Commentary

Measurement: Data In & Data Out

The data-driven marketing industry is extremely focused on 50% of the story — the “data in” component that drives targeting and personalization — but what about the “data out” element?  Are you spending the same amount of time on measurement as you are on targeting?

It’s easy to overlook this element because it’s the one that’s been around the longest.  Some marketers aren’t fully aware of the strides that’ve been made in measurement during this current age of big data.  Most marketers focus on creating interactions or engagements, then applying attribution models and media mix models to determine performance, relying on delayed revenue reports to measure success.   While these are certainly ways to determine success on longer running “always-on” marketing efforts, there are lots of in-flight ways to measure results as well, which can used for shorter in-market bursts.  

Measuring performance in-flight relies on two distinct types of data: causal or non-causal.  Non-causal is whether or not your audience made a categorical purchase during the flight of your campaign.  This is a great way to determine if, overall, your efforts were successful at driving a purchase.  

Causal is significantly more difficult to achieve, as it relies on a direct data feed of purchases from a partner who has access to this information and the ability to tie this back to a unique identifier.  There are companies which offer these solutions, but there are certainly more hurdles to overcome on this kind of reporting.

In-flight measurement is key to the optimization of your efforts and requires you use the right data to optimize toward.  If you get access to the data in close-to-real time, your marketing program can be automated, and performance can increase substantially in a very short window.  

Marketers have been optimizing digital campaigns for 22 years, but in most cases that optimization is based on a behavior other than actual purchases (e.g., engagement, interaction rate, or the dreaded click rate).  More options for gaining access to that purchase data have become available in the last five to six years.  So marketers are now able to optimize the entirety of their efforts — from website imagery, email to of course advertising — against purchases.

Interestingly, click-through optimization is finally dead.  I don’t know very many marketers who optimize to click through anymore, which enables the entire industry to breathe a sigh of relief.  The click-through rate was the industry’s first and most substantial mistake, even though at the time it was seen as a breakthrough.  It was a proxy for action, but not the intended goal.  Of course, when we came up with the CTR measurement, click rates were well above 10%!  I remember clearly when customers were concerned because their CTR dropped below 5% on one of their banners – it was like the sky was falling!

Closing the loop — driving the data out the back-end of the equation to provide measurement, and feeding that back into the system as “data in” — is where marketers have always wanted to be. Now they can finally get there.

Next story loading loading..