What is an attribution model?
An attribution model is the rule that decides which marketing touchpoints get credit for a conversion — and how much each one gets. When a customer interacts with several channels before buying — say a Meta ad, then a LinkedIn post, then a branded Google search — the model determines how that one sale's credit is divided among them.
That sounds academic, but it's one of the most consequential choices in growth marketing. Switch models and the same channel can look essential or worthless, which changes where you spend your next ad dollar. The right approach isn't to search for the one "true" model — there isn't one — but to understand what each model rewards and choose deliberately.
Single-touch vs multi-touch.
Attribution models generally fall into two primary families:
Single-touch models
Single-touch models give 100% of the credit to one interaction — either the very first or the very last. They are simple and easy to explain, but they ignore the rest of the buyer's journey.
Multi-touch models
Multi-touch models spread credit across every interaction, weighted by position or recency. They reflect reality much better — most SaaS purchases involve several touchpoints — but they require capturing the full journey with reliable first-party server-side tracking.
For most SaaS teams with considered, multi-step buying journeys, multi-touch is closer to the truth. But single-touch models are still useful lenses — especially first-click for judging what starts customer relationships.
The models, one by one.
First-click (first-touch)
Gives 100% of the credit to the first interaction. (Like crediting the person who introduced you.)
- Good for: Valuing demand generation and top-of-funnel awareness.
- Weakness: Ignores everything that nurtured and closed the deal.
Last-click (last-touch)
Gives 100% to the final interaction before conversion. (Like crediting whoever was standing there at the finish line.)
- Good for: Short cycles and judging closing channels.
- Weakness: Over-credits branded search and direct, and starves the channels that created the demand — the default that misleads most.
Linear
Splits credit evenly across every touch. (Like giving every runner on a relay team the same medal.)
- Good for: Full-journey visibility with no positional bias.
- Weakness: Treats a decisive touch and a trivial one as equal.
Time-decay
Gives more credit to touches closer to the conversion. (Recent conversations weigh more than old ones.)
- Good for: Longer sales cycles where late-stage nudges matter.
- Weakness: Undervalues the early awareness that started everything.
Position-based (U-shaped)
Gives 40% to the first touch, 40% to the last, and splits the remaining 20% across the middle. (Rewards the introduction and the close, with some credit for the nurture in between.)
- Good for: Valuing both discovery and conversion at once.
- Weakness: The fixed weights may not match your actual funnel.
W-shaped / custom
Extends position-based by also crediting a key middle milestone (like lead or opportunity creation), or lets you set your own weights. (Custom-tailored credit alignment.)
- Good for: Teams with well-defined funnel stages who want the model to reflect them.
The same journey, every model.
Take one customer who saw a Meta ad, then engaged on LinkedIn, then converted through branded Google search. Here's how a single $149/mo conversion's credit is distributed under each model:
Under last-click, Meta looks worthless. Under first-click, it looks essential. Neither is "wrong" — they answer different questions. This is exactly why you compare models rather than trust one, which is what Signal Sparrow's attribution models → let you do side by side.
Which model should you use?
There's no universal answer — match the model to how you actually sell:
The best practice for SaaS: don't default blindly to last-click, view a few models together, and make sure whichever you choose still reconciles to real revenue. A model that distributes credit beautifully but doesn't tie to Stripe MRR is telling you a nice story, not the truth.
Attribution windows matter too.
The model is only half the setup — the attribution window (the lookback period in which a touch can earn credit) is the other half. For trial-led SaaS where the paid conversion happens weeks after the click, a 7-day window credits no one, while a 90-day window tells a completely different story. Set the window to match your real sales cycle, and it's best to reattribute your history against different windows to pressure-test your assumptions.
Common mistakes to avoid.
Compare models, keep the truth.
The healthiest way to use attribution models is to see several at once, on data you trust. Signal Sparrow computes first-click, last-click and every multi-touch model on the same journeys, reconciled to Stripe — so switching models redistributes credit without ever changing your total revenue. Whatever tool you use, the principle holds: compare deliberately, set the right window, and reconcile to real MRR.
