Site Reporting : Differences between BC and Google Analytics

[Guest]

Hi, my client has raised an issue they are seeing on their website reports via Adobe BC vs. reporting information from Google Analytics. 

• They are trying to do a comparison to view traffic / activity on their website. By using both reporting from admin console and comparing to Google Analytics.  These figures are completely different. 
  For example -  a simple comparison would be SITE VISITS

• We have used the SAME time period on this comparison : i.e. : 31May - 16 July to view traffic
  Google reports 2,627 visits
  Adobe BC reports 5,144 visits

• In this time on the website, Adobe report a spike on 20 June, which shows an unusually high number of visits with 1084 visits. The same event is not recorded on Google Analytics on this day it shows only 20 visits to the site.

• Another comparison to show is UNIQUE VISITORS
  This reports from Adobe BC 4248 visitors 31May - 16 July to view traffic
  Google reports 1909 visitors 31May - 16 July to view traffic

Any reason why these are so different?  We are not expecting them to be spot on, but thousands of visits or sessions difference is major.   Any ideas?

Thanks in advance, Megan

[smitchell80 · Accepted solution]

Hi Megan,

Our analytics work fundamentally differently like other services (Google, etc) so there can be discrepancies in the data between them.

For example, Google Analytics counts a page visit as a 30 minute session. If the person is inactive for 31 minutes and then clicks again, it counts as 2 visits.

As for BC individual visits are given a 1 hour session before they time out and are counted as a new visitor. Unique visitors are kept unique via a cookie for up to 1 year.

Every time a user interacts with the site this is refreshed (so if you view a page once and then 30 minutes later, the timeout would not occur until 1hr 30 min after you viewed the page the first time).

Google relies on JavaScript to record usage whilst BC does it each time it serves content. We use a number of methodologies to detect bots including user-agent string, repetitive hits from same IP or location amongst others. There's the reason why the numbers are different and cannot be compared. You can compare the results provided either by Business Catalyst for a long period or by Google Analytics, but not one method with the other.

Whilst there may be differences in the exact number of reported visits (as Google Analytics and BC have different ‘count’ processes and Bot handling mechanism’s) the main use for visitor statistics should be to view trends over time, which you should find are similar between GA and BC reporting.

Hope this helps!

-Sidney

[be.free]

I had the same question. Very, very helpful to know. Thanks for a comprehensive response Sidney.

[FraserMcCulloch]

Sidney

Why don't Adobe just use the same reporting method as Google, the industry standard, and save every single BC partner and business owner the confusion of 2 reporting tools or thinking your thing is bigger than it really is ?

I've just looked at a client site for the time frame 20 June to 16 July

BC versus Google

Total Visits 3934 v 1914

Unique 3076 v 1370

From your explanation above, surely BC would deliver lower visitors than Google ? Every single one of my client sites have GA installed and in every instance monthly visits are more than double in BC compared to GA.

F.

[smitchell80]

@Fraser,

Those numbers appear a bit off please log a ticket so we can investigate your numbers further. 

On a side note we are currently working on improving this a bit more to closely align with GA.  Look for this option to become available in a very near future release.

Kind regards,

-Sidney

[FraserMcCulloch]

Hi Sydney

Thanks for the suggestion but I would have to do this for all my sites !

We have a saying in Scotland, "we don't complain, we just go elsewhere". I'm a BC fan but I stop using BC analytics many years ago.

Here's another one with attached screenshots from BC an GA

BC v GA

Total Visits

3411 v 2131

Unique Visits

2710 v 1670

[Daryl Barnes]

Sidney,  I really hope you guys do make the BC reporting more in line with the industry standards.  Our numbers show similar to what others are reporting where the visits and uniques are more than double what Google Analytics reports.  This serves no one well.

[be.free]

I actually prefer the way BC does it.

The fact that GA relies on JS will immediately illuminate many users who either don't have a correct version or for security reasons don't have it turned on - that is one valid reason GA will miss visits.

The argument Sidney makes for the reason BC applies the logic they do make more sense than GA and for that reason I discourage the use of GA which is mostly unnecessary. There are many other tools out there that apply different methodologies anyway.

BC please DON'T become another Google clone!

[Guest]

Back to my orginal post - perhaps this comes back to what the client wants? My client who raised this is **really super** interested in stats, all the action on their website! They have a business that runs around the globe.  I do have other clients who go "What do you mean analytics...? "  Either way, it seems that the client should choose 1 tool to monitor and stick with it - so at least there is a benchmark/starting point to show what activity happens on a site.  Google Analytics is such a bigger beast with so many other amazing features that are not available on the BC Platform, so perhaps again opt to show the client both and see what rattles their interest for their own need?  For my say I think it comes back to the client you're helping, how wide their business stretches, what their site does for users etc. By defualt I plug it into all sites. Nothing worse than getting down the track with a client who suddenly takes interest in their site statistics and you can't help them with information deep enough.

[Daryl Barnes]

Wayne, no one has a problem with the way BC actually collects the statistics.  The problem is in how they calculate what is a real visit by a human versus what is just a spider or bot.  And what constitutes a "unique" visitor versus a return visitor.

What method it uses to collect the data is a non-issue.  Studies show that users without javascript or with it turned off typically represent less than 2% of visitors to the average website so it does not account for the over double the amount of traffic that the BC reports show.

However, now that I think about it, the reason Google Analytics is more accurate as to real human visitors is likely because it uses javascript.  The overwhelming majority of spiders and bots that crawl the web don't process javascript so that is likely why the raw server logs that BC uses for reporting are so off in regards to reporting real human visitors.

Clients don't care how many spiders are accessing their site, you know.  Spiders and bots don't purchase our customers goods. 🙂

So BC needs to be able to filter out the spiders and bots better in its reporting since it doesn't self filter using javascript.  Not sure how realistic it would be to be able to do that though.

[cristinel@adobe]

Hello everyone,

I would like to jump in this discussion as I think it is just the right timing to have it since we are in the middle of an analytics infrastructure update and we're starting to gradually change the reporting to correct some of it's current flows.

First, I'd like to make some additions to Sidney's detailed description of how a visit is calculated in BC compared to GA, before moving on to discuss about the reasons behind those differences.

Business Catalyst - (until 4 months ago)

• A visit was reset after 1 hour (regardless of whether the user continued or not the session)

Business Catalyst today

• A visit is reset after 30 minutes of inactivity or after 12 hours (regardless of whether the user will continue to use the system)

Google Analytics

• A visit is reset after 30 minutes or at 12:00 AM
• Also, a visit is reset when the referal has changed

At first sight, these differences in the way these two analytics solutions are calculating visits are not big enough to justify the differences between GA and BC you've highlighted in your examples. So where do these differences come from, if not from visit definition?

The main reason and the bulk of the difference comes from the fact that BC Analytics is integrated solution and it aggregates the visits that come into all site domains/URLs. A BC site, with a single custom domain (my site.com) has the following URLs:

• mydomain.com
• (optional): www.mydomain.com
• mysite.businesscatalyst.com
• mysite.worldsecuresystems.com

By comparison, when setting up a GA account, most of the times a user will set it up for a single domains from the ones above, while some traffic comes through the others. GA will not count those visits & visitors, while BC will do it. To avoid this situation, you should set-up redirects towards a single domain or set-up GA to count the visits in all domains under the same profile.

One other reason, probably with a smaller impact, is some content might not be tracked by GA while it is tracked by BC. One example here are the email campaigns that could be viewed in browser. Since the campaign likely does not uses the site page template and does not include the GA code, it is not counted there.

Because of all these reasons together, the numbers will likely never look the same. The differences are so big mainly because of the set-up and not because we count bots and spiders into our reports.

Getting back to what I was saying at the start of the message, with the release we're going to launch next week, we have incorporated a beta version of a new analytics engine. While the admin console does not includes any clear link or interface on how to access this (so that we don't confuse customers) the tool will be available for partners to try and I will publish a separate post on this forum with instructions on how to enable the new engine. Until now, we've run some internal tests with the tool, and the results now shows comparable but not identical results between GA and BC.

Starting with the next release, you will be able to play around with the new engine and give us feedback on the results. Stay tuned for the announcement,

Cristinel

[Neil Eisenberg]

Thanks for the info.

Neil

Neil Eisenberg

Design Intervention Studio

"The Art of the Web"

19 Birch Lane

Woodstock, NY 12498

845-679-8052

www.designinterventionstudio.com

[Daryl Barnes]

Cristinel, I really appreciate your explanation for the team's proposed fixes.  Can you follow up with instructions on how to enable the new analytics engine?  Or has it already been posted somewhere?

[pChirita]

Hi Daryl,

You can go to http://YourSiteDomain/AdminConsole/#!/Utilities/ExternalReporting.aspx

and check the "Enable Quick Reports" box.

Please note that the engine is still in ALPHA and it will be down every once in a while 😉

Cheers,

Paul.

[Daryl Barnes]

Are there any plans to make the new reporting changes live?  Or is it already live?  If not, wouldn't it make sense to incorporate this into the new "Beta Features" section of the Site Settings?

[Bogdan_Ripa]

We plan to add it to the beta features list with the December 4th release.