Although our article mentions Google Analytics, the ideas can apply to any service out there.

Google Analytics (GA) isn’t accurate. A quick Google search brings up endless articles, lists, videos, and opinion pieces on why we shouldn’t use GA, why we should use GA, why GA is evil, why GA is a productivity killer, and why GA is unhealthy. But mainly why GA isn’t accurate. Despite its age it’s still being used, and despite its inaccuracies people still hold it up as truth. It’s not.

Which might be fine in a vacuum. The problem comes up when you compare Google Analytics against something else.

Don’t get me wrong, Google Analytics is a great tool. It tracks a variety of stuff, it’s free, it’s simple, and you can pretty much find an answer to any question you have about it. But, we need to be realistic.

We’ve been asked countless times at AdButler why our statistics might not match Google Analytics, and it’s a good question. Both services work off scripts placed on the page, and run in the user’s browser. Neither are new technology. In theory, our impression numbers should reflect your counted page views. However, they can’t be one-to-one. And, inevitably, when the numbers don’t match, which side do you trust? There’s important considerations, like GA being free, AdButler and GA not really having the same purpose, and the difference between an impression and a pageview. However, these are almost philosophical differences.

I have a more basic explanation, which I like to call the train metaphor. If you bear with me, I promise it’ll all make sense.

You’re standing on a platform…

Your website is a train station. Let’s start there. Whenever someone visits the station it’s a big deal, so the station sends out two trains, one to AdButler and one to Google Analytics (both of which are also train stations). The two trains are carrying a simple message: someone has visited, so let’s count them! Once those trains hit their destinations the tally goes up by one, and in AdButler’s case, the train returns with some cargo, like a banner or VAST tag.

In a perfect world, both trains will hit their destinations every time and the numbers would be the same. But sometimes they don’t. Maybe one track is longer than the other (a bad internet connection). Or one train leaves before the other one does (AdButler usually loads before Google Analytics because it’s higher on the page). Or if the passenger sneaks by one train (using an adblocker), or by both trains (disabled JavaScript), so the trains don’t even know someone has arrived. There’s even the chance that one train hits a bad rail and flips off the tracks entirely (any sort of computer error)!

Usually, it all comes down to that: one train makes it, while the other doesn’t. If someone opens your page and navigates away before it finishes loading (especially if your page gets stuck loading for a minute), AdButler will record the impression, Google Analytics won’t. The same thing happens when AdButler records a click, but GA doesn’t record a visit on the destination page: one train didn’t make it, and that’s the truth of it.

So what do we do?

The first step is to turn on AdButler’s Accupixel Tracking. You can find it in the settings part of your AdButler dashboard. While still in Beta, Accupixel will make your statistics more accurate, and eventually, it will be turned on by default for everyone. Our enhanced tracking will lead to a smaller discrepancy between services. But that only solves half the issue.

Below are a few things that will keep things in perspective:

1. Even if the numbers matched, they wouldn’t be accurate. Trends are what is important, not hard numbers. If your numbers are trending up, that’s good. If you see a dip, that’s bad. But don’t sweat the digits.
2. More tracking services will not be more accurate. First of all, all tracking is going to be flawed (see number one). Second, putting more and more tracking scripts onto a page is going to slow it down, which is going to cause more tracking issues, and probably make people unhappy. Pick a few, hopefully ones that look at different things, and stick to them.
3. No service is perfect. The train metaphor is about Google Analytics, but really it works for any service. Maybe you run a waterfall (one train moving through various stations), or you’re getting discrepancies between two services (one train splitting into two trains), or maybe one service does it differently (now there’s a runner, or a boat, or a bike). Some services are definitely better, and Google Analytics is free for a reason. But those trains can always run into problems.

4. Keep it updated. A service like Google Analytics is always updating their technology. Stay up to date (maybe follow this blog?).

   Perhaps those aren’t really the tips you might have wanted, but you can find hard advice elsewhere. Just keep those tips in mind. Trains are a wonderful, sometimes accurate technology. Even so, they still need tracks.