Attribution is making up stories (and that’s actually fine)

There I was, sitting in another meeting about email marketing ROI, watching an earnest analyst explain why our campaigns were responsible for 23.7% of conversions. The presentation was immaculate. Charts. Bullet points. Numbered points. Everyone nodded thoughtfully at the attribution model that had somehow managed to peer into the souls of our customers and divine exactly which touchpoint deserved the credit.

“So,” I asked, channelling my inner toddler (his name is Derek) with an inconvenient question, “are we going to stop doing email if this number is bad?”

Collective pause.

“Well, no. Email is essential.”

“Right. So what’s the actual question we’re trying to answer?”

And that is when I realised we’d all been performing in the longest-running show in marketing: Attribution Theatre. A frequent production where everyone knows their lines, the staging is solid, and absolutely nobody can explain the plot.

The great attribution fairy tale

Let me tell you about Colgate toothpaste. Not because I have strong opinions about dental hygiene (but, you know, clean them regularly), but because it perfectly illustrates the madness of attribution.

Picture this: You see a TV ad for Colgate. Your brain files it somewhere. Later, you’re wandering through the supermarket aisles (lost, perhaps forever), and Colgate sits smugly on the shelf. Later, an online ad pops up while you’re procrastinating on social media. You mention to your partner that maybe it’s time to try a different toothpaste, because this is what passes for stimulating dinner conversation in adulthood.

Fast-forward two weeks: you’re out of toothpaste, you’re doing the weekly shop, you see Colgate, you buy it.

Now, which touchpoint gets the credit? The TV ad that planted the seed? The in-store display that nudged the consideration? The online ad that reinforced the message? The conversation with your partner that made it feel like a joint decision? Or did you run out of toothpaste and need a replacement?

Plot twist: It was all of them. And none of them. And some of them. And a few of them.

The two stories we tell (and why we need both)

Here’s the thing about attribution that nobody talks about in those glossy case studies: it’s not about finding truth. It’s about building narratives that help people make decisions. And like any good story, you need to know your audience.

Story One is for you and your team—the people elbow-deep in campaigns, trying to figure out if that new creative is working or your targeting has gone entirely off the rails. This story is messy, provisional, full of “we think”, “it looks like”, and “let’s test this theory.” It’s attribution as a debugging tool, not a source of ultimate truth.

Story Two is for leadership—the people who need to understand whether marketing contributes to growth and deserves more budget to expand into new channels. This story is cleaner, more confident, and focused on capabilities and reach. It’s attribution as organisational communication, not tactical optimisation.

The problems start when we confuse these stories. When we try to make tactical debugging look like strategic certainty, or when we expect boardroom narratives to guide day-to-day decisions.

(It’s like using a sledgehammer to hang a picture frame—you’ll make a hole, but probably one a lot more sizable than you intended.)

The honest truth about why we measure

Let’s be honest for a moment (that hat DOES make you look older). The main reason most of us care about attribution is budget justification. We’re not trying to unlock the secrets of human behaviour or create perfect customer journey maps. We’re trying to prove that marketing works, so we can get more money to do more marketing.

And you know what? That’s perfectly legitimate!

But it means the real questions aren’t “which touchpoint converted the customer?” but “are we building our capability to reach more people more often in more places?”

This is where attribution actually becomes useful—not as a microscope for examining individual customer decisions, but as a telescope for spotting growth opportunities. Can we afford to test that new channel? Do we have the operational muscle to scale what’s working? Are we building reach and frequency in ways that support long-term growth?

The fundamentals never went away

While we’ve been obsessing over multi-touch attribution models and customer journey orchestration, the fundamentals have been sitting in the corner. Like the reliable friend nobody invites to parties anymore, they are reach and frequency, brand awareness, and creative that doesn’t make people want to hide under their desks.

But here’s the delicious irony: the data is starting to back up what we knew all along. Getting your message in front of more people, more often, with creative that doesn’t insult their intelligence still works. It still drives growth. It still matters more than knowing whether someone clicked your display ad before or after seeing your social campaign.

This doesn’t mean attribution is pointless—it just means we must consider how it fits into the bigger story. It’s a supporting actor, not the star of the show.

Getting comfortable with uncertainty

The hardest part about attribution isn’t the technical complexity or the data integration challenges. It’s accepting that you’ll never understand the whole story. That customer who just bought your product? They’ve been on a journey that started months ago with a conversation you’ll never hear about, influenced by factors you’ll never measure, making decisions based on criteria that would make your targeting algorithms weep.

And that’s… actually fine.

The companies that succeed aren’t the ones with perfect attribution models. They’re the ones who get comfortable with uncertainty while still making wise decisions. They measure what they can, acknowledge what they can’t, and focus on what moves the needle: understanding their value proposition and getting it in front of their audience as creatively and consistently as possible.

What this means for actual humans doing actual marketing

So, what do you do with all this? First, stop pretending attribution is about finding truth. It’s about building useful stories that help you make better decisions. Ask yourself: what question are you trying to answer, and what will you do differently based on the answer?

Second, embrace the dual nature of measurement. Have one conversation with your team about optimisation and uncertainty. Have a different conversation with leadership about growth and capabilities. Both are valid. Both are necessary. Neither is the whole truth.

Third, remember that everything contributes. Your job isn’t to find the single touchpoint that “worked”—it’s to create a system where multiple touchpoints work together to build awareness, consideration, and eventual conversion. More reach, more frequency, better creative. It’s not rocket science, but it is discipline.

The epilogue (or what happened next)

That meeting about email ROI? We eventually had a much better conversation about what questions we were trying to answer. Not “did email work?” but “how can we make our email program contribute more effectively to our overall growth strategy?” Not “which touchpoint converted?” but “how do we build our capability to reach more people with messages that actually matter?”

The attribution model still produces precise percentages and confident charts. But now we treat it like what it always was: one way of looking at the data, not the final word on reality—useful fiction that helps us tell better stories about what we’re doing and why it matters.

And sometimes, when I’m feeling particularly philosophical (usually after too much coffee), I think that’s actually more honest than pretending we have it all figured out.

After all, in a world where someone can see a TV ad, discuss toothpaste over dinner, and buy a product two weeks later, maybe the real attribution treasure was the uncertainty we collected along the way.