Why Measuring Your Conversions by Source is a Mistake

By Renold Liu

There is a lot of talk about segmenting your sources and there are countless articles about how to properly set up your Google Analytics to track where things are coming from using standardized UTM codes, but this may not actually be the best way to measure your digital marketing results.

Put down your pitchforks and let me explain.

The most important thing to remember is that each marketing source will affect other sources and – while separating sources is important for different reasons – you must realize that the results of the marketing campaign are due to the whole effort rather than its discrete parts.

Not splitting up digital sources when measuring results may seem counter to what everyone is saying, but let’s walk through a very common buyer’s journey to illustrate how ALL sources affect each other and produce a cumulative effect.

Consider this example:

Let’s assume that you’ve done your research and you want a new BMW X5.

Because you’ve been researching BMW X5s on research sites, some dealer advertising technologies will know that you’re interested in X5s so you get served a Facebook ad of a beautiful BMW X5 at “Metropolis BMW” and a 0.99% lease offer.  

You think: “Not bad, let’s see what else is out there,” and you fire up Google and search “BMW X5” and  get served SEM BMW X5 ads from “Metropolis BMW”, “City BMW” and “Westgate BMW.” Because you recognized “Metropolis BMW” you click on that dealer’s ad. It takes you to the New BMW X5’s search results page, where you find the one you like.

You read up on it, scroll through the pictures, but for whatever reason you don’t fill out any forms and you leave.

A few days later, you are reading about Lebron dominating on ESPN.com and Metropolis BMW retargets you with the X5 you’ve looked at. You think “That was a pretty good X5, I should really show the wife and see what she thinks.”

Later that evening, you want to show your wife the X5. You fire up your browser and type in “www.metrobmw.com” and navigate to the BMW you remembered. Your wife thinks it’s pretty good and you fill out a “Get ePrice” form to get the best price.

SUCCESS! Digital marketing conversion!

To recap, this was the buyer journey, the ads, and the associated digital marketing sources:

AdDisplay Ad for BMW X5SEM Ad for BMW X5Retarget Ad for BMW X5Direct Website Visit

So, what would you attribute the conversion to?

Direct traffic? But there was a SEM ad click in there. What about display? There were no clicks – but it did prompt the buyer to chat with his wife. How about Facebook? It started the whole thought process.

Your Google Analytics would have reported Direct as the traffic source for this conversion. But, as you can clearly see, completely attributing it to Direct isn’t correct.

Ok smart guy, so then how would you measure your digital marketing?

Because each source interacts and affects other sources, the only trustworthy way is to measure “one level up” at the campaign or thematic level. For example, “one level-up” at your dealership could be the campaign you’re running: new cars, used cars, specials, fast sellers, slow sellers, etc.

This is because your new car campaign message will rarely cross over to your used car campaign messaging and vice versa. As a result, the people who resonate and react with one campaign typically won’t resonate with the other.

The measurement and subsequent comparison then becomes, which campaign or theme gave me the best ROI? Did promoting new cars on all channels result in better ROI than promoting used cars on all channels?

So next time you see yourself judging your marketing results by channels try to remember that digital marketing is all a big happy family and that digital marketing, as a whole, really is more than the sum of it’s parts.