9 replies to this topic

[copywriter]

I've wondered this for a long time. My GA traffic has always been WAY higher than other tracking tools show.

For instance, for one particular site, GA consistently says I have 6,500 uniques a month. But another program I have that tracks visits shows around 400 uniques a month.

[Jill]

It's usually the other way around. I wrote about it in the last newsletter:

Does Google Analytics Count Bots as Visitors?

Perhaps those other programs are just missing lots of traffic for some reason?

[copywriter]

Is that something Google said? (No bots?) Where did you find that info?

[Jill]

I don't recall. They likely say it somewhere, but it's always been my impression that they filter out bots. I imagine some can get through, if they're intelligent enough, however.

[Jill]

In researching it more, the consensus is that bots typically don't execute javascript and therefore don't get counted. But it's likely there are some sophisticated bots that do execute javascript.

[copywriter]

But I wouldn't think there would be THAT much of a difference, ya know? I mean 6500 vs. 400? Curious.

[JKMitchell]

Is it possible that GA counts uniques differently to your other stats package? In the past I've looked at several and the "rules" for what constitutes a unique visitor seems to be different for all of them.

Having said that, that's one heck of a difference - are the total number of visits different by a similar factor?

[copywriter]

It's not a stats program. One time I noticed it was with Google's own Website Optimizer. GA would show thousands of visitors, but WO would show hundreds.

Then, I installed a live chat program. It also tracks visits with total and unique. It's doing the same thing. GA shows 1,000s and the chat shows 100s.

[Michael Martinez]

Unfortunately, you cannot assume that Google Analytics is tracking visitors the same way your chat program is. All sorts of factors could be affecting the counts, including (but not limited to):

• Placement of the Javascript code on the page (hits tend to be lower the farther down the document the code is placed)
• The algorithm on the back end (written for two different purposes, by two different teams)
• The availability of either remote service (even Google Analytics may be unreachable at times)
• The criteria used to determine a "visit" (such as date ranges, use/non-use of IP addresses, cookies, etc.)
• Minimum requirements for activity to quality as a visit/view/hit

It's better to compare analytics to analytics but even there you'll run into disparities for the above and other reasons. For the most part, if you see the same spikes and dips around the same times on the different services you're probably tracking visits consistently. However, if the two (or more) services don't show the same traffic patterns then one (or more) may be misconfigured.

[copywriter]

Thanks Michael.