Most ad reporting tools tell a comforting story: someone clicks an ad, lands on your site, buys a thing, you give the ad credit. Reality looks more like this:
A prospect sees an Instagram reel. Doesn't click. Three days later they Google your brand name. Click your AdWords brand-search ad. Bounces. A week later they hit your site directly from a podcast you sponsored. Convert.
Who gets the credit?
The four attribution models that matter
Every "attribution" debate boils down to which touchpoint along that journey you decide to trust:
- First-touch: the very first ad someone saw. Great for measuring top-of-funnel awareness, terrible for performance budgets — your prospecting campaigns will look like rockstars and your retargeting will look broken.
- Last-touch (aka last-click): whichever channel was last before purchase. The default in Meta Ads Manager, Google Analytics, every off-the-shelf dashboard. It overcredits anything bottom-of-funnel — branded search, retargeting — and badly undercredits awareness spend.
- Linear: split credit equally across every touchpoint. Fair, but assumes the third tap on a TikTok ad mattered as much as the YouTube pre-roll that introduced you.
- Time-decay: weight credit more toward touches close to purchase. Acknowledges that recency matters without entirely giving up on awareness.
Why this isn't academic
If your dashboard runs on last-click and you've been killing top-of-funnel spend that "doesn't convert," you're probably starving your funnel without realizing it. Conversely, if your dashboard runs on first-touch and you're scaling spend on awareness, you might be paying for clicks that would have converted anyway via your retargeting layer.
Picking the right model isn't about finding the "true" answer — it's about matching the model to the question you're asking:
- "Which campaigns deserve more budget?" → time-decay or position-based
- "Which campaigns are doing the discovery work?" → first-touch
- "What's my CAC at the channel that closed the deal?" → last-click
What Persequor does differently
We compute all five models (first, last, linear, time-decay, and position-based) on the same data, side by side. Click a campaign in the dashboard and you see how its ROAS shifts under each model. The right call is rarely "use this one model" — it's "watch the gap between models." A campaign that looks great on last-click but invisible under first-touch is a retargeting campaign cannibalizing other channels' work. A campaign that's strong under first-touch but absent on last-click is doing real awareness lifting that single-touch reporting will gaslight you out of funding.
The best attribution decision isn't picking a model. It's seeing where the models disagree.
Connect Shopify and an ad platform in Settings → Integrations and the multi-touch view lights up automatically.