Google Analytics Visits Metric: Change. Implications. Opportunities.

symmetry 3Recently the Google Analytics team changed the way sessionization algorithm. In lay terms… how the start, duration and end of a Visit is computed.

A minor version of the butterfly effect occurred, one small change in a part of the system caused a few other smaller changes in other parts of the system. Some people freaked out. Others wondered what the fuss was all about. Still others wondered what they were going to eat for lunch. :)

If you were in the first two categories this blogpost (delightful video actually!) is for you.

Some of you know that I am the co-Founder of Market Motive, and also lead the Web Analytics Master Certification course (which is one of eight courses in Internet Marketing certification). The course is a structured program that teaches the gamut of web analytics. You start at the beginning (I know!) and evolve to an Analysis Ninja-hood.

Here's the rough path: Web Analytics 2.0 immersion > Adv KPI analysis techniques > Segmentation (Oh I love this so) > Actionable reports > Surveys > Testing > Multi-Channel analytics > Competitive Intelligence > Dissertation > (If you pass the rigorous evaluation) > Certification.

A key strategy for having the most kick butt course in the world is that it is astonishingly fresh. In addition to the structured curriculum we do weekly hour and half calls. Guess what the latest video in the Web Analytics Master Certification course was? Did you guess GA Visits changes? Ding, ding, ding!

I wanted to make an exception and share that video with all of you. It will hopefully give you an idea of the love and unique approach that Market Motive brings to teaching, along with sharing some very critical context and guidance on this important issue.

This important video covers:

00:00 – 02:14 Intro. Which metrics are impacted?

02:14 – 09:20 Context. How do web analytics tools currently compute visits? And if you use Omniture, WebTrends, CoreMetrics, Piwik, what important question should you ask your vendor.

09:20 – 12:45 Impact. Estimating how much the change impacts you! Details, and grab the custom report.

12:45 – 17:50 The change. Clear articulation of what the change actually was. The new sessionization algorithm.

17:50 – 21:48 The implications. With examples that impact on Average Page Views per Visit, Bounce Rates, % New Visits, % Returning Visits.

21:48 – 31:20 Multi-Channel Funnels. Impact, new sweet yummy implications, and digital media attribution analysis. The one report that will probably cause you to hyper-ventilate a little bit.

31:20 – 37:02 Closing thoughts. My four point guidance on the implications and opportunities.

There is one important thing the video does not cover. At the time Google Analytics switched over to the new sessionization algorithm, coincidently a bug also impacted GA users data. The bug, temporarily, made it seem as if the impact was crazy because of the sessionization change. This bug was identified, data was fixed, that "craziness" is gone. Hence I'm skipping covering that.

Okay are you raring to go? Sink your teeth? Jump into the knowledge pond?

Here's the video:



Full of surprises, right?

I hope the video shared the reports and data you need to figure out for yourself in what way this change impacts you, the context you need to figure out how to internalize the change and share concrete information with your management team (rather than the FUD so easily accessible via tweets, Plus, Facebook posts, etc). For some of you the changes might be larger than the range I've mentioned in the video. But consider this: You know have a MUCH better idea of how your customers behave. You would not want to be blind to that (and have no change in data!). I suspect not.

I particularly enjoyed sharing with you the multi-channel funnels part, and the implications on attribution analysis. Is it not amazing that without the monkey gyrations – JavaScript tag updates, variables capturing, praying for rain – other tools might put you through (*if* they do multi-channel unique visitor analysis), that you can get this amazing data?

Beyond what's in the video consider the Assisted Conversions report, Top Conversions Paths, Time Lag and Path Length details. Earning a load of money and happiness because, finally, you'll deliver big consistent impact to the company bottom-line is now reality.

In closing… Change is always hard to accept, especially when it comes with even the slightest impact on status quo. But if there has to be progress in life, then change is just the thing that puts us in a higher, more optimal orbit. It makes a better existence possible.

Go give the new data and reports a try. Thinking in a new way will require effort and brain power. But real happiness is worth it.

All the best!

As always, it is your turn now.

Did the video clarify how the visits algorithm works today? Did it help you internalize the change, and its value? Given my scenario of Google > Bing > Email > Mom > Conversion, would you have changed the GA sessionization algorithm in a different way? If you downloaded the custom report, how much did "zero visits" (multi-visit sessions) impact your company?

Please share your feedback on the video format, and the content of the video via comments.

Thanks.

PS: A small present for you. I've created a special page that collects every single thing I've written on this blog. All 443,192 words (as of today). The posts are categorized into 11 clusters, making it easy for you to find what you are looking for.

Access it here: Digital Marketing & Analytics Knowledge

Comments

  1. 1
    Joe Teixeira says:

    Hi Avinash!

    I watched the entire video – it's great! I think that everyone reading should take the time out to watch it. It's 30+ minutes but it's worth it, even if you're an expert in the field. However, I will say that my Mom has JavaScript tags so she can count as a Campaign :).

    Some thoughts:

    – I think that, more and more, people are starting to accept that there are data "imperfections" (things like 0 Visits and multiple campaigns per session) in web analytics. Which is good, IMO, because I find that once people can get over that hump (that their data is not perfect), then real analysis, insights, and actions can happen.

    – This video is excellent because it brings to light not just the change in sessions, but also how web (Google) analytics considers a session. Custom reporting is a great teaching / educational tool for this. It reminded me of the following Custom Report that I created a few months ago to show how some of the Entrance / Exit oriented metrics out of Google Analytics work and how they relate to pages:

    http://goo.gl/6wIdv

    Although Visits isn't a metric on this report, I would try to see what happens when you add visits to this group. Notice anything "funny" about Visits (and how it relates to another metric on it)? This will show you what a Visit really means in GA.

    Thanks Avinash!

  2. 2
    Nick says:

    Google analytics made powerful update recently, Thanks for sharing tips related to that.

  3. 3
    Himanshu says:

    Hi Avinash! Your teaching style is superb and i am a big fan of your voice. You may not know, but you have tremendous voice clarity which makes it very easy to understand what you say and the way you change the intonation of your voice and reinforce important points during presentation, make concepts go right into the head. It is really pleasant to hear you. I infact urge you to come up with more video tutorials and turn this blog into a v-blog. And just for the sake of SEO and some others who aren't blessed with high speed net connection, you can always include a video transcript at the end of each video.

    Now back to the post. I think this new change is going to have a tremendous impact on conversion rate or any metric which uses visits in calculation. Majority of businesses still compute conversion rate on the basis of number of visits. Since visits will go high,the conversion rate should go down. I find it really difficult to convice the senior management that conversion rate should be calculated on the basis of unique visitors and not visits as Google itself calculates the CR on the basis of visits (People tends to believe Google more than me). SEOs who don't understand this difference will have a hard time explaining why the website CR has gone down. Do you have any news regarding Google changing this behaviour any time in the future and start using visitors as a main metrics esp for calculating CR.

    • 4

      Himanshu: In both of my books I've spent at least some time emphasizing the value of using Unique Visitors when computing conversions rather than Visits.

      Using Visits for conversion implies that your marketing, and web experience, is built to convert on every single Visit by a consumer. Jump into bed right away. Most customers don't behave that way. Using Unique Visitors means your start to think of doing business with a person at their pace.

      The "standard" definition of conversion Rate does use Visits in pretty much every tool, GA and Omniture and WebTrends and the rest. That is not expected to change anytime. But the GA API and Custom Reports means that it is increasingly easy for us to do that computation ourselves – if we want to.

      Thanks for the kind words about the video, I am glad you found it to be of value.

      Avinash.

      • 5
        al says:

        Hi Avinash,

        Thank you for your videos and your great book!

        I did have one question that I cannot seem to find the answer to in any of the Google Analytics books I have bought. If you had two URLs in your all traffic sources report that differ only by a port number and you wanted to combine the two URLs into one, how would you do that? http://www.domain.com:80 and http://www.domain.com –> http://www.domain.com

        I just can't seem to get that search and replace filter working right no matter what I try.

  4. 6
    Ojil Casi says:

    Not Able to find the multichannel funnels. Avinash any earlier posts from you on where this can be found?

  5. 7
    De4e says:

    Thanks for Post. New tools are impressive.

    There are little bug with number of conversion since 8/28(0's only in MultiChannel reports) , hope it will be fix soon.

    I've already find few interesting points, thank this stuff. By the way – interesting to know, that many peoples click same ad in ppc more than 2 times.

    At last there are no need to use tools like "multitouch analytics" or parse google cookies now.

    • 8

      De4e: If conversions between the standard goal reports in GA and MCF reports don't match then that is expected behavior at the moment. MCF is moving mountains and picking up boulders and looking for the needles in the haystack. For now, and just for now while in beta, there is a slight delay between the GA standard reports (near real time) and MCF (24 hrs).

      There is more context in this help article:

      https://www.google.com/support/analyticshelp/bin/answer.py?answer=1319312

      This delay bit will of course change with time.

      But want to stress that with MCF you are not doing tactical analysis. You won't learn anything from looking at what happened in the last 24 or 48 hours. This is by nature strategic analysis and you'll look back last several weeks or a month. In those cases this type of a delay matters less (even if the delay will ultimately go away).

      Avinash.

  6. 9
    James Gurd says:

    Hi Avinash,

    Thanks for such a useful video – really helpful and easy to follow.

    One question based on your scenario:

    If the email contains 10 links and each link has a different tracking code to enable analysis at link level (so same source, medium & campaign UTM but perhaps different Content UTM), if the visitors clicks on 3 different links from the same email, will the new GA measurement of visits then count that as 3 separate visits?

    If so, that to me is going to be a headache for analysing the conversion cycle because it might distort the number of stages.

    I'd be interested in your thoughts on this.

    Thanks,

    James

    • 10

      James: If the email links have the same campaign parameters then it does not matter if the person clicks on all of them to come to your site (before the 29 mins of inactivity expire). It will be one visit.

      If each email link contains different parameters then per the new algorithm each click will be a new visit.

      It would be important in this context to evaluate if you really need to have each link have its own parameters. Most of the time people don't need that. But if you do need that then there might be a different way to collect the "this was a different link in a different part of th same email campaign" information. Perhaps use event tracking or custom variable after people come back. Perhaps use segmentation, based on where each link landed. Something else, depending on why each link had its own complete set of parameters.

      Thanks,

      Avinash.

      PS: Please also checkout my reply to Max on email campaign individual links tracking below. It provides a solution to the issue you've highlighted.

  7. 11
    Josh Braaten says:

    Thanks for explaining the changes occurring with sessionization. It certainly seemed like a necessary evolution along the way to a more robust future of Multi-Channel Funnels. I've been playing with funnels since they launched and am in love. Inbound marketing got a new ally in MCFs… I just hope enough marketers brave the learning curve and master the insights provided.

    BTW, Joe hints that you may have coined the newest trash talk in web analytics: "Your mother has JavaScript enabled!" I tend to agree with him. :)

  8. 12
    Kirk says:

    Thank so much for sharing that video with us.

    I have my head in the sand and was not even aware of these changes (I must admit I have been holding out on switching to the new GA UI for as long as possible and so I miss some of these new changes from time to time), so it was a welcome wake-up call and informative video.

  9. 13

    Thanks a lot for this video! This is very helpful, not only to validate I understood this visit change right but also to make sure I can thereafter spread the correct information internally :)

  10. 14
    Kris G says:

    Hi Avinash,

    The video was an amazingly helpful video that answered a lot of my questions about GA (and other tools) tracking methods. I have always been unclear and have heard mixed thoughts on some of the exact questions that this video has answered.

  11. 15
    Max says:

    Are you kidding me? A different sourceID will trigger a new visit? So within an email, I have 2 links with different sourceID(headerlink, and footer link) that goes to the same page, if I click on each link once within a minute, it will be counted as 2 visits?

    Most analytic tool has multi-channel attribution solution without this change. I honestly think this updates is meant to attribute more 'visit' to Google search result, appearing Google is bringing more visit to your site. And I cannot believe you justify this updates base on one particular made-up scenario.

  12. 20
    PP says:

    Pretty philistine comment, but looks like there is a residual word after you edited the first sentence. :)

  13. 22
    Nt_1 says:

    Thanks for this post, very clear and … reassuring for those puzzled by the change ! Google's lucky to have you as their evangelist…

    The one thing I would point out though, is that the Custom Report looking for "Visitors with 0 Visits / by the segment: Keyword" will only show Keywords which were used exclusively as 'second in a session'. If the same keyword was used as the first in a session, the Unique Visitor count it will be aggregated in a line that does not have '0 Visits'. So unless I'm mistaken, it is not a measure of the whole 'multi-source Visit' phenomenon, but only measures a small subset of it.

    In some cases (especially long tail websites) I found that using Pages as a Segment gives much more accurate results – typically closer to 20/30% than to 1/2% – but I think the exact data is not available as it would need accessing non aggregated data.

    Or maybe it's just too late and my brain has stopped functioning properly?

  14. 23

    hi avinash, great post… this is quite a big news, but it's live right now?

    i still have to play around to understand how does it work properly, but it kinda reminds me "istances" in omniture, not exaclty the same stuff but is kinda similar… maybe like them you follow this path also for computational reason ?

    for sure will be easier to explain to client how does this work, is more "natural" and less "technical".

    in the end, don't forget to get this up to date http://www.google.com/support/conversionuniversity/
    :)

    greetings from italy :)

    tommaso

  15. 24

    Hi Avinash,

    Great video – your booms ripped my eardrums out haha!

    One question – correct me if I'm wrong, but you are saying bounce rate for certain sources may increase due to the session change – but this should not mean that this channel is discounted if it is significant in the funnel reports, right? It maybe be a big player in the assisted conversions (or top of the funnel), but the bounce rate signifies a bit of work in keeping people interested, correct?

    Cheers,

    Jon

    • 25

      Jon: I want to be very cautious about this.

      What Google Analytics will show you is how your customers are actually behaving. If they come from source one, do nothing, leave, come back later, that's a bounce. That is all I would read into it. We can try and understand why these people leave etc. [Bonus helpful post: Six Tips For Improving High Bounce / Low Conversion Web Pages]

      If they come back again and convert etc then of of course the new reports will properly credit the source. Still does not mean you should not try to understand why they bounced in the first place! :)

      -Avinash.

      • 26

        That's what I meant, you put it much more eloquently!

        Another question – in the multi channel funnels, is "Social Network" channel made up of:
        custom campaigns tagged as such only,
        referrals from facebook, twitter etc only,
        or both (and anything else that Google Analytics deems as being from a Social medium/source)

        This doesn't seem clear from Google help
        cheers
        Jon

  16. 28

    Great video! I for one welcome the change :)

  17. 29
    Ameya says:

    Hi Avinash,

    As usual a very informative post. Just a doubt…. in the Custom Report when I enter Landing Page instead of Keyword Dimension the numbers change drastically…couldn't figure out why…your views would help!

    Thanks

    • 30
      Nate S. says:

      Ameya,

      Not sure if anyone ever answered your question, but in this case I think it's because the 0 Visit filter is considering a Visitor's 2nd, 3rd, 4th (and so on) entries back into the site. Typically those won't have a qualifying landing page, because they're continuations of the original visit.

      I am curious though, Avinash, how to get more than just Search Engine traffic in the report you've provided. Can I get a count of any visitor who came in via multiple touch points? Source didn't seem to do the trick as a dimension in a test I ran.

      Thoughts?

      Nate

  18. 31
    Sridhar says:

    Hey Avinash,

    This is an AWESOME video and I really liked it!

    I made every marketing guy in our office to watch this video! And even they liked it.

    Thank you for your video and I would also like to know, how the same reflects to our adwords campaign.

    Even I see the same in Adwords interface like zero clicks but one conversion? How does it happen, can you please explain me?

    Regards,

    Sridhar K

  19. 32
    Pascal says:

    What about visits (from the same visitor withing 29 minutes) that came via google but used a different keyword? Are they also counted as different visitis? Do they also have a different attribution in terms of "assisted conversions"?

    Cheers
    Pascal

    • 33
      Himanshu says:

      Hi Pascal!

      Any change in traffic source (paid search, organic search, email, ad version, keyword) will result in a new visit. So if a visitor come to your website via 29 different keywords on Google (even within 30 minutes), it will still result in 25 different visits. All keywords will have different attribution in terms of assisted conversions.

  20. 34

    Brilliant video: you explain a very complicated concept in a way which makes it very easy to understand what was happening in the past and is happening now.

    I also like the way you do remind us of the concept of micro-conversions, such as engagement goals, towards the end. I think these new reports will teach us to value those far more highly.

    Having said that, I also wish there was some way of configuring the reports so that they don't necessarily start with the default view which summarises all the conversions for all the goals. In the case of an ecommerce site, for example, it might be better if the default view was based on transactions not the total of all the goals.

    But for now it's enough to remember to explore the menu, just as you showed in the video.

  21. 35
    Paul Tyler says:

    Brilliant, thanks Avinash and love your enthusiasm and excitement. This explained all the changes much better than any post I have read so thanks.

  22. 36
    Judah says:

    Monkey gyrations. LOL.

    I'm taking that one (but I will give you multichannel attribution).

  23. 37
    Kevin says:

    Avinash, always love your post.

    I have a site though that is still showing 0 after the 18th of August any reason this may be? I have set the custom report exactly as shown and still have 0 with UV and PV recoding numbers and 0 visits just for the 29th of Aug.

    Kevin

  24. 38
    Calin says:

    Thank you for a great video, Avinash. I've been waiting for multi channel funnels since I saw them as USPs for a paid Analytics solution.

    What's still unclear for me (and may have to test) is what happens when:

    – a visitor arrives from organic search, closes the browser, remembers the website URL and returns direct (types URL), after less than 29 minutes, converts

    – another visitor arrives via referral link, leaves, returns direct after 31 minutes, converts

    Who gets credit, after Aug 18, for the visit, and the conversion, in each case? Is "direct" still stripped of any merit except a mention as the last step in the multi-channel funnels?

    If my obsession for direct traffic and its misteries shows, it's because it's there ;).

    Calin

    • 39

      Calin: The cases you are describing have nothing to do with these changes, they would work exactly the same way in any web analytics tool.

      First case: 1 visit. Original referrer gets credit for conversion, organic search.

      Second case: 2 visits. Originally referrer gets credit for conversion.

      The recent changes impact multiple visits within the same session where the visits are things other than Direct.

      If you want to learn about Direct traffic and how it works…

      ~ Excellent Analytics Tip #18: Make Love To Your Direct Traffic

      -Avinash.

      • 40
        Calin says:

        Thank you, Avinash. I've read that post when it was published, I'll read it again. And again.

        Calin

  25. 41
    Charlie says:

    Great video Avinash! I was trying to understand what really happened on Google Analytics since some days ago, and your video helps me a lot!

    The only data I don´t get is "0" -as you showed on the video- when checking analytics after August 11th. I´m using Michael custom report for the New Google Analytics… but seen no difference on data.

    My next task is to learn about the multi-channel attribution ;)

  26. 42
    Modi says:

    Awesome video Avinash, thank you for sharing!

    Even though Google said the recent change will have a 1% impact in the analytics results, that 1% for large websites & organisations is a massive figure. Being able to understand the new algorithm and how sessions are being perceived is crucial.

  27. 43
    Palashi says:

    Hi Avinash,

    Great post, really good pictorial representation of the whole issue clearing up all the doubts about the visit change.

    In your post you mentioned that the change in the visits has impacted other metrics as well, would you say that it has impacted the imp metric ecommerce conversion rate too, and what impact would this issue have on all the other ecommerce metrics using visits in their calculation and what is your take on it ?????

    Regards,
    Palashi

    • 44

      Palashi: Ecommerce and Goal Conversions are defined as Outcomes divided by Visits.

      If the number of Visits change then it will impact the Conversion Rate number. Not the total number of Conversions of course (making it every more easy to show context – in this case the raw number).

      You will now also get a new number, Assisted Conversions. The number of times ad / media channels assisted in conversions happened in the future via other media channels.

      Avinash.

  28. 45
    Kim says:

    This video has been extremely helpful. But I still don't understand how to interpret the Multi-Channel Funnels Overview when specifying Conversion Goal or Type.

    When I look at my Goal #3 it shows 13 Conversions and 2 Assisted Conversions, but the Visualizer doesn't show the assisted conversions. Instead it just shows Organic Search: 7.

    I can't figure out how to make heads or tales out of this data.

  29. 47
    Rohan says:

    Great article and video.

    When I look at the MFC data for my site I see that the Organic Search conversion area is completely embedded in the Paid Search conversion area.

    What is the inference you draw from this? Are we spending money on SEM where SEO is already effective?

  30. 49
    Matt says:

    Thanks for sharing the very informative video.

    I am not finding the multi channel funnel anywhere within my GA accounts. Does it have to be activated somehow?

    • 50

      Matt: MCF is only available in the new version of Google Analytics. Click on the link that says New Version on the very top of the page.

      Then in the orange bar you'll see a navigation element called My Conversions. Click on that. Then click on Multi-Channel Funnels. Boom! You're in business.

      Avinash.

  31. 51
    Len says:

    Turns out is doesn't affect us at all but still a great video, learnt lots. I love the idea of putting javaScript on moms.

  32. 52
    Gonzalo Paternoster says:

    Hi Avinash!!

    I don't know if you remember me. A few months ago you provide me some advice when I was hurting and you made a huge difference in my life and the life of my family. So thank you!!!

    I have a quick question.

    1. I have a site that is mostly an affiliate site with outgoing links so most of my visitors visit only 1 page and say a short time on the site. To help with my stats for Google so I be doing nofollow links for them, "_blank" target links etc. Any advice would be great.

  33. 54
    dasu says:

    hi Avinash,

    nice post!

    i have a question here, since the update to session happened, i am now seeing almost two visits for each adwords click , for an adwords ad clicking through , if any traffic source information changed(utm_term or glicd), the new visit generated should not be counted to the keyword be searched, so is this making sence my adwords visits is twice as high as my adwords click?

    • 55

      Dasu: Here is how it works….

      Scenario 1: I got to Google. Search for for Lightning. Click on an Adwords ad for Lightning. Visit the website. Browse and leave and never come back again.

      What will Google Analytics show? One Visit to the site.

      Scenario 2: I got to Google. Search for for Lightning. Click on an Adwords ad for Lightning. Visit the website. Browse. Then go back to Google. Search for Pink Lightning. Click on an AdWords ad. Come back o the exact same site again.

      What will Google Analytics show? Two Visits. One for the keyword Lightning. One for the keyword Pink Lightning.

      I hope that helps.

      Avinash.

  34. 58
    Andy Miller says:

    Great post Avinash, your enthusiasm is infectious! The video was very simple to follow and understand.

    Using the custom report on some of our retail clients we have found that the "zero visits" from multiple visit sessions impacted (on average) just under 1% of traffic.

    The MCF's are brilliant, we are finding some absolute gems of insight which clients are also loving. BOOM :)

  35. 59
    MIke says:

    One item that we still do not understand in Analytics is bounce rate. I found a post you did in 2007 but not recently. Could you provide an exact explanation of how Google determines it. The one key item of confusion is if an ad is clicked on entry, is it considered a bounce? Also is there time on page that would not count it as a bounce? Thanks for the help.

    • 60

      Mike: Any session (visit) where there was only one "hit" stored is a bounce. That hit can be a page view (this is most typical), or it can be an event or click or some type or anything else.

      So simplest explanation (for any tool not just GA): Visits with single page views are bounce.

      If you've implement the WA tool correctly then none of these things are a bounce:

      I come to your site, see one page, click on a ad and leave the site. Not bounce, you have to ensure the outbound click is captured as an Event or "Fake Pageview."

      I come to your site, see just one page, click play on a video, watch the whole thing, leave. Not bounce, you have to ensure you have video tracking implemented correctly.

      I come to your site, see just one page, click on a link to download a pdf, leave. Not a bounced visit, you have to ensure you are capturing the download click as an Event.

      So on and so forth.

      If you don't know how to do any of the above hire a consultant who can help you: http://www.google.com/analytics/partners.html

      All the best!

      -Avinash.

  36. 61
    Atle Egenes says:

    Hi

    Great video! I have one question to the Top Conversions Path. In my stats its rather strangely grouped with the same sources chunked togheter and no mixing of them: http://twitpic.com/6lbb6l

    Any idea why this happens?

    • 62

      Atle: It is simply showing you that in the third path the same person came back twice via Organic search before converting. In row four it was three times.

      It is hard to read anything from that single screenshot (and not other context of your business) but what it is showing is that in your case, for some reason, people just tend to come via one channel before converting (though they use that channel multiple times). A little weird, but nothing wrong with that. :)

      Avinash.

  37. 63

    Hi Avinash

    Have just watched your video and found it very interesting regarding how Google is now evaluating visits. So whilst a number of visits via Google, Bing and Email within a 29 minute timespan would have counted as one visit before, which would have been not very informative, now the information is more correctly recorded as three visits.

    I like the way you explained how to calculate the impact of organic search v other methods such a social media. Considering the financial value put on Facebook and Twitter the actual impact of their channels seems to be nominal. Could this be because of your own activity in these channels or would that be a fair assessment of the impact of social media v organic search?

    Anyway great video delivered with enthusiasm and clarity.

    • 64

      Joe: The value of any channel, in your case Social, will depend on how you participate in it and how much of your audience is there. Done right, and if a large percent of your audience is actually there, it can be quite a productive channel.

      The great thing is that you don't have to decide based on your gut. Let your data guide you. In my case I compute the Per Visit Goal Value (standard metric in Google Analytics) for each of my acquisition channels and use that to decide if i should do Social more or Search more or Affiliate or… other things. In my case it turns out that Search simply dominates but

      All you need is to ensure you are tracking the economic value of your Macro plus Micro Conversions. Here's a post that might help: Compute Website Economic Value

      Avinash.

  38. 65
    Rendy says:

    Has anyone told you how nice to see the video content being explained/broken down with time ranges in the post?

    It really got me into watching it and I can focus straight on the bits that I need.

    Well done, Avinash!

  39. 66
    Rammesh Perumal says:

    Completely refreshing and a lot info to digest about Google analytic for me. A great video how to use the analytic stats to improve our website performance. Thanks for sharing.

  40. 67

    What's the change of the August 12 session change in this case?

    You click CPC of site A, browse site A and bookmark a page of site A
    Then you return to google or another site for further browsing
    Within 30 minutes, you use the bookmark to go back to site A

    Source has changed and yet it hasn't… => 2 visits attributed to CPC or still 1 visit?

    I have several clients who have a widening gap in visits/CPC-clicks since August 12.

  41. 68
    Venky says:

    Hi Avinash,

    I have an interesting question to ask on Google Analytics – Capturing some key metrics on Page Views, Event Tracking….for games. Is there a way we can capture the amount of time spent on a game by a player A, player B, player C. Say the games are played on a hosted website.(To track the amount of time spent) Also what are the limits of event tracking in google analytics, what we cannot do- say for example to check a performance of player, player actions (his clicks), his button interface, etc.

    What are the interfaces between the player and the game…(record his actions, events, trends)?

    • 69

      Venky: Event Tracking (and Custom Variables) are two "open frameworks" in Google Analytics (and some other tools as well) to collect data. There is a simple structure to them and a hierarchy, but beyond that how you implement these frameworks and use the data is up to you.

      There is a ton of native flexibility that will allow to to measure what you want (time spent, accomplishments in games, paths taken, fake weapons purchased yada yada yada). But you have to think really hard as to what you want to collect and measure, because you literally can do anything. The best option is to work with a GACP (or the relevant consultant for a different tool) who can guide you optimally. Here's a list: http://www.google.com/analytics/partners.html

      One final thing… Sometimes even if you can get a rocket to commute to work, it is best to use a bike if you only have to go three miles. So as you evaluate your options don't close your mind to other specialized gaming only tools that might meet your needs.

      -Avinash.

      • 70
        Venky says:

        Thanks for your quick response Avinash. I will definitely look into the details you have mentioned. Your blog and videos are helpful. I am working on my GCAP (Individual Certification). The most important thing you did bring during the presentations is, data analysis is tough, tools can definitely perform but you are the one to identify, capture, omit and add value to data.

  42. 71
    Alex says:

    Very interesting, thanks for sharing – and very well presented.

    PS: although my mom did not call during watching your video, I did indeed do a few emails, and even had lunch :)

  43. 72
    Lovett says:

    Very informative post and awesome presentation.

    Thank you!

  44. 73

    Very interesting, thanks for sharing – and very well presented.

    PS: although my mom did not call during watching your video, I did indeed do a few emails, and even had lunch :)

Trackbacks

  1. […]
    Colpa (o merito) delle ferie mi ero perso una novità enorme nel mondo della Web Analytics, Google cambia il modo con cui le visite vengono “registrate” su Google Analytics.
    In realtà ne aveva già parlato Marco, e anche il blog ufficiale di GA, ma io ho incocciato solo questo post di Avinash.
    Fino prima di questo update (live se ho capito bene dal 12 agosto) una “visita” in Google Analytics veniva attivata ogni volta che
    […]

  2. […]
    Naturally Avinash Kaushik has written an extremely good post on Multi Channel Funnel reports (and more!) which should be a vital read for everyone. In particular, the video embedded in his article contains a very clear explanation of why the changes to the calculation of visits have been introduced as well as a great introduction to the benefits:
    http://www.kaushik.net/avinash/google-analytics-visits-metric-change-implications-opportunities/
    […]

  3. […]
    If you use Google Analytics, this post (and accompanying video) is a must-read. Avinash, an expert on web analytics, especially Google Analytics, explains how the session algorithm changed and the resulting impact this has on various reports and metrics in GA.
    […]

  4. […] Analytics: New Source Starts New SessionGoogle Analytics now starts a new session when any traffic source value for a user changes; http://eicker.at/GASession […]

  5. […]
    Google Analytics sessionization
    […]

  6. […] Video: Understanding recent changes to the Visits algorithm in Google Analytics […]

  7. […]
    . Da ich selber Sprecher war, will ich aber an dieser Stelle kein Recap machen, sondern allen Interessierten unsere Präsenation zur Verfügung stellen. Thema der Präsentation waren Multi-Channel Trichter bei Google Analytics. Seit Ende August gibt es bei Google Analytics in der neuen Version im Reiter Ziele eine neues ” Modul”, die Multichannel Trichter. Ingesamt ein sehr spannendes Thema. Für alle die sich auch für weitere Recaps interessieren – weiter unten mal eine Liste der wichtigsten Blogbeiträge.
    […]

  8. […]
    My favorite portion, aside from the new interface, is the change in the conversion attribution algorithm, which in turn also changes how GA defines a page visit. Before, a “visit” was defined as the user landing on the Web site. Regardless of how many times the user came back, one session was counted as one visit no matter how many referring sites there were and conversions were awarded to the last referring site before the purchase. The problem with this was that it was impossible to accurately track visits from mailing campaigns, organic traffic, etc. This has all changed.
    […]

  9. […]
    Voy a hablar de GA porque probablemente sea lo que la mayoría usa y conoce. Y voy a dar por sentado que obviamente conoces su interfaz y los conceptos básicos en los que se basa. Dime que sabes exactamente cómo se define una visita en GA (tienes grandes blogs de analítica web que te lo han explicado…y ¡ya estás tardando en aprenderlo!)
    […]

  10. […]
    Almost two years ago, Google made some fundamental changes to how they sessionize their web visits. One side effect of this change involved how they count visits from third party referring sites (including email, social media, etc.); understanding this change is critical not just for A/B testing email, but for generally understanding email analytics. And of course the example above illustrates some unexpected side effects of hash-based bucketing.
    […]

Add your Perspective

*