What Attribution Actually Is (and Why It's Mostly Guesswork)

When a new customer finds you, several things probably happened along the way. They saw an ad, maybe got an email, perhaps a friend mentioned you, then they searched your name and called. So here's a simple question with a surprisingly slippery answer: which of those things gets the credit for the sale?

That question is what attribution tries to answer. And the uncomfortable truth, which most marketing tools won't tell you plainly, is that attribution is mostly educated guesswork dressed up as fact.

What attribution is

Attribution is the practice of deciding which of your marketing efforts deserves credit when a customer takes action. If someone saw three different things before buying, attribution is the rule you use to split the credit between them.

The trouble is that there's no single correct rule, so several competing ones exist — and they disagree:

• "Last touch" gives all the credit to the final thing the customer interacted with before buying. Simple, and it systematically over-credits whatever happens to come last.
• "First touch" gives all the credit to whatever introduced them. Also simple, also wrong in the opposite direction.
• "Multi-touch" tries to split credit across everything. More sophisticated, but it's still just a formula for dividing credit — it's guessing at proportions nobody can actually observe.

Each of these will give you a different answer about which channel is working. None of them is the truth, because the truth — what was actually going on in the customer's head — isn't visible to any tracking tool.

Why it's mostly guesswork

Attribution can see that a customer touched several things before buying. It cannot see why they bought. It has no way of knowing whether the ad convinced them, or whether they'd already decided and the ad was just the last thing they passed on the way in.

That distinction matters enormously. Two customers can have identical, traceable journeys — same ad, same email, same search — where one was genuinely persuaded by the marketing and the other had decided a week ago. Attribution counts them the same. It's measuring the footprints, not the decision.

So when a report tells you "this channel drove 40 sales," what it really means is "this channel was the one our rule decided to credit for 40 sales" — a much weaker claim.

What this means for you

None of this makes attribution worthless. Knowing the rough paths customers take is genuinely useful. The mistake is treating any single attribution number as the final word on what's working, and shifting real money based on it.

The practical posture is to hold attribution loosely:

• Use it for a general sense of which channels are involved, not as a precise verdict.
• Remember that whichever model you're using is flattering some channels and shortchanging others — by design.
• Be especially suspicious when one channel looks like a runaway winner. It may just be the one your rule happens to credit.

Most importantly, attribution can never answer the question that actually matters — not "which thing did they touch last?" but "did our marketing actually cause this sale, or would it have happened anyway?" That's a different and more honest question, and it's covered here: would it have happened anyway?

---

If you'd like help making sense of which of your channels are actually pulling their weight — beyond what the attribution report claims — that's something I do with Canadian small businesses. You can get in touch here.