and complete the checkout.

      At each page copy the url you see in the browser and copy it into Paid Track naming it Step One and Two and so on and so forth. In the end hit Save and you'll have you conversion process in front of you. Start with a focus on the step that shows the highest abandonment.

      PadiTrack is an extremely easy tool to use.


      • 36

        Hi Avinash

        Thanks for the advice. I am on a bit of an extended Christmas break and am looking forward to implementing your advice. I will let you know how it goes.

        I am no guru on running a small online business but after seven years I can confirm to all your readers that your advice is spot on and I will be making a determined effort this year to implement your advice this year


      • 37

        Hi Aninash

        Thanks for all your help, can I please raise a micro issue

        I am having difficulty setting up a funnel in Paditracks. It all seems so easy but I cannot validate the first page of my checkout. The URL is

        When setting the funnel up I select

        1. The radio button "Page Path ?"
        2. Matches exactly

        There is no help section on Paditracks and I am really keen to get thongs up an running

        Would you or any of the other contributors have any advice or suggestions as to how I should proceed

        Many Thanks

        Colin Clear

  23. 38

    Excellent post! as usual :)

    I have quick questions:

    How can I get CTR metric for other channels except paid advertising in GA.
    I can see the Count of the visit metric in GA,but I can't find it in custom reports metric. where can we get that? ( I can see in the post that count of visit is greater than 4)

    Please clarify.


    • 39

      Sumit: You'll get the CTR from the ad channel you are using. For example your CTR for your email marketing campaign will be in your Email Service Provider's analytics platform (though some of these integrate with GA and Omniture and WebTrends and everyone). Your display campaign CTR will be in Yahoo! Display Campaigns analytics tool.

      Count of Visits is a Dimension Drilldown, please look in the Visitors folder. In the screenshot you are referencing is creating a segment not a custom report. But you can use it in both places.


  24. 41
    Sandeep Ail says

    Hi Avinash,

    Always love your posts. There is something that occurred to me now and realized that you are the best person to shed some light on this topic.

    We are now firmly in the age of smartphones. While website metrics are very relevant and important, there is an app ecosystem that is growing exponentially.

    Getting accurate sales and conversion data for digital marketing efforts for your app being sold in say iTunes store is a very big challenge for developers and digital marketers alike.

    This elusive data keeps you from taking accurate decisions for your digital marketing spend for the apps being sold in the appstore.

    You are probably the best person to decipher this ecosystem to provide some actionable data.

    Hope we can see some posts from you in 2012 on this front.

    Will also be very willing to brainstorm with you on this topic and share any appstore data that i already possess.

    Wish you and all the readers a very Happy New Year!

    • 42

      Sandeep: Depending on the provider you use (iTunes or Android Marketplace or Amazon Android Store) you'll get a bit more information or a bit less information. With a little bit of pain for Android Marketplace you can get referrer tracking for your sales and campaigns which is nice. More information here:

      But this is still more painful then it needs to be. As you mentioned… with the remarkable rise of smart phones and application consumption, it is only a matter of time before we get detailed insights into our sales (and user behavior in apps).


    • 43


      I think this is a very interesting issue and raises the whole issue of attribution.

      There is an article in the December issue of the Harvard Business Review titled
      " The Future of Shopping".

      In the article the author outlines how shopping behaviour will be influenced by technology
      and how retailers need to embrace this new technology. Apps and smartphones are most definitely part of the new technology ecosystem.

      The challenge for those interested in web analytics is how to attribute sales to a particular channel.

      Would be interested to hear from anyone else that has read the article.


      • 44

        Colin: There are many threads in your comment (a good thing), let me see if I can reply to some of them separately.

        It is increasingly common that multiple campaigns (even digital) touch the same person prior to conversion/goal completion. We have to get smarter at attributing value of campaign spend to those conversions. I personally believe that "attribution modeling" is a good step one but insufficient because there is no model that is sans bias / judgement / guessing. Media mix modeling is harder, but the right answer.

        It will become increasingly common that multiple devices influence our purchases online and offline. Some of this is impossible to know (me researching on my mobile phone while standing inside a Carphone Warehouse), others we might get better at. We are just going to have to live, learn and adapt.

        Finally I do not believe this is in any way a unique web analytics problem. No company has ever been able to attribute any value to Magazines or TV to any degree of confidence. In fact they are much worse than what we can do on a bad day with digital campaigns. This is a business analytics problem and we are only starting to understand it, thinking of clever solutions will come next. :)


        • 45


          I think you are right that media/marketing mix modeling gives a fairer/better answer than attribution modeling, but not only is it much harder, it is also significantly more expensive. In fact it is so expensive that only a very small number of large brands can afford to do it. But from my experience, the cost and complexity comes in large part down to scarcity of data: getting the right data, at the right level of granularity and for a long enough time series. The huge opportunity we now have is that we are moving from a world of scarcity of data to a world of abundance of data about customer behavior and attitudes – and much of that data is freely available. And with the rapid adoption of predictive analytics tools, many of them based around low-cost/open source tools like R, the other "hard part" of the problem is becoming easier to solve for even smaller organizations. But it takes knowing how to make sense of all of that data, understanding how to fit the data together in a way that mirrors the value-creating logic that a marketers recognizes, and knowing the data sources intimately. And getting that right is the job we have ahead of us.

          All in all, as you say, digital marketers do have it a lot better than offline marketers in being able to attribute marketing value – and it will hopefully be something that the confluence of big data, the commodization of business/predictive analytics and good old fashioned sound research design will make rapidly better.


  25. 46
    Abhishek Mishra says

    Hi Avinash,

    Many thanks for this brief intro about Metrics & KPIs; much sufficient details.

  26. 47

    Great article!

    I LOVE the assisted conversions metric! I wish I could implement something like that with my employers system (I use it personally and for all future clients)…however, I have yet to convince them the importance of channel segmentation (to my exhaustion- it makes my job harder!).

  27. 48
    Astonish says

    Hi Avinash,

    I am a newbie in this domain, but I have been following all you posts and every post of yours has helped me a lot.

    We are currently working on creating a digital engagement index for all our online customers. But the whole confusion is that we are neither looking at the page depth not looking at events per visit or the visitor frequency to the website. We only have customer transaction and demographics information to use.

    Our website has a few online services like bill pay etc. Now using these online services as segments we are calculating the overall profitability and demographic information for those segments. Then looking at the profitability of the customers using each of these online services we are going to rate the online services and then accordingly sum them up for each customer based on the usage to get a digital engagement index.

    Can you please let me know how correct or how incorrect we are?

    Thank you

  28. 49
    Krishnan says

    Hey Avinash, Awesome post, as usual!

    First Q:
    Assisted conversions in the L bucket (Acq, B, Outcomes matrix): Curious how you decided this metric should be in the L (rather than S or M) and ahead of a metric such as micro conversion.

    I'd have placed in S or M at least.

    Is it because its a relatively new metric (also not available in all tools) and small businesses may not have time/know how/resources to make use of it? Though useful for all businesses, for a small business, its critical to understand what step in the conversion process a particular channel plays (first, assisted, or last) and is make or break in designing the channel strategy – both communication/investment.

    Second Q: Bit tangential.
    Say, traditionally, the ratio of Assisted/Last is high for Paid (>3, assists more than closes sale) and low for organic (<0.2, closes more than assists), if there has been a reversal in the relative ratios (paid and organic both ~1) recently, what would be your first thoughts?

    Third Q: Can you recommend any resources that explore attribution modeling? I was hooked when I read about it in your WA 2.0 book.


    • 50

      Krishnan: Every metric could apply to every size of the business, so picking the right metric for a business is a lot more art than science. I used my experience, understanding of a business of that size and applied these criteria:

      1. Expected level of analytical sophistication available in the company.
      2. Expected level of complexity of marketing execution.
      3. Expected ability to actual use the data to drive meaningful action.
      4. My secret agenda at driving behavior right for a company in that bucket.

      By those criteria Assisted Conversions are not optimal for most Small and Medium sized businesses.

      To your second question…. the ratio would lead me to figure out how best to optimize the portfolio of channels I use. Other MCF reports (order of the touch point etc) would help glean those insights.

      To your last question… sadly most stuff on attribution modeling is probably out there via a Google search, here is no definitive location. Some agencies have put out whitepapers. You mention you have Web Analytics 2.0, as you read these whitepapers keep in mind the pros and cons I've shared for different models in the book. Be especially wary of the MCU attribution model.


  29. 51

    This is very useful and practical information – thanks for collating it and presenting in such an easy to consume manner.

    I have a doubt on the interpretation of the assisted conversion feature. While the feature does tell me the relevance of the source once they have come to the site, it doesnt tell me how effective the source is in click throughs (maybe a channel does well once people come in, but i lose a lot on CTR). Since the channels are all paid for in different terms (while Adwords is CPC, mail is almost CPM), how do we normalize for the overall effectiveness of the source.

    While I do understand that this information is crunchable in excel collating the data from all sources, is there a way of visualizing that too in an off the shelf GA report or by customizing it ?

    • 52
      Dave Rekuc says

      Guru, CTR never stops being an important metric for measuring the effectiveness of your ads. Once the visitor is on your site, you're now worried about converting them. Typically the last ad clicked prior to a conversion gets all the credit for the conversion, however, the assisted conversions report helps you see which campaigns or ads are being interacted with, but not being attributed a conversion on a typical conversion report.

      An offline example I like to use is someone calls your office and talks to salesman 1 for some information on your product. The next day he calls back and talks to salesman 2 that closes the deal and sells him the product. A typical conversion report would give salesman 2 all the credit for closing the sale, this assisted conversion report shows you that salesman 1 assisted in the sale by giving the customer the information they needed on that first call.

      I hope this helps clarify the use of an assisted conversions report :)

  30. 53
    Daniel Milstein says

    That is so true, Avinash. Bounce rates are a sign that you are getting the wrong set of audience. Those are folks who get to your content and realize it's not what they are looking for. CPR is definitely provides a good indepth analysis and insight on the relevancy of the ads.

    I really enjoyed reading your post. This will definitely help a lot of people the key metrics that really matters.

    Thanks for sharing.

  31. 54

    Mr. Kaushik, that was a fantastic article but I'd like to expand on a few points.

    In the first example (small business) I find the terms get confusing. Email and social media both have costs involved, but they usually don't have a direct CPC. For email services we use, it's all CPM — we're charged by emails going out, not per link clicked — and for social media, it's usually measured in a per-hour value over CPC.

    I know these are simple examples to keep the math easy, but I don't think they're really being taken into account properly — or the order is confusing. Is the CPC being dictated and thus the cost for email clicks is $50 and the cost for social clicks is $200, or is the CPC being set because of these associated costs? If it's the former then I think it's off-track, but if it's the latter then I think it needs to be expanded on. For example, if it costs the brand $50 to send out 1,000 emails does this number include the costs associated with acquiring the email address and writing the email? Does the social media cost include networking to achieve the connection and writing the post?

    Regardless, the summary is that in this case the social media order costs $15 more than the email order. If the cost to create and ship the product is $15, then the email order actually costs $20 and the social order costs $30 — in neither case is the brand 'shipping a crisp $5 bill along with every Social Media order'. They're actually sending out an extra $15 for the social, and $5 for the email.

    Instead, it would probably be better describing the situation in this way:
    In ONE HOUR (I'm using a time reference here since obtaining email addresses and social contacts takes time and would complicate things fast):
    It costs us $12.50 to get 1,000 emails opened (let's assume that's how much it costs to send enough emails that are opened by 1,000 people on the contact list out of who-knows-how-many people it was sent to that ignored/deleted the email — note I'm using the same cost ratio as the original).
    $50 to get 2,000 impressions via social media (let's assume in this case our social media expert is paid $50/hr and currently has 2,000 active contacts in the network)

    50 people clicked the email link and 200 people clicked the social link. 10 of the email clicks and 10 of the social clicks converted to purchase a product that costs us $15 to make that brings in $50 worth of revenue.

    The email's cost-per-conversion is therefore $12.50 / 10 = $1.25 while social's cost-per-conversion is $50 / 10 = $5.

    Since we charge $50 per product that costs us $15 to make, then the profits from email sales are $50 – $15 – $1.25 = $33.75. The profits from social sales are $50 – 15 – 5 = $30. While each social sale costs us $3.75 more, it still increases our total revenue by an additional $300.

    I still agree that it's important to drop things whose CPA are not optimal, but the first example used either has confusing wording or needs to include a simplified big picture.

    • 55

      Scott: You are absolutely right that this, as is often the case, is a matter of Garbage In, Garbage Out. In companies large and small it is very hard to collect correct total cost of any campaign.

      But we have to do the best we can because everything costs something, and we very rarely attribute cost to our marketing.

      Let's see if we can simplify… I think, broadly speaking, there are three types of costs involved in our marketing efforts:

      1. People (salary, benefits, retirement :))
      2. System (software – cloud or hosted, computers whatnot)
      3. Advertising

      So if you do paid search your actual cost is primarily 1+3. Cost of people involved, adwords is zero and advertising is the cost of clicks.

      If you do email marketing, as you say in your comment then 1+2+3. Cost of people, fixed costs for the email program and cost of sending each email.

      If you do social media your actual cost is 1. The person's time. At least for now. Or 1+2 (if you use a paid "enterprise social media presence management software").

      So on and so forth.

      Try to compute cost to the best extent possible across all campaigns. Try to get as close to all three of these costs as you can. Then make CPA decisions. It is possible you won't get perfection across campaigns, but it is a significantly more optimal place to be than not attributing any value to our acquisition efforts.

      PS: I'm assuming that once the person shows up on your site all other costs (serving site, taking the order, fulfillment are the same) to keep things simple.

  32. 56

    Thanks for this post!

    It does seem that the tools you are offering could simplify the process of analyzing performance of digital marketing campaigns and website. I will certainly look into this!

  33. 57

    Thank you for an insightful post. Having worked in online advertising for Microsoft I agree with all of your insights here.

    I'm actually in the process of writing a dissertation on public relations and metrics and have just purchased your book.

    Glad I've found your blog too!

  34. 58

    I've had this in my to-read list for over a month now but finally got to it and of course, it didn't disappoint.

    Thanks for the tips on which metrics to follow!

  35. 59
    Mike McGuire says

    Hi Avanash (or anyone)

    How can you figure out how many paid search visitors come back and purchase, how many times and how much do they spend… etc


  36. 61

    LOVE it AK! Especially the idea of focusing on CPA (not just on lead gen).

    I like to think of all of our small business clients' sites as if they were sales funnels. Of course, you want leads and sales, but it can be tough to stay committed to a marketing strategy unless you can ID something higher up in the funnel – like "increasing non-branded organic traffic."

  37. 62

    Hi Avinash,

    I defined my metrics in the following way, untill today found a better explanation from your end for acquisition, behavior and outcomes, which is very correct. Also it's comes directly from none other than You (Analytics Guru);

    Acquisition: Pageviews, Visits, New/Returning
    Behavior: Bounce Rate, Avg Time on Site, Pages/Visit, Loyalty & Recency, Navigational Summary of Landing pages
    Outcomes: Goals/Transactions,

    Lot to learn from you and your blog everytime i dig deep inside.

    Ranjan Jena

  38. 63
    christophe says

    To all,

    I am new in this business so please excuse my ignorance but when we speak about AOV has been the Sum of Revenue Generated / Number of Orders Taken. Does Sum of revenue generated include e.g. Transport cost, VAT etc… many thanks

    • 64

      Christophe: I did a quick Google search for "google analytics average value" and got this page with the definition:

      "Average Ecommerce Value: The average value of all transactions to your website over the selected time frame. This is calculated as the total value of transactions (not including tax and shipping) divided by the total number of transactions. The currency is the local currency defined for that particular site."

      Hope this helps.


  39. 65

    Avinash the master,

    How do you define the size of the business?..How do we know that the site we are handling fits the criteria of a mid size level?..

    Can it be done based on the number of total visits (excluding paid)? 10k monthly visits for small, 11k-100k for mid, 100k larger for large?

    • 66

      Jim: In some ways it does not matter what the definition of big, med, small is. See the goal I've set on top for where each company is in its evolution and what they are trying to solve for. Use that to identify where you are and use it. Perhaps the simplest way to think about it.

      But if it helps, Fortune 1,000 is a big company list. The company currently ranked 1,000 has revenues of $1.9 billion each year. Should give you a sense of size.

      Fortune 5,000 is medium size. Company #5000 today has revenue of $5 million.

      Rest is small.


  40. 67
    Tammie Jumalon says

    Quite beneficial, look forward to coming back and learning more.

  41. 68

    Hi Avinash,

    Brilliant article. Thanks!

    Do you have any benchmarks for "Non-transactional revenue per Visitor"?

    By non-transactional revenue, I mean the website monetization revenue. That would be really helpful for our business.


  42. 70

    Hi Avinash – thanks so much for the informative post, I love the simple breakdown of metrics…it's easy to get stuck in the weeds when you measure acquisition sources.

    I did want to let you know that the "Paditrack" link in your blog is broken. Upon further research I found that they discontinued their awesome free tool :( but Google has an alternative funnel tracker they published – here's the link to the article i found about this:

  43. 71

    Wonderful article. Pretty relevant even today.

    The insights of data science and analytics surely help understand the user behavior for better conversion rates.

  44. 72
    Brian Robben says

    A post like this shows I have a ton to learn when it comes to web metrics and optimizing my blog.

    I'm excited to dive into the action steps listed and learn as much as I can though. Please continue with the excellent content!

  45. 73

    Hello! This information is hugely helpful, and I enjoy your casual writing style. Just a small point of clarification. In your explication of "Cost per Acquisition" above, I believe there is a typo here: "Tip: Remember this is just cost, not profit. If your product costs you $15 to make then, in the above scenario, you are shipping a crisp $5 bill along with every Social Media order!"

    Per the provided image, I believe it should be $5 with each Email order and $20 with each Social Media order.

    Thank you for teaching me today, Avinash!

  46. 74

    I happened to read your blog as directed by the course of Introduction to Web Analytics by University of Illinois. The content to me is like a big sea as I am just infant for the subject. However I am able understand your mantra to some extent. But still I am not in a position to drop any comments/contribution.

    The content provide is greatly help me to understand the web analytic in three different categories of business and three different categories.

    I will check with the web team of my organization whether they do this kind of analytics on corporate web site. I will come if any to get clarify.


  47. 75

    Thank You Avinash. A big follower & admirer of your, and your analytics knowledge.

    Thanks for this beautiful article on the important metrics. It's important to record your targets and numbers on a daily basis while trying to meet the goals. Without numbers around the right metrics, there is no success or achievement.

    It's important to be clear what metrics meets your succes, and then recording on a daily basis to gauge your improvements.

  48. 76

    Very nice article…It gives lot of information, I love reading it and very easy to understand.

    Thanks Kaushik


  1. […]
    Best Web Metrics / KPIs for a Small, Medium or Large Sized Business,

  2. […]
    Avinash Kaushik, Best Web Metrics / KPIs for a Small, Medium or Large Sized Business, Occam's Razor, 12 Dec 2011. Explains the power of cost per acquisition, bounce rate, checkout abandonment rate, macro conversion rate, CPA + click-through rate, loyalty.

  3. […]
    Avinash Kaushik discusses the best web metrics/KPIs for businesses of all sizes at Occam’s Razor.

  4. […]
    Avinash Kaushik starts off our Monday morning with some great technical info on web analytics on his Occam's Razor site: Best Web Metrics / KPIs for a Small, Medium or Large Sized Business

  5. […]
    Best Web Metrics / KPIs for a Small, Medium or Large Sized Business.

  6. […]
    Indicatori de urmarit pentru un site mic/mediu/mare Avem la dispozitie o suita intreaga de date si unelte care pot ajuta orice tip de site sa creasca si sa castige mai mult. Ce e insa important sa urmaresti in cazul tau? Un articol foarte bun semnat Avinash Kaushik, cu exemple si idei pentru mai multe tipuri de site-uri, de la cele mici la business-urile mari.

  7. […] 网站分析中,获取大量数据并不艰难,艰难的是建立考核体系,获取有效的高质量分析见解,以实现数字化的网站运作与营销活动。本文主要从Avinash的文 章意译而来,有翻译不到位或错误,望大家指正,同时建议英文好的朋友直接阅读原文《Best Web Metrics / KPIs for a Small, Medium or Large Sized Business》。 图一:网站运营最佳KPI/度量 […]

  8. […]
    Avinash Kaushik/Occam's Razor: Best Web Metrics / KPIs for a Small, Medium or Large Sized Business

  9. […]
    Um beim Thema Web Analyse zu bleiben, direkt noch ein umfangreicher Artikel von Avinash Kaushik: „Best Web Metrics / KPIs for a Small, Medium or Large Sized Business“. Kaushik nennt die wichtigsten Metriken zur Messung und Optimierung Ihrer digitalen Marketing-Kampagnen und Website.

  10. […]
    Metrics matter; especially when you are using a system which can automatically optimize your process in order to maximize those metrics.

    Don’t get yourself in a jam; remember this next time you decide to measure click acceptance instead of actual sales to drive your online marketing effort. Clickthrough rates are useful as a measure by proxy, but they can be misleading.

  11. […]
    Best Web Metrics / KPIs for a Small, Medium or Large Sized Business (Englisch) Ein umfangreicher Artikel, der sich mit Metriken der Web-Analyse beschäftigt und einzelne Metriken passend zu spezifischen Unternehmensgrößen (Klein-, Mittel- und Großunternehmen) vorstellt.

  12. […]
    Best Web Metrics / KPIs for a Small, Medium or Large Sized Business (Englisch) Ein umfangreicher Artikel, der sich mit Metriken der Web-Analyse beschäftigt und einzelne Metriken passend zu spezifischen Unternehmensgrößen (Klein-, Mittel- und Großunternehmen) vorstellt.

  13. […]
    För att kunna göra en rättvis värdering av dina insatser måste det finnas en målbild. Och till skillnad från traditionell marknadsföring där man väntar till slutet av en kampanj för att analysera resultaten så sker digital marknadsföring i realtid och måste optimeras och justeras i realtid. Beroende på om du vill bygga kännedom, driva försäljning eller skapa engagemang bör du sätta olika mål. Begär alltid en s k ”post-buy report” från din mediebyrå. Exempel på mätetal: Klick, annonsvisningar, besökstrafik, bokningar, leads, sales, kommentarer, retweets, likes.

  14. […]
    Avinash Kaushik covers simple metrics for small business in his post Best Web Metrics / KPIs for a Small, Medium or Large Sized Business. Are you following these KPI’s for your business?

  15. […] bevakar sökord i sociala medier och analyserar även "passion" och känslaMätetal (KPI:er) för att analysera digital marknadsföring och försäljning Senaste rönen om vad som påverkar Googles sökresultatSvenska exempelOriflame – The […]

  16. […]
    First, there are a set of MR jobs for various descriptive analytic metric e.g., bounce rate, checkout abandonment etc. I find the blog site of Avinash Kaushik to be the best resource for web analytic. It’s better than reading a book. I am implementing many of the metrics defined in this post of Avinash. Second, I will have a set of MR jobs for predictive analytic on web log data e.g., prediction of user conversion, making product recommendation.

  17. […]
    Cela dit, miser sur une stratégie exclusivement fondée sur les médias sociaux présente un risque, dans la mesure où le modèle se cherche encore. Enfin, n’oublions la dimension budgétaire ! Un article intéressant a démontré par exemple que le coût d’acquisition dans une campagne email était 4 fois moins important que dans une campagne social média. Alors, ces métriques sont à suivre… de près!

  18. […]
    One of the easiest ways to find proven online metrics is to actually just Google "Key Performance Indicators (KPIs) for XXXX", where XXXX is your goal.
    For example, here's some great articles:
    KPIs for a Small, Medium or Large Sized Business:

  19. […]
    De vraag die hieraan voorafgaat is: ‘Wat zijn mijn belangrijkste KPI’s?’. Is het antwoord op deze vraag niet duidelijk, dan zal de hoeveelheid beschikbare data je eerder verder van je doel houden, in plaats van dichterbij brengen. Pas wanneer je snapt wat je wilt meten en waarom, kun je sturing geven aan de data die je verzamelt. Veel bedrijven schaffen dure analysepakketten aan, al voordat is vastgesteld wat het pakket zou moeten kunnen en wat er verzameld moet worden. Resultaat: een geweldig systeem en geweldige rapportages, die vertellen wat je niet nodig hebt en nog erger, die je afleiden van je werkelijke doelen.

  20. […]        网站分析中,获取大量数据并不艰难,艰难的是建立考核体系,获取有效的高质量分析见解,以实现数字化的网站运作与营销活动。本文主要从Avinash的文 章意译而来,有翻译不到位或错误的地方,望大家指正,同时建议英文好的朋友直接阅读原文《Best Web Metrics / KPIs for a Small, Medium or Large Sized Business》。 […]

  21. […]
    No discussion of social media ROI would be complete without mentioning Avinash Kaushik, Google guru and author of the blog Occams Razor. His many useful posts tell you which metrics are right for which circumstance. Two of my favorites are here and here … challenging reads, but a good way to start thinking about the metrics that matter.

  22. […]
    Cela dit, miser sur une stratégie exclusivement fondée sur les médias sociaux présente un risque, dans la mesure où le modèle se cherche encore. Enfin, n’oublions la dimension budgétaire ! Un article intéressant a démontré par exemple que le coût d’acquisition dans une campagne email était 4 fois moins important que dans une campagne social média. Alors, ces métriques sont à suivre… de près !

  23. […]
    There are a number of other metrics that can also be used to display how something the client is doing is – or isn’t – working. Do they think their product or service page is the bee’s knees, but it lacks substance and suffers from high bounce rates? These are action points you can improve upon to show value for the work you do. Not sure what metrics to look at or work with? Give this piece from Avinash Kaushik a read – it’s chock full of metrics to consider based on the size of the business in question.

  24. […]
    Do you know which metrics to track? 3 metrics beginners should watch Web metrics & KPIs for small businesses

  25. […]
    Your analytics data must have a purpose that tracks the purpose of your website. Your website’s purpose may be to drive sales, signups, or other conversions, or it may be an online equivalent of a business brochure. Whatever you want your website to do, tell your analytics tool what the objective is by focusing only on the Key Performance Indicators relevant to that directive.

  26. […]
    Best Web Metrics / KPIs for a Small, Medium or Large Sized Business… via @avinash

  27. […] * Articles on Bringing the Sexy Back Out of the Box Best Web Metrics […]

  28. […]
    If you need more help, the amazing Avinash Kaushik lays out his best web metrics for you in this blog post. The best piece of advice I can give here is both tried and true: stay away from vanity metrics.

  29. […]
    Additional free resources: 1) review and study Avinash’s blog post on the best metrics and key performance indicators for small, medium, and large businesses.

  30. […]
    Long article on business metrics/KPIs by Avinash – check out the first section on customer acquisition.

  31. […]
    After you have defined your business goals, Avinash Kaushik recommends looking at critical metrics/KPI’s using the Acquisition-Behavior-Outcome Framework. Using this framework will show you the end-to-end journey of your visitors. Although some KPI’s vary per company, here’s a breakdown of metrics that every company should measure, no matter what.

  32. […]
    Pick key metrics to monitor and choose goals that tie in with your business objectives. The metrics that a small business should track are Cost Per Acquisition, Bounce Rate, Checkout Abandonment Rate and Macro Conversion Rate. For a medium sized business some additional metrics include Click-through Rate, Page Depth, Loyalty (visit count per person), Micro Conversion Rates and Per Visit Goal value. For a big business some other metrics are % new visits, Events/Visit, Days to Conversion and % Assisted Conversions. Check out this awesome blog post by Avinash Kaushik for more details.

  33. […]
    “Reporting squirrel (riportoló mókus)” bárki lehet, mondta Avinash, de “Analysis Ninjának (elemző ninja)” kell ahhoz lennünk, hogy értelme is legyen. És ennek egyik kulcsa a helyes kérdések feltevése, ami komoly kreativitást igényel. Avinash beszélt még arról, hogyan választhatja meg egy kis, egy közepes és egy nagy cég, hogy milyen KPI-kat használjon, hiszen teljesen más erőforrásaik, céljaik lehetnek.

  34. […]
    Only post content that is “incredible” and of value to your audience. Don’t put time and effort into creating content that no one cares about. Pay attention to the end-to-end customer experience, and focus on acquisition, behavior, and outcome metrics.

  35. […]
    It is difficult to measure and track all of the important metrics separately. That is why we have dashboards to give us an executive overview of the overall performance. Now you know how to decide what needs to be featured on your dashboard. The metrics you track should help you craft your action list for improvement. A great starting point is the overview of best web metrics KPIs for small, medium and large business. Before creating a dashboard, understand what makes a dashboard actionable. As an example, use these downloadable custom reports that target your report audience.

  36. […]
    Avinash Kaushik on key metrics for small, medium-sized, and large ecommerce sites.

  37. […]
    KPI’s or “What Do You Want to Know”: Think about how Acquisition, Behavior, & Outcomes impact your business. Check out Google’s amazing Avinash Kaushik and his views on KPI’s.

  38. […]
    Obviously at the end of the day you care about your business making more money while providing value to your customers. How you define your key performance indicators (KPIs) will give you an idea of which kinds of conversion rates you should be optimizing. Don’t be afraid to include as many conversion KPIs as you think are relevant to your success. Think in terms of purchases (if applicable), contact requests, content downloads, quote requests, email newsletter signups, form completions, store locator searches, or any other step in your sales process that gives you more information about your prospect and gets them closer to becoming your customer.

  39. […] Auf seinem Blog erläutert Avinash Kaushik die seiner Ansicht nach besten Analytics-Metriken für kleine, mittlere und große Unternehmen. […]

  40. […]
    Not purely B2B, but these are Avinash’s fav KPIs:

  41. […]
    Spiller innholdet ditt på lag med de strategiske målene dine? Hvor er du, hvor vil du, og hva trenger du for å komme dit? Tenk brukervennlighet og tone-of-voice, men også om innholdet ditt hjelper deg å nå målene, eller KPIene dine. Avinash Kaushik har skrevet konkret om gode KPIer her.

  42. […]
    Welke data heb ik nodig? De vraag die hieraan voorafgaat is: ‘Wat zijn mijn belangrijkste KPI’s?’. Is het antwoord op deze vraag niet duidelijk, dan zal de hoeveelheid beschikbare data je eerder verder van je doel houden, in plaats van dichterbij brengen. Pas wanneer je snapt wat je wilt meten en waarom, kun je sturing geven aan de data die je verzamelt. Veel bedrijven schaffen dure analysepakketten aan, al voordat is vastgesteld wat het pakket zou moeten kunnen en wat er verzameld moet worden.

  43. […]
    Digital Design & user Experience Best Practices: Happiness + Profits (Blog Post by Avinash)
    Best Web Metrics / KPIs for a Small, Medium or Large Business (Blog Post by Avinash)
    Market Motive (Avinash’s Online Marketing Training Company)

  44. […]
    Avinash in one of his blog posts, strongly advocates aligning the KPI as per the organization goals. So, he goes about segmenting the objectives as per – Acquisition, Behavior, and Outcomes, followed by Organization Size.

  45. […]
    It takes a while for the majority of visitors to a site to convert, according to Google digital marketing evangelist Avinash Kaushik, so any evaluation needs to take an “end-to-end” view of the various important activities.

  46. […]
    7. Read these two articles on 3 Metrics Google Analytics Beginners Should Watch and Best Web Metrics / KPIs for a Small, Medium or Large Sized Business

  47. […]
    These are all questions you can answer with hard data using totally free tools like Google Analytics that are painless to set up and really powerful. If you really want to get into this, there’s a great post by Avinash Kaushik, that’s worth a read.

  48. […]
    In a blog post, Kaushik identified the most important metrics for different sizes of business:

  49. […]
    Kaushik, Avinash (2013). Occams Razor. Best Web Metrics / KPIs for a Small, Medium or Large Sized Business.  Retrieved at:

  50. […]
    El tener métricas de marketing digital adecuadas te permitirá tener mayor control de los datos, instrumentos y resultados generados de tu estrategia de marketing digital. De este modo podrás saber si tus resultados están siendo alcanzados o si tus esfuerzos de marketing no están funcionando. A continuación te muestro una tabla derivada de la publicación de Avinash Kaushik, donde se agrupan métricas de marketing digital de acuerdo al tamaño de tu empresa:

  51. […]
    Eine detaillierte Beschreibung der Kennzahlen-Empfehlung von Avinash Kaushik gibt es hier. In dem Artikel “Ab durch den Sales Funnel – Wie man potenzielle zu zahlenden Kunden macht” findest du weitere Empfehlungen speziell für den Sales Funnel.

  52. […]
    Wenn du nach einem schnell umsetzbaren Kennzahlenpaket suchst, dann sind die Kennzahlen von Avinash Kaushik​ die richtigen für dich. Sobald du diese im Griff hast, kannst du immer noch weitere Kennzahlen hinzunehmen.

  53. […]
    Eine detaillierte Beschreibung der Kennzahlen-Empfehlung von Avinash Kaushik gibt es hier. In dem Artikel “Ab durch den Sales Funnel – Wie man potenzielle zu zahlenden Kunden macht” findest du weitere Empfehlungen speziell für den Sales Funnel.

Add your Perspective