Rethinking the Marketing Attribution Challenge
January 22, 2019
I came across an interesting take on procrastination a while back. It went like this: Stop thinking about procrastination as a personality problem. Instead, think of it as an alarm that blares because something is missing, and that something is what’s really stopping you from springing into action.
When it comes to attribution, marketing has a procrastination problem. How else can you explain that it took us until 2017 before more than half of us planned to use multi-touch attribution when — in our heart of hearts — we’ve known multi-touch was a “must” from the day we learned about the shortcomings of single-touch B2B marketing attribution?
When it comes to more intelligent attribution, why delay? I mean, we’re talking about a sixth sense for making decisions that drive higher lifetime customer value and profitability. We’re talking about understanding why customers became customers, and which pieces of content (often previously unheralded) played a pivotal role in that journey. We’re talking about making it known that marketing drives revenue in a bigger way than once thought. Why delay that?
Something must be missing. Maybe we’re putting too much pressure on ourselves, thinking we need to make the leap from single touch to a yet-to-be determined, scientifically perfect marketing attribution model, when what we really need, and what we can actually accomplish, is incremental progress.
Here are a few ways to rethink your marketing attribution challenge so that you can fondly look back on 2019 as the year you advanced your team’s ability to accurately attribute what’s responsible for your success.
There’s More Than One Way to Achieve Multi-Touch Marketing Attribution
Many of us default to last-touch attribution because it’s the easiest to measure. Multi-touch (or multi-channel) attribution is harder to measure, but it’s worth it because it tells the story of how our entire marketing strategy is working. There are two basic types of multi-touch attribution:
In this model, you assign values to various tactics based on predetermined rules such as frequency, recency, or perceived value of the interaction. For example, you might assign a higher score to a later-funnel interaction, or for viewing a video demo than for downloading an ebook. This model tends to be limited when rules are assigned arbitrarily rather than based on sound data.
This model relies on machine learning to apply value to interactions and improve the weighting over time. Algorithmic attribution pulls in sources like cookies, CRM data, historical sales data, and other technologies to analyze the effectiveness of online and offline tactics.
These definitions might be helpful in that we now know what we’re shooting for, but it doesn’t really get us closer to our goal, does it? For many B2B marketers, the best way to start making headway is to shift the focus from clicks to conversions. By following “the curve of marketing math success” (below), we can gradually transition ourselves from being click-focused (only measuring what happens with content) to become conversion-focused (measuring the actions our content drives).
To climb this curve, we need to take advantage of the relevant, actionable data at our disposal. That data can be quite different depending on the company and the company’s data-gathering practices. But we can all take advantage of tools that handle algorithmic attribution on our behalf, like LinkedIn Conversion Tracking.
Using this tool, marketers can better understand attribution and true ROI because it records website conversions tied to your LinkedIn ad campaigns. That means you can show the true value of your ads and branding campaigns that aren’t necessarily the “last touch.” With this information, you might decide to prioritize different ads and campaigns than you would’ve previously because new data shines a spotlight on how valuable they really are.
View Marketing Attribution Through the Lens of Customer-Centricity
If we were to give ourselves a goal to come up with an advanced, algorithmic attribution model by the end of the year, what would that look like? Most of us can’t begin to imagine it, and that might be what prevents us from progress.
Instead, it might be helpful think about attribution in these terms: what do we need to know in order to deliver a better experience to our prospects and customers?
For example, you may want to know which thought-leadership content your “ideal customers” tend to access during their journey so that you can replicate the satisfying experience to create more ideal customers. Now, we’ve moved from algorithmic abyss to a question that can actually move our attribution forward: what data do we need to make this happen?
By viewing attribution from the lens of your customers, and how it can potentially benefit them, we take the pressure off ourselves to create an attribution model that would make Einstein blush. Rather, we can address smaller opportunities to create a more customer-centric marketing program, and we can actually work with our technical teams to pull these iterative improvements into our marketing attribution models.
You’re not procrastinating. You just haven’t figured out what you’re missing yet. For more advice that can help you make marketing attribution hay this year, check out our guide, Solving for Marketing ROI.