Conflicting data in Google Analytics

[Relative]

Hi Guys

I've been looking at the data for a client in Google analytics and I was wondering if anyone knows why some of the data doesn't tally up. In my case its the following:

Under Aquisition
No of Sessions for a Keyword shown in Organic Search tab (compared to)
No of Clicks for a Query in the Search Engine Optimisation tab

For example, for a brand term, Google are showing 17 Sessions in Organic Search.
For the same term Google are showing 90 Clicks in the Queries section of Search Engine Optimisation

OK, we know that Google are a little cloak and dagger regarding keyword data but surely Sessions and Clicks for the same keyword should be identical unless I'm missing something.

[katemorris]

There are few things that could be varying the data.

The first thing to remember is that Webmaster Tools is only showing data for the URLs in the index they associate with that account. This data is different if you are looking at domain.com or www.domain.com. If you had a few subdomains on the same root domain and on the same analytics tracker, but were looking at WMT for www.domain.com, you would be missing information from those subdomains.

In addition to that, there might be issues with your GA tracking code firing. It could be missing from some pages in the index. You could have a number of PDFs on your site that would incur clicks, but would not be tracked in GA.

The other issue is the difference in the data sets. WMT shows clicks only, not visits to your site. People can be clicking a result in Google and immediately changing their mind. That would result in a click but no visit. WMT also doesn't distinguish between returning and new clicks. It's just total.

The two data sets are completely off and each have their own drawbacks. To get the best data, you'll need to ensure there is a tracking code on every page. Then account for files like PDFs that can get clicks, but not visits in GA. That's a place to start. But it's still different data and not ever going to be the same.

These are some big differences and that does point to something being wrong, but there would need some in-depth analysis to point to exactly what is causing it.

[Relative]

Hi Kate and thanks for your reply.

I completely agree. Reporting on a keyword basis is a little old hat now and we should look at overall visits to the site along with other pre defined KPI's agreed with our client (nice article Kate). However; the question remains.

Google provide 2 sources of overall traffic volume (GA and WMT) and they seem to show completely different results.

(Copied from above)
Here's another one for 1st May - 25th May:
GA - Organic search - Total sessions  - 874
GWT - Search Analytics - Total Clicks  - 431 (New System, all filters removed)
GWT - Search Analytics - Total Clicks  - 1755 (Old System, all filters removed)
CRAZY RESULTS!

A lot of the data provided by Google is calculated on No of Visits. If the data showing No of visits is incorrect then the rest of the data becomes incorrect. Taking the results above and using GA as a benchmark, the two results shown in WMT are approx 50% more and 50% less than the bechmark. So the results you report could be 50% either side from the actual. Does this make sense? If WMT is more accurate then I would have to recalculate the results provided in GA.

Example (using data above)
if Total Sales = 10
(GA Results of 874) Conversion Rate  = 1.14%
(WMT New Results of 431) Conversion Rate  = 2.32%
(WMT Old Results of 1755) Conversion Rate  = 0.57%

The old saying of "You can't argue with maths" is only applicable if the data you're working with is correct in the first place.

[katemorris]

This isn't actually an issue of Google hiding or not reporting accurately. Bing and Yahoo also have started not processing any keyword data to analytics sets. Google still shows some, but it's only for non-logged in users.

The reports for WMT are more accurate, but you should not try to report on a keyword basis. It's a losing battle. I wrote an article on how to change up reporting and analysis in today's world. https://moz.com/blog/stop-thinking-keywords-think-topics

[Relative]

Thanks for your replies Martijn and Alick300.

I know they hide the keyword data but why show it in the first place. If Google want to give their users the best possible user experience then why show conflicting data and confuse their users?

The issue I have with this is that if this brand term in question is showing 17 sessions in GA but 90 clicks in GWT, which system is showing the correct overall Session Volume or does that metric simply not exist?

If I select a date range 1st May - 25th May and the same brand term, here's what I get:
GA - Brand term - 17 Sessions
GA Queries - Brand term - 90 Clicks (shown in GA from webmaster tools)
GWT Search Analytics - Brand term - 111 Clicks

Each result would have an affect on Total Sessions for that date range. This is only looking at one keyword and looking at other keywords it shows a similar discrepancy between the two systems.

Here's another one for 1st May - 25th May:
GA - Organic search - Total sessions  - 874
GWT - Search Analytics - Total Clicks  - 431 (New System, all filters removed)
GWT - Search Analytics - Total Clicks  - 1755 (Old System, all filters removed)
CRAZY RESULTS!

To me this means Google Analytics isn't reporting accurate visitor data which we use to make data driven decisions. When I'm asked by my client "How many visitors have we had this month", do I reply "Between 431 and 1755"?

[Alick300]

completely agree with Martijn Scheijbeler

[Martijn_Scheijbeler]

Well honestly you're missing something. The clicks and the sessions aren't identical between the two because they hide the keyword data in Google Analytics. That's why it will almost never match again in GA v.s. GWT.