AdWords and Omniture metrics don't agree

[mcculloughphoto]

I am  comparing metrics from an AdWords campaign and Omniture metrics. I find two dates that AdWords says we received more visits from the campaign than we received in total traffic to the site on those dates. After accounting for base traffic to the site AdWords seems to be around 30-40% high.

Are there any known issues with Omniture not counting visits?

Has anyone else experienced this problem and what was the cause?

-Dave

[AkhilSharma]

An excellent question: why they dont agree.
To understand this we need to understand how both of them works and then we will understand why is it so.

Adwords: captures the visit the moment it is  clicked from the search engine/content network. So for example if an user clicks on the Adwords ad, immediately the visit is incremented in adwords database.

Sitecatalyst(Reporting and Analytics): will register the visit the moment visitor had landed onto your website and the tag is fired.

Now theoretically the journey between the ad clicked and landing onto our website is very short(sometimes in milliseconds upto a few seconds) depending upon the user internet speed.

However, technically it is a long journey which passes through  several technical challeneges. For example what happens if:

1)due to low speed of the network, the user closed the browser without even seeing your landing page.
2)has javascripts disabled or blocked by his browser.
3)he had landed on the website but immediately closed the browser even  before the sitecatalyst tag was fired.
4) user had reached your website landing page but due to technical (internet disconnection/slow speed) issue, the page was not loaded/half loaded which resulted in the user abandoning the website in itself.

Ideally there could be many possibilities which resulted in the difference, but the key lies in understanding why there is a difference and how to minimize the difference.
For example: when there is a huge difference: i will drill down deeper to identify this segment which is giving the highest difference may be a city/country specific, device specific, internet speed specific.

Then identifying how to reach out to them. May be by making landing page faster load and lesss heavy page etc.

Hope this helps. Please mark it solved if it answers your query.

[mcculloughphoto]

It is helpful in directing me where to look but I doesn't resolve the problem.
I get the concepts involved but I'm having trouble with the size of the discrepancies. Here are two specifics.

On 1 Sept Adwords said that 133 mobile users and 0 computer users clicked on an ad that took them to a specific page. When I check the Omniture metrics for that date/page I see that page had 103 visits for the day. We average about 15 visits/day to that page and according to Omniture 89 visits originated on that page. That is a difference of 44 visits (133 - 89 = 44) or almost 1/3 of the clicks that AdWords billed for did not record in Omniture. I've checked Omnitures Javascrip report and of the 693 visits to the site on that date, 500 users had java script disabled and 193 had java script enabled. I also checked cookies - 685 enabled, 4 disabled and 4 unknown.

On the same date Omniture reported 693 total visits to the site. We average about 300-350 visits/day so we were up about 350-400 for the day. Adwords charged for a total of 755 clicks from ads to the site on that date.

As to the possible causes that you mentioned here is what I know -
1) Slow speed of the network - this is not an issue. This is a site for a major manufacturer with dedicated servers and a top notch IT department.
2) Javascripts disabled or blocked by browser - I understand this is a problem with Google Analytics but doesn't seem to be a problem for Omniture. According to Omniture java script was disabled on many of the devices used by visitors but it did record this.
3) Closed browser visit recorded - this is a possibility but that could be the case for a couple of visits, not for 30% to 50% of total visits from AdWords.
4) Technical difficulties resulting in user leaving - This is a fast, smooth running site. This isn't an issue.

Additionally, we are running an ad campaign that uses a YouTube video. The video is not running on our site so the only views are from the ad or organic - which are next to nothing because this has only been online for about 2 weeks. For last month YouTube reports 2,802 views of the video. For the same month AdWords billed for 3,233 views of the video. That is over a 13% discrepancy (431 views) within their own system.

I get the concept that there can be a difference but when I have a 30% to 50% discrepancy I've got to understand what is going on.

Do you know if there are any problems with Omniture not recording visitors under certain circumstances? For instance do they record ALL visitors that have javascript disabled or do they miss some of them?

I can't be the only person running into this problem.

[mcculloughphoto]

I just got through reading up on Omniture's  javascript report. After adjusting for mobile phone users there were only 4 users (out of 693 for the day) without javascript enabled.

[AkhilSharma · Accepted solution]

Hi,

1) Slow speed of the network - this is not an issue. This is a site for a major manufacturer with dedicated servers and a top notch IT department.
Answer: Speed of the network depends on the two factors one from the server side and other from the user side.As you indicated that server side speed is top notch then it means what might be lacking is the speed from the user side. As most of the missing clicks are from mobile around 44,  it may be that these 44 users who had clicked didn't reached to the website and dropped in between.

2) Javascripts disabled or blocked by browser - I understand this is a problem with Google Analytics but doesn't seem to be a problem for Omniture. According to Omniture java script was disabled on many of the devices used by visitors but it did record this.
Answer: this is not a problem with Omniture(Sitecatalyst-Reporting and Analytics) so lets leave this one out.And as the numbers also show that most of the visitors can be tracked with or without javascript enabled.But at least we can now say that whoever reaches the landing page gets tracked by omniture(Sitecatalyst-Reporting and Analytics).

3) Technical difficulties resulting in user leaving - This is a fast, smooth running site. This isn't an issue.

Answer: Again it might be from the user side.Please check the size of the landing page as well as the speed and time it takes when browsing through mobile. 
As for tracing these missing visitor might be very difficult as for sitecatalyst they didn't reached the landing page or not stayed till the javascripts tag/beacon is fired.

As for the possible workaround and identification, try the technology report and specifically mobile reports which will help you to identify the problem. 
Also please note that its quite difficult to identify those who had not reached the sitecatalyst tracking but for 30-50% of the discrepancy its worth trying.

Let me know if that helps.

[AkhilSharma]

Hi David,

Since we haven’t heard back from you recently,hope we were able to answer all your queries and the issue got resolved to your satisfaction. 
I will go ahead and archive this case as resolved.
However, if you have any additional questions simply respond to this message and it will be reopened.
I would appreciate if you could spare your valuable time to participate and identify the correct answer which would be of great help to us.

Thanks and Regards

Akhil 

[Locutus243]

David,

Did you ever manage to get to the bottom of this?

We are having some similar issues whereby AdWords is showing 25% more clicks than Adobe is showing visits. I've done some testing and the journey seems to be working OK. While I get all of the points by AkhilSharma I find it hard to believe so many clicks would not load the landing page because of technical issues.

I wondered if you managed to solve it and if so how you went about doing it?

Mark