One deal, two stories: first-touch vs last-touch attribution

A prospect finds you through a search ad in February. They download a guide, go quiet for four months, then come back through a webinar in June and sign in July. One deal, say 40,000 in first-year revenue. Which touch gets the credit for that sale? The February ad that started the whole thing, or the June webinar that was in the room when they finally bought?

That question is the whole of first-touch versus last-touch attribution. Attribution just means deciding which marketing effort earns the credit when money comes in. First-touch hands the entire 40,000 to the search ad. Last-touch hands it to the webinar. Same deal, two completely different stories about what worked and where your next budget should go.

Most owners assume there is a correct answer and they are quietly getting it wrong. There usually is no single correct answer, and the fear of choosing wrong is not the thing that should keep you up at night. What actually costs you money is switching models depending on who is in the room. For the wider view of how the models fit together, our guide to marketing attribution covers the full set; this piece is about the two you will reach for first.

First-touch

Credits what brought them in

• Flatters brand, content and demand gen
• Hides everything that closed the deal
• Best when you are judging what fills the top

vs

Last-touch

Credits what closed them

• Flatters sales, demos and the bottom of the funnel
• Hides the work that created the interest
• The default in most tools, because it is easy

What first-touch credits, and what it hides

First-touch gives all the credit to the earliest recorded interaction. The blog post someone read, the ad they clicked, the conference where they first handed over a business card. It answers one specific question well: what pulls new people into your world in the first place.

That makes it the natural ally of demand generation, the top-of-funnel work whose job is to create awareness where there was none. The blog, the paid ads, the sponsorships, the events. If you want to know which channels are genuinely filling the top of your funnel with people who eventually buy, first-touch shows you, because it only ever rewards the thing that came first.

The flattery runs toward anything early. A cheap awareness channel can look like a hero purely because it happened to be the first thing a future buyer touched, even when the deal was really won by a rep who spent six weeks answering hard questions. What first-touch hides is all of that closing work. Read it on its own and you will conclude your sales team and your bottom-of-funnel content did nothing, because by design they never appear.

What last-touch credits, and what it hides

Last-touch turns it around. Every bit of credit goes to the final interaction before the sale. The demo request, the pricing-page visit, the sales call that came right before the signature.

This one flatters the closing end of your funnel: your sales team, your retargeting, the case-study and pricing pages people read once they are already leaning in. If your question is which activities are present at the moment a contract gets signed, last-touch answers it. It is also the default in a lot of tools, because it is the cheapest thing to measure. You do not have to stitch a long history together, you just look at the last event before closed-won.

The cost is that it erases everything that built the demand to begin with. The brand campaign that first made someone aware of you eighteen months ago gets nothing. The webinar series that warmed up a cold market looks like dead weight. Lean on last-touch alone and you will be tempted to cut the very programs that fill the top of your funnel, then wonder six months later why the pipeline thinned out.

When each one is the honest choice

Pick based on the decision actually in front of you, not on which model sounds more sophisticated.

When you are trying to answer where should we spend to get more of the right people to notice us, first-touch is the honest lens. It rewards the channels that start relationships, which is exactly what a demand-generation budget is meant to buy. Judging that budget by last-touch would punish it for doing its job.

When the question is which late-stage plays convert people who are already interested, last-touch is fine, and its simplicity becomes a feature rather than a shortcut. A short, mostly-inbound sales cycle, where people find you and buy within a couple of weeks, does not have enough touches in between for a heavier model to earn its keep. Last-touch will tell you what you need to know without the overhead.

The real trouble is the long, involved deal: months of touches spread across ads, content, events, and several conversations before anyone signs. There both single-touch models mislead you in opposite directions at once, and you will want to spread the credit across the journey instead. That is the case for multi-touch attribution, which shares credit among the interactions rather than crowning one.

Why consistency beats being right

The move that saves you the most money has nothing to do with which model is more accurate. Pick one. Write down which one and why. Then use it for at least a couple of quarters before you let it decide anything.

The failure we run into most is not a company using the wrong model. It is a company using a different model every time someone needs to win an argument. Marketing pulls up first-touch to show the CEO how much demand the brand campaign created. Sales pulls up last-touch to prove they carried every deal over the line. Both readings are true. Both come from the same list of closed deals. And now the two teams who should be planning next quarter together are fighting over whose numbers are real, when the numbers were never in conflict at all. They were answering two different questions.

The value of any model comes from watching the same measurement move over time. If first-touch says search drove 30 percent of your pipeline this quarter against 12 percent last quarter, that trend means something and you can act on it. Switch models between those two quarters and the comparison turns into noise, and every decision built on it becomes a guess dressed up as a number. A single source of truth for how you count credit is worth more than a marginally more accurate model nobody trusts because it keeps changing shape.

What attribution is actually for

Attribution earns its keep only when it changes a decision: a budget you move, a channel you cut, a program you back with more money. If your model cannot survive being tied back to a specific closed-won deal in the CRM, it is decoration, and expensive decoration at that. Getting campaigns to report against revenue in the first place is its own piece of work, which we cover in campaigns that connect to revenue.

First-touch and last-touch are both rough, both genuinely useful, and both far better than the common alternative of crediting your campaigns to nothing at all. Start with the one that matches how your buyers actually buy, hold it steady, and reach for something richer only when the simple version stops answering the questions you are asking of it.

Common questions