The Importance of Multi-Touch Attribution: Why Getting in the Weeds Improves Decision-Making
Here’s how you can give credit where it’s due and make the right marketing investments for your brand.
by Eric Matisoff
posted on 08-06-2018
Do you remember what exactly convinced you to catch “Black Panther” at the movies? Was it a text from a friend? A trailer you watched on YouTube? An interview with Chadwick Boseman that popped up on your Facebook feed? A coupon from your local cinema?
Consumers are fed a dizzying amount of information through a variety of channels — Facebook, email newsletters, Instagram, display ads, stores, mailers, and more — so it’s understandable why companies have a hard time understanding exactly what led them to make a particular buying decision.
But here’s the rub: as a marketing organization, it’s critical that you figure it out. Tried-and-true attribution models, such as first-touch and last-touch attribution, aren’t the only way to understand a customer’s path to conversion. Different situations warrant different models, so your analyst teams need the flexibility to offer views into attribution that are customized to the needs of various groups within your organization who are trying to improve the customer experience. This is why multitouch attribution increasingly has become a necessity for better measurement and campaign optimization.
The case for multitouch attribution
Consumers, of course, do more than just grab tickets to the movies. They buy your products and services. And while a simple text from a friend might be enough for them to throw down dollars for dinner and a movie, branding matters in influencing which movie and restaurants they pick. What’s more, the process for picking a movie to watch is very different from buying a new car. The roadmap to the final “yes” in the customer journey can be convoluted.
In an ideal world, it would be nice to throw advertising dollars at every channel possible, but, in reality, marketers have a limited marketing budget and need to maximize ROI on those dollars. The solution: figure out the touchpoints, or attribution, so you’re not throwing away good money by spending precious dollars on channels that are doing little other than adding to the noise-to-signal ratio.
And since consumers access marketing information through a variety of devices (67 percent of shoppers regularly use more than one channel to shop), and take in a large variety of content along the way, multitouch attribution is essential so you can figure out exactly which channel and which device are best for each marketing campaign.
Not yet convinced? Surveys indicate that many enterprises are increasingly becoming convinced of the value of multitouch attribution. Fifty-six percent of marketers believe it’s important, while 33 percent say it’s paramount. Attribution also makes companies 15 to 30 percent more efficient, allowing them to deliver better and more seamless experiences to customers.
How to do attribution right
As a marketer, you want to understand if you’re making the right marketing investments and be empowered to answer questions like, “Should I be moving dollars from social to invest more in our mobile app?” Smart enterprises follow a data-driven marketing approach. This is where multitouch attribution will help.
There’s a wrinkle, however. Attribution is a very tricky landscape to maneuver and there are significant ramifications if you get it right — or wrong. With so much at stake, what are the best strategies to score one better with multitouch attribution?
First, do right by data. What’s the point of all that analysis if your data is flaky? Make sure you have all the data you need from every channel and touchpoint. If your company hasn’t integrated search data into your tracking and measurement systems, doesn’t have quality display impression data, or half your website isn’t yet tagged properly, then the data you have likely isn’t as robust as it should be. Accurate, scalable, and actionable data is the foundation of a valuable, data-driven attribution ecosystem.
Second, be prepared to act. Analysis in and of itself is not the end goal. Action, based on what the analysis is telling you, is. This is where it’s also important to have C-suite buy-in, so all stakeholders understand and are fully invested in the final outcome.
I’ve seen customers who are very successful, and they all have one thing in common: they all have executive management that’s very data driven, supportive of their analyst teams, and gives them a lot of liberty to make the right decision. With attribution — which allows you to clearly see what’s working and what’s not — there are winners and losers within your organization. So, unless you have the support of your executive team, it’s very difficult to have an effective attribution program.
Pick an attribution model that will work for you. There are many ways of analyzing consumer data, and each model has its advantages and disadvantages. You might already be familiar with first-touch and last-touch attribution, where all the credit is given to the first or last point of contact with a brand, respectively. All the other channels along the way get none of the credit in these models. Understandably, such methods of analyses don’t paint a complete picture, but they’ve been popular because they’re easy to set up and use.
More complex models such as linear, time decay, or U-shaped attribution models assign varying degrees of credit to touchpoints along the way. In a linear attribution model, each touchpoint along the path to conversion (for example, social, and paid and search) is given equal credit for the conversion. With time decay, touchpoints closest to the conversation receive the most credit. In a U-shaped model, the first and last touchpoints that helped create the lead are prioritized, and the remaining credit is divided equally among all other touchpoints along the way. Depending on your organization, one of these models may be a better way of answering the question: How did my customer get to yes?
Take one of these for a test drive. The multitouch attribution model that will be right for you is not always static. It goes without saying that what worked yesterday might not work today if your data and goals have changed. Before investing resources into launching an analysis from end to end, try a test data sample to see if you’re able to garner any more insights than with the models you already have in place.
Look beyond marketing channels and orders. Multitouch attribution needs to be available at more granular levels of data than social vs. display vs. search, etc. In my analysis experience, consumers often click through on multiple paid search terms and campaigns before making a purchase. Or perhaps it takes five internal searches to find the right product — which term should get credit? Applying different attribution models to keywords over time can provide interesting insight into the mind of your consumers, and allow you to optimize so much more than just your high-level marketing budget. Finally, remember to adjust your attribution model for micro-conversions like registrations, video completions, and product views. Certain channels may be better at driving registrations rather than orders, but that is easily forgotten when only focusing at the bottom of the funnel.
Customers engage with brands today on more devices and through more channels than ever before, so it’s difficult to figure out exactly what led to the final sale of a product or service. But a multitouch attribution model can help you make better data-driven decisions, and transform the insights you gather into action that leads to more marketing efficiency and less waste of your valuable marketing dollars.
This article originally published on InfoWorld.
Topics: Digital Transformation
Products: Analytics, Audience Manager, Campaign, Experience Cloud, Experience Manager