First Touch vs Last Touch Attribution:Run Both
First touch vs last touch attribution: one shows what finds customers, the other what closes them. Run both and read the gap to see what creates your demand.
First touch vs last touch is a question of credit. First touch gives it all to the first interaction and shows what finds customers. Last touch gives it all to the final interaction and shows what closes them. Run both, because the gap is your most useful number.
First touch vs last touch attribution: you need both
Here's the verdict on first touch vs last touch. Don't pick one. Run both, permanently.
Most tools hand you last touch by default. It's the easiest to compute, so that's what you get. And it lies in a very particular direction. It flatters whatever sits at the bottom of the funnel and starves everything that created the demand in the first place.
First touch has the opposite bias. Each is wrong, in opposite directions. That's exactly why running both is more honest than trusting either.
A sale rarely has one cause. Someone reads you for months, sees a post, replies to an email, searches your name, buys. Which of those gets the credit? There's no objectively correct answer. Only useful ones.
The models, and what each one is good for
Before you choose, know the menu.
| Model | Who gets the credit | The question it answers |
|---|---|---|
| First touch | The first interaction | What finds people for me? |
| Last touch | The final interaction | What closes people? |
| Linear | Every touch, evenly | A fair naive answer. A decent sanity check. |
| Time decay | Recent touches count more | Suits short sales cycles. |
| U-shaped | First and last heavily, the rest shared | Credits both finding and closing. |
| W-shaped | First, last, and the point a lead became qualified | Suits longer cycles with a clear mid-point. |
Six models. You don't need six. First touch vs last touch, side by side, is enough, with the discipline to look at both.
Free guide · 26 pagesStop paying for what does not work.The Attribution Playbook: how to find which marketing produces money.Get the free guide →Why the gap matters more than either number
Line your channels up twice. Once ranked by first-touch credit, once by last-touch credit. Now look at how the order changes.
That takes thirty seconds and it's where the real intelligence is.
A channel that looks strong on last touch and invisible on first touch isn't producing customers. It's taking credit for customers something else produced. A channel that dominates first touch and vanishes on last is your engine. And it's exactly the one your last-touch view has been quietly telling you to cut.
Think of it as two groups.
Demand creators. They rank high on first touch and low on last. Content, a podcast, a community, organic social. They introduce you to people who buy later through some other door.
Demand harvesters. They rank high on last touch and low on first. Branded search, retargeting, direct traffic. Genuinely useful. But they can't manufacture demand. They only capture it.
Here's the trap. Fund the harvesters by cutting the creators and you'll watch the harvesters starve over the next two quarters. You'll wonder why the cheap channel stopped working. The harvester's efficiency was borrowed. Attribution is how you see the loan.
An example, framed as an archetype
Say you write a weekly piece for your own site and also run a small branded search ad. Last touch says search is your best channel. It sits right next to the sale, so it gets the credit.
But look at the same customers on first touch. Most of them first arrived through your writing. They read for a while, then typed your name into a search box later. Search didn't find them. Your writing did, and search collected the receipt.
Read only last touch and you'd cut the writing to pay for more search. You'd have killed the thing that fills the pipe. This is the most common way a solo founder destroys their best long-term asset while believing they're being disciplined. A longer treatment of why slow channels need protecting is in how to know which marketing channel makes money.
Set your window first, or both models mislead you
Every system has a lookback window. It's how far back the tool will look for touches that count.
Tools ship with a default that has nothing to do with your business. If the window is shorter than your sales cycle, it erases the start of every journey. That doesn't just weaken first touch. It kills it. And it quietly wipes out every channel that works slowly.
So find your real median time from first touch to money. Set the window comfortably longer. If you don't know your cycle, that's your first finding, and a useful one. The detail is in how to set an attribution lookback window.
How to read first touch vs last touch side by side
You're not looking for the right answer. You're looking for disagreement.
- Both rank a channel high. Probably solid. Protect it.
- High on first, low on last. A creator. Judge it on a long trial, not a month.
- Low on first, high on last. A harvester. Useful, but don't feed it by starving a creator.
- Low on both. The candidate for cutting. Check your tagging and window before you decide.
Then write the finding down in one sentence, out loud if it helps. "Search closes what writing finds." That's a strategy, and it fits on a sticky note.
The limit you can't remove
Neither model sees what no tracker sees. A recommendation in a private group. A colleague forwarding your post. A name heard on a podcast. Both models will give those credits to something else.
That's why this is a compass, not a ledger. Use the models to compare channels. Don't expect either to match your accounts. Add a free-text "how did you hear about us?" on your main form, and use it as a correction beside the tracked data, never instead of it.
If a number contradicts you, check the tagging, the window and the model. If it survives, believe it. That's the whole point of building the instrument.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →What to do this week
- Switch your tool to show first touch and last touch next to each other.
- Write one sentence for what each model answers.
- Name your creators and your harvesters.
- Check nothing you rely on for first touch is being cut by a short window.
Not every tool can do this. If yours can't, that's worth knowing before you build more on it.
Which one to trust when they disagree
They will disagree. That's the feature.
When first touch and last touch tell the same story, you can move fast. When they tell opposite stories, slow down. Before you act on the gap, run the boring checks. Is the tagging consistent? Is the window long enough? Did the join between the form and the person actually carry the source?
If you've never confirmed that last one, the gap may be an artefact. A form that drops the source makes everything look like direct traffic, and direct traffic looks like a harvester. You'd cut the wrong thing with great confidence.
Once the plumbing is clean, trust the disagreement over either number. A model that always agrees with you isn't measuring anything. It's flattering you.
What each model is not telling you
First touch has blind spots too. It gives the whole sale to whoever happened to show up first. That could be a lucky click that did nothing to persuade anyone. If you read only first touch, you'd overpay for introductions that never turn into anything.
Last touch gives the whole sale to whoever was standing at the till. If you read only that, you'd fund the till and starve the shop window.
Neither one tells you why someone bought. Neither sees a private recommendation. Both are lenses. Used together they show you the shape of the journey, and that's enough to make a better decision than you'd make blind.
And better is the bar. Not perfect. A year of slightly better monthly decisions stacks into something that looks, from outside, like a gift for marketing.
Free guide · 26 pagesStop paying for what does not work.The Attribution Playbook: how to find which marketing produces money.Get the free guide →Why this changes what you write and post
There's a payoff beyond channels. Look at first touch by individual piece, not just by platform. Which posts, topics and entry points introduce the people who later pay?
You'll probably find a handful doing most of the work. Make more like those. Stop optimising for the number you can see, which is traffic. Traffic optimisation reliably gives you popular content that sells nothing.
That is the quiet gift of first touch. It tells you what your market wants from you, in the first place it ever met you.
Free assessmentHow trapped are you, really?Take the Corporate Suffocation Index, the free solopreneur readiness assessment.Take the assessment →Frequently asked questions
First touch vs last touch: which is better?
Neither. They answer different questions. First touch shows what finds people for you, last touch shows what closes them. Use both and read the gap between them.
Why is last touch the default in most tools?
Because it's the easiest to compute. That convenience comes with a bias. It flatters the bottom of the funnel and gives the channels that created demand no credit.
Do I need a complicated multi-touch model?
No. Linear, time decay, U-shaped and W-shaped exist, but two models side by side beat one clever model. Start with first and last, and add others only if a specific question needs them.
What does a big gap between the two tell me?
In first touch vs last touch, a big gap separates demand creators from demand harvesters. A channel strong on last touch but invisible on first is likely collecting credit for work something else did.
The full method, with the checklists, is in The Attribution Playbook, a free 26-page playbook.
How trapped are you, really?
Take the Corporate Suffocation Index, the free solopreneur readiness assessment.