The Paradox of Multiple Attribution Models

January 29, 2016 Mark Emmons


Marketers are pretty good amateur jugglers. Only they’re not keeping balls, clubs or rings flying through the air. Their specialty is juggling complex data.

New technologies allow them to slice and dice data more precisely in hopes of understanding the real impact of their business. Take, for instance, the question of attribution. Now, marketing teams can search for clues using multiple reporting models. First-touch attribution. Last touch. Conversion touch. Multi-touch. W-shaped. U-shaped. Time Decay.

That’s why data-driven marketers can find themselves juggling several different sets of often conflicting results. Sure, it’s a neat trick catching and tossing all of those balls, uh, attribution models. But more information actually might be giving you less insight into your marketing efforts. It leads to confusion.

“That’s the paradox of multiple attribution,” said Adam New-Waterson, the chief marketing officer at LeanData. “You’re convincing yourself that there are multiple balls in the air when the reality is that there’s only one ball. The problem marketers face isn’t a lack of models. It’s a lack of understanding what their data really means.”

What they need to do is keep their eye on the right ball, he added. But figuring out which numbers are relevant in your reporting and then applying them in a meaningful way for your business is a challenge.

“You can make numbers say just about whatever you want,” added Jason Seeba, the chief marketing technologist at BloomReach. “You can tell whatever story you want with data. But if you’re looking at too many metrics, or the wrong ones, you’re going to end up making the wrong decisions.”

Or you may become tempted to make no decision at all. There can come a point when the sheer amount of information becomes overwhelming. You can’t process all of it in any meaningful way. Analysis paralysis can take hold.

different lenses

Through different lenses

“That’s why the main point to me is really understanding what the different attribution models mean and finding what works best for your business,” Seeba added.

If that juggling metaphor is too abstract, here’s a real-world example. Your company closed a $120,000 deal. There were six marketing touches. In a first-touch model, the entire $120,000 of credit goes to the initial person in an account who showed up on your radar. The same thing happens in the last-touch model. In an unweighted, multi-touch model, each of the six touches is awarded an equal $20,000 share of influence.

“You’re trying to look at the deal through all of these different lenses,” New-Waterson explained. “You’re using the same $120,000 to rationalize several different arguments. But the problem is you only have one, $120,000 figure. You just can’t have several sets of computations based off of one deal and still have it make sense for other people in your business to understand.”

Savvy marketers understand the limitations of first-touch, last-touch, and unweighted models — how they don’t give a complete picture of the buyer’s journey. However they often feel they need to “dumb it down” so the information can be understood by a wider audience. That includes busy executive team members who expect concise, digestible explanations and not deep-in-the-weeds analysis. Even if you’re drilling down with a sophisticated form of weighted multi-touch attribution, the visibility still can be cloudy.

That’s because data is always open to interpretation.

Let’s say, BloomReach’s Seeba said, that a successful deal has included somebody attending a webinar, prospects visiting a trade show booth and an executive dinner. All of those touches played a role. But how do you determine which one — or ones — were the most pivotal to pushing the deal through?

In the Weeds

In the weeds

It’s complicated.

“One of the interesting things about attribution is that everyone likes to try and take 100-percent credit for deals,” Seeba added. “Somebody on your team might feel that the trade show booth was core to closing the deal. But they don’t take into account the other 50 marketing touches that happened. It’s not necessarily being done selfishly. People are saying, ‘This deal wouldn’t have happened except for X.’ But the reality is there are a whole bunch of ‘except for X’ factors in every deal.”

It’s not about getting “credit,” either. Yes, marketing faces increasing pressure to defend its budget and prove Return on Investment (ROI). But without understanding what has been successful — and unsuccessful — in the past, it’s impossible to craft predictable, scalable campaigns that will help drive engagement, fill the pipeline and swell the company coffers in the future.

“It’s a straightforward problem and yet so difficult to solve,” said Matt Heinz, president of Heinz Marketing. “The challenge for most B2B purchases is that it’s not a short line. It’s not linear. And there’s a complicated mix of touch points and attribution moments that happen along the way. It’s never clean. It’s never simple.”

And it’s not like you can attribute an 18-month-long, six-figure enterprise deal for a complex IT solution to somebody downloading a white paper. Business doesn’t work that way.

Heinz contends that most B2B marketers are not yet grappling with the attribution issue in a thoughtful manner. He sees many companies content with tracking things like first touch in their reporting without thinking through the complexity of the buyer’s journey. At the same time, he notes there are more advanced models and methodologies emerging to give marketers significantly deeper insights in their reporting data.

“But it’s also true that the more data that you’re pulling, the less likely it’s going to add up to 100 percent,” he added. “You’re never going to have a perfect story. You need to make the most of that data along with your intuition and instincts.”

New-Waterson echoes the sentiment that it’s crucial to figure out the system that works best for you.

“And then you stick to it,” he added.

After all, that’s a lot easier than frantically juggling a bunch of models and hoping they all don’t come crashing down to the ground.


Cracking the Attribution Code

Marketing teams are under more pressure than ever to quantify their influence and justify budgets. At a time when marketers are expected to prove Return on Investment, LeanData is exploring the challenges they face and how it is possible to improve account reporting.


Previous Article
LeanData and the Business of Giving Back
LeanData and the Business of Giving Back

SAN JOSE — Corporate team-bonding events can take many forms. But this one began with a quick training sess...

Next Article
Account-Based Marketing Won’t Be the Death of Anything
Account-Based Marketing Won’t Be the Death of Anything

This has been proclaimed as “The Year of Account-Based Marketing” by several industry thought-leaders, incl...