Why attribution will never tell you the whole truth

Dear reader,

I've watched founders move their entire budget because one attribution model said a channel was carrying the business. Then they switched models and the story flipped completely. Both can't be right, and that's the point worth understanding.

Attribution is a model, not a measurement

When someone sees an ad, reads a review, clicks an email, and finally buys, who gets the credit? There's no objectively correct answer. Every model just picks a rule and applies it.

• Last click hands all the credit to the final touch and ignores what warmed them up
• First click does the opposite and over-rewards discovery
• Linear spreads it evenly, which is tidy but rarely true
• Data-driven guesses at weightings you can't fully see or check

None of these is the truth. They're different lenses on the same blurry photo.

What to do instead

You don't need perfect attribution. You need decisions that survive being slightly wrong.

1. Pick one model and stick with it so trends stay comparable
2. Watch overall blended numbers — total spend against total revenue
3. Run the occasional geo test or pause a channel to see what really moves

If a channel only looks profitable under one specific model, be suspicious. Real winners tend to look good from several angles.

The honest goal isn't a spreadsheet that assigns every penny. It's knowing roughly where growth comes from and not fooling yourself. If your channels are all claiming the same sales and the maths no longer adds up, start with a FREE call and we'll find a view you can actually trust.