The Next Generation of Visual Website Optimizer is launching on 29th April 2014 See What's Coming

Categories

Find Articles

I see a difference between my Google Analytics and VWO Data. Why?

VWO and GA (Google analytics) have different tracking methods. This is the reason why you see the difference between the GA and VWO data of your website. To make it easy for you to comprehend your data, here are the main points of difference explained that might be causing the confusion for you:

1. You’re Comparing “Visit Starts” in GA with Visitors Data in VWO

The “visit starts” data in GA is actually the total number of visits on the website. So if a person visits your website and closes it (GA loses the previous session here), only to visit it again after some time, GA will start a new session for the same visitor. As a result, the “visit starts” option that shows you the number of visits on a website will count this session again in its data and reflect 2 visits.

On the other hand, the visitors count shown in your VWO tool is the unique visitor count that became a part of the test. The default cookie life of a VWO cookie is 100 days, so this visitor is counted only once by our tool as long as the VWO cookie lasts in his browser.

If you want to see a fair data comparison, you should visit the “Overview” page in your GA to check the data listed under unique visitors. And this data should be compared with your Visitors data in VWO. Still, the exact data match wouldn’t be possible here as well, as GA and VWO have different tracking methods. Usually, there’s a discrepancy of 5-10% in this data.

2. GA Reports Take 24-48 Hours to Update, VWO Reports Provide Real Time Data

So if you have paused any test, make sure that you at least wait for up to 48 hours before you compare the data.

3. You’re Comparing the Date Selector Data in GA with the Date Selector Data in VWO

When you use date selector in GA, it shows you unique visitors only in that particular date range (even when they may have visited the site before the selected date). But date selector in VWO doesn’t segregate unique visitors for the selected date range.

For example, if a visitor visits a page (under test in VWO and is also being tracked in GA) on 4th May, he will be tracked as a unique visitor for 4th May in VWO and GA. If now the same visitor comes to the website on 5th May, for VWO he will not be a unique visitor and won’t be counted in our records for 5th May. On the other hand, for GA, if the date selected is 5th May, he will be again shown as a unique visitor.

So if the date selected in VWO and GA is 4th May, you will see one unique visitor in both of them, but if the date selected is 5th May in VWO and GA, you will see zero unique visitors in VWO whereas GA will show you one unique visitor for the same date. Now if you select a date range of 4th -5th May, you will see one unique visitor in both VWO and GA.

4.  You’re Comparing Results in Different Time Zones

Apart from being conscious about selecting the same date range, different time zones can also reflect a huge difference in the data that you see in your GA and VWO account. While your GA might be having the time zone that was chosen by you earlier, in VWO we have GMT as the default time zone. So before you start comparing any data or even begin with your tests, you must make sure that your date range and time zones are same in both the tools.

You can change your time zone in VWO any time you want. But you must keep in mind that when you change the time zone, your previous data will still remain according to the previous time zone and the new data that will be collected will be in accordance with the new time zone.

This can be a huge data mess to understand and can spoil the entire outcome of the test for you. This is why we highly recommend that you do not change your time zone in between the tests, and wait for all your tests to complete before you switch your time zone.

5. You Have Multiple Tests Active at the Same Time

When you have multiple tests active on the same webpage simultaneously, make sure that you choose different slots in GA for each of your tests. If you are pushing the data for multiple tests in the same slot, your VWO and GA data will not match.

6. You’re Doing a Split URL Test and You have Not Integrated VWO with GA

In the case of a split URL test, the control page is first loaded in the browser before the visitor is redirected to the variation page. At this time, before the redirection, many times GA code can begin to execute on the control page itself and this could trigger a pageview on the control page for that particular visitor.

This can often happen because VWO’s code is asynchronous in nature, so it may get activated at any time on the web page and not necessarily at the beginning of the page.  So in cases where GA code is activated before the VWO code in a split URL test, anyone going to the variation will also be counted in the control’s pageviews as well. As a result, you’ll see that the pageviews for your control page are almost the double of your variation in a split URL test.

To solve this issue, you can create segments in your GA to avoid the visits of the variation page from being counted in your control page data. You can do this by selecting that the page is equivalent to the variation page but not equivalent to the Control page.

7. Your Bounce Rate is Going Down in GA, But Not in VWO

As explained in the previous point, in the case of a split URL test, GA code may get activated on the page before the VWO code. So, it can pick the pageview of the visitor for the control page, before the visitor is redirected to the variation page. GA treats this redirection as an engagement because the visitor still stays on the same page even after being redirected to the variation. For it, this just triggers a visit to another page on the same domain.

This is why you might notice that in the case of a split URL test, your bounce rate has gone down after starting the test, while it remains almost the same in your VWO data.

We suggest that you check the bounce rate through custom variables in your GA to know the accurate data. You can use the non-interactive event in GA to make sure that Google doesn’t count this redirection as an engagement (or change your bounce rate) and only treat it as a pageview.

As soon as this event is put into action, GA starts collecting the data accordingly, however it may take 24-48 hours before this new data begin to reflect in your custom variables reports.

8. You’re Using the DoubleClick Code of GA for Tracking and Have Set the Prefix

As long as you have a prefix set when using the DoubleClick code of GA, your tracking won’t work. Make sure that no prefix is set, if you want tracking to work perfectly with the DoubleClick code.

9. You’re Using Async Code of VWO which Uses Low Timeout Values

In case you are using async code of VWO, it uses some timeout values by default. If you feel that your visits are less than expected, you can increase these timeout values to include users with slow internet connections. In the async code of VWO, you will observe two parameter settings_tolerance and library_tolerance with values equal to 2000 and 2500 respectively. These are timeout values of VWO code and are in milliseconds. When you increase these values, your visitors with slow internet connection will also become a part of the test.

In addition to these well defined reasons there can be some user specific reasons:

1. If a user has javascript disabled on his browser, he won’t be included in the test.

2. In case a user has cookies disabled on his browser, he won’t be included in the test.

3. The URL(s) on which you are measuring the traffic inside GA(or any other analytics tool) should be the same URL(s)(test page pattern) on which you are measuring the traffic inside VWO.

VN:F [1.9.22_1171]
Rating: 4.8/5 (8 votes cast)
I see a difference between my Google Analytics and VWO Data. Why?, 4.8 out of 5 based on 8 ratings
3 Comments
Neil
July 25, 2013

This was a very helpful assessment and insight into the differences we are seeing.

Giovanni
February 11, 2014

You declared “GA Reports Take 24-48 Hours to Update, VWO Reports Provide Real Time Data”.

But in GA I can click on refresh report, and the date of its generation is updated.

So, if I refresh the report, is your declaration still valid?

Smriti Chawla
February 11, 2014

@Giovanni – Refresh report in GA does not necessarily mean that it will show latest data. It is just a refresh from browser cache. For more information, you can check out this link: https://support.google.com/analytics/answer/2450241?hl=en

Leave a comment →


Raise Support Ticket

Top

Raise a Support Ticket!

Your Name

Your Email

Your Phone

Your Query

Or, contact us through email support@wingify.com

A B Testing Tool