Life Without Multi-Touch Attribution: How Your Marketing Measurement Program Will Change Under Incrementality

Marketers will first embrace incrementality for its cookieless nature but will come to find out that its benefits dramatically exceed even the best results of multi-touch attribution in their marketing measurement program.

As I wrote in a previous post, Marketing Measurement Cannot Be Reliant on Individual Customer Journeys: Incrementality is the Fix, upcoming limitations to third-party cookies are going to severely limit the way that marketers will be able to target consumers, personalize ads, measure effectiveness, optimize program performance, and a slew of other tactics that this industry has taken for granted in the past. 

This isn’t a “maybe this will happen” prediction. Google is committed to killing third-party cookies. They’ve already stated that it’s not just about how Chrome collects data. Their plan is to block any industry or vendor attempts to bypass their policies with other ID programs.

When it comes to a marketing measurement program specifically—the key ingredient to performance optimization—legacy models such as multi-touch attribution (MTA) will need to be replaced by cookieless solutions like incrementality. To learn more about this methodology, check out The Single, Most Important Difference Between Incrementality and Multi-Touch Attribution (MTA). 

Making such a big transition from multi-touch attribution—something most marketers know a lot about—and into incrementality—something most marketers know very little about—can be a painful shift. Although I laid out a change management plan to help those of your moving into incrementality, today, I’m going to turn the page on MTA and help you better understand what life will be like under incrementality. 

MTA = lagging data vs Incrementality = leading insights

Under multi-touch attribution, the way marketers plan and buy media really doesn’t change much. In a perfect world, MTA solutions track all of the brand touchpoints with individual users across channels in order to deliver a unified view of the customer. Going back to an example that I’ve shared multiple times, the following is a basic user path to conversion as they interact with multiple marketing exposures over a week before finally converting on the website:

Because each channel can only track its individual exposures, channel practitioners can only measure the contribution of their channel to conversion. So, in this marketing measurement program example, the Facebook practitioner would see in their platform that a Wednesday ad click created Saturday’s sale, a YouTube marketer would see that a Tuesday video exposure drove the sale, and an email marketer would think that their Monday blast was what caused the sale.

Multi-touch attribution was supposed to solve this by sitting on top of all channels, provide a unified view, and then apply smart math to determine how much of the $150 sale should be attributed to each exposure—i.e. the Facebook ad might receive $20 in credit, the YouTube ad $10, and the email $25, and so on. 

The glaring problem here, of course, is that it’s virtually impossible to track EVERY marketing exposure. What about offline media such as TV, radio, print, billboards, etc? Or other digital ads that aren’t trackable? Of course, if there were actually 10 exposures but the MTA system only tracks 5, then the math is absolutely flawed. And more importantly, what if an ad exposure had zero incremental value in driving a sale? MTA will try to give credit where no credit is due.

Champions of MTA argue that the insights are directional, but that’s just a nice way to say it’s flawed. And, with the third-party cookies going away, it will be even harder to provide a unified view and that flawed approach will be rendered virtually invalid (except in a few specific use cases).

Champions of MTA argue that the insights are directional, but that’s just a nice way to say it’s flawed.

On the other side, incrementality relies on a test and learn approach. Like the science experiments that we all learned in middle school, marketers use Test and Control groups to expose marketing campaign elements (channel, message, ad, etc.) to one group and not the other. By comparing the results, they can better understand the incremental value that the campaign element has on performance. 

The key benefit here is that because incrementality compares the Test and Control groups, all of the nuances and subtleties of individual exposure tracking that can be so easily missed by multi-touch attribution are negated and the causal impact of your media investments is revealed. Incrementality simply says, “Your business is X% bigger (or not) because the tested tactic(s) was present .”  It doesn’t attempt to explain everything going on with your programs, just gives a clear result of whatever you have decided to test. 

And so, here we have the main functional difference of MTA vs Incrementality approaches to marketing measurement: 

• Like a rearview mirror, multi-touch attribution is a passive, lagging dataset that marketers use to understand past campaigns in order to better plan subsequent programs
• Like a front windshield, incrementality is a premediated, leading dataset that marketers use to answer burning questions about program efficacy in order to improve current programs

Life under incrementality: a deeper connection between media and measurement 

A test and learn marketing measurement approach such as incrementality will require a more rigorous process than MTA. Practitioners need to plan out their tests, sort them by priority, execute those tests, and analyze the results. Then, they can make smarter, data-driven decisions on budget allocation, messaging creative, bids, and other choices that can help to improve the efficacy of their programs before they begin. 

Incrementality tests can illuminate key gaps in the way marketers understand their activity before the campaign runs. That way, marketers can start with an optimized campaign and then improve results even higher as it runs with in-flight optimization.

So, yes, an honest assessment here is that incrementality will take more time and effort than MTA.  

But, here’s the good part and why it’s worth it…

Because incrementality tests need to be planned out methodically, there’s a level of premeditated insight gathering that in essence forces marketers to have a deeper connection between their media and their measurement. You have to think more intentionally about what you want to measure and how best to run those tests in order to get the insights you need. 

If anything, MTA facilitated a bit of laziness on our part as marketers. With incrementality, marketers will have a closer link to the two sides—media & measurement—which most might agree is the ideal state of things. 

Plan your experiments, identify your most burning measurement questions, then test them 

What are some of the questions worth testing with incrementality? 

Here are some of the most popular, burning questions we’ve heard from marketers that they want incrementality tests to answer:

• Does this channel drive incremental value or would these conversions have happened anyway?
• What is the impact of one channel on another? If I spend more in one channel, does it help/assist another?
• What is the optimal level of investment for a specific channel? Are we spending too much or too little?
• What is the effect of upper-level tactics on lower-funnel activity?
• How does my offline media assist my online media performance? And vice-versa?
• Are my promotions and sales worth the discount I’m offering or am I just leaving money on the table?

These are questions that MTA just never was able to answer consistently. And, if it did, the known flaws of that approach can bring up questions of accuracy. Not having confidence in your measurement solution might be even worse than having no measurement solution at all. 

Worried that your marketing measurement program will be disrupted?

The worst thing you can do right now is to wait. 

The end of the third-party cookies is imminent. Skai’s Impact Navigator is an incrementality testing platform that can serve as the technical foundation for your cookieless measurement program. For all of the power of incrementality experiments and measurement, historically, execution has been incredibly resource-intensive. Impact Navigator is a service platform that enables marketers and businesses to measure the incremental impact of advertising tactics on their on and offline business metrics. 

To learn more about incrementality, and to see Impact Navigator for yourself, please contact us to schedule a brief demo.

Even if you’re not ready yet to switch, it’s a good idea to see what’s available so you can start thinking about how you’ll handle the coming changes to MTA and make the most informed decision down the road.

Book a Meeting of Skai