What everybody needs to know about tracking

Posted by Dave CollinsAnalytics, Google Ads

The era of being able to accurately track your online marketing efforts is over. Or more accurately: the era of thinking you could accurately track your online marketing is over.

Only a few years ago, we thought we could track almost every single keyword that was used to find our websites. The reality was that we could track the keywords that were used to click on our websites in the search engine results. If we’d have left it at that, the difference would have been minor, but we didn’t.

Once we could track our keywords, we then wanted to link them to conversions. We wanted to know which keywords resulted in the greatest number of enquiries, signups and sales, and why not? The information was there. Except it wasn’t.

A few years ago we worked with a company who were spending a great deal of money on their AdWords ads. Initially they were confident that the ads were sending targeted visitors to their website, but with time they became accustomed to the smörgåsbord of data that was there for the taking, and started trying to track their ROI more accurately.Precious

After months of implementing an impressive series of steps to track their conversions to previously unheard of levels, they saw their worst fears confirmed. Their AdWords account wasn’t anywhere near as profitable as they’d thought. In order to confirm this they decided to freeze their AdWords account for a seven day experiment to see what impact this had on their sales.

If I remember correctly it was just over 24 hours later that they turned their AdWords ads back on, as the immediate and dramatically negative impact on their sales was all the proof that they needed. Their brief but costly experiment had taught them that conversion data is not to be relied on.

This particular company was fortunate in that they not only had high volumes of fast-converting data to reach this conclusion quickly, but also had the foresight to test their theory before fully committing to it.

I won’t bore you with the variety of different reasons that there may be disconnects in your conversion tracking, but I will ask that you at least consider the idea that your tracking may be dangerously misleading.

Let’s take a step back. If I were to build bird houses and sell them in local markets and craft fairs, I would be able to precisely track my ROI, based on the costs of my time and raw materials, the costs of my selling at each event and the sales from each.

The same rules, however, do not apply online – at least not for most businesses. And this goes far beyond the issue of which attribution model you choose to work with. You simply cannot track all your sources, and more importantly you have no idea how representative this data may be.

When a person fills out one of the forms on the SoftwarePromotions website, we can see where that person came from as a referral source.

In days gone by we used to also ask people where they heard about us, and the disconnect between what people said and where they actually came from was enormous.

So even when you can track, you can’t necessarily rely on the data produced.

We also recently started working with a company that I first met at a conference many years ago. When they filled out the form on our website I could see that they “found us” through Google. This was in fact true, but also quite incorrect.

I can only imagine how many businesses close their AdWords accounts in error because of incorrectly calculating a poor ROI.

I assume that this is significantly less than the number of businesses who spend far too much on their accounts but never realise.

The house, in this case Google, always wins. And they don’t even need to rely on statistical probability – they can weight the system as they see fit.

Conversion tracking is an extremely useful indicator. Using it as anything more can be deadly.