Home › Blog · Attribution · 9 November 2026 · 5 min read

First-touch vs last-touch attribution: which to use

First touch credits the channel that introduced the customer, last touch the one that closed the sale. A worked example shows when to use each.

First-touch attribution gives the whole sale to the channel that first brought the customer to your site, and last-touch attribution gives it to the channel they came through just before buying. Use first touch to judge what finds new customers and last touch to judge what closes sales, and read them side by side.

What attribution is

A customer rarely buys on the first visit. They see an ad, look around, leave, come back through a search, leave again, and buy a week later after clicking a reminder ad. Each of those visits is a touch. Attribution is the rule you use to decide which touch gets credit for the sale.

No rule is the true one. Each rule answers a different question, and the useful skill is knowing which question you are asking.

First touch

First touch credits the channel of the customer's first recorded visit. It answers: where do my customers come from?

It suits these decisions:

  • Choosing where to spend to reach people who have never heard of you.
  • Judging prospecting campaigns, which show ads to new audiences.
  • Grouping customers by how you acquired them, for example to compare their value over time. See customer lifetime value by channel.

Its weakness is memory. The first touch can only be the first one you recorded. If a customer first visited on a phone and bought on a laptop, or their cookie expired in between, the recorded first touch is really a later one.

Last touch

Last touch credits the channel of the final visit before the purchase. It answers: what brought the customer back to buy?

It suits these decisions:

  • Judging retargeting, e-mail and branded search, whose job is to close.
  • Short buying cycles, where the first and last touch are usually the same visit.
  • Day-to-day checks, because the last touch is the most reliably recorded one.

Its weakness is that it rewards the channel standing nearest the checkout. A reminder ad shown to someone who was about to buy anyway gets full credit. The channel that introduced that person gets none. Read alone, last touch pushes budgets toward retargeting and away from the campaigns that fill the top of the funnel.

A worked example

Suppose a shop has ten orders in a week, each worth 100 USD, for 1,000 USD in total. The customers arrived in four different ways.

OrdersFirst touchLast touch
4MetaMeta
3TikTokMeta
2GoogleGoogle
1MetaGoogle

Now count the revenue under each model.

ChannelFirst-touch revenueLast-touch revenue
Meta500 USD (5 orders)700 USD (7 orders)
TikTok300 USD (3 orders)0 USD (0 orders)
Google200 USD (2 orders)300 USD (3 orders)
Total1,000 USD1,000 USD

Both columns add up to the 1,000 USD the shop actually took. They tell different stories about TikTok. Under last touch, TikTok earned nothing and looks like a channel to switch off. Under first touch, it introduced three of the ten customers.

If the shop stopped TikTok on the strength of the last-touch column, those three customers would likely never have arrived, and Meta's retargeting would have had fewer people to bring back. Meta's last-touch number would fall too, and nobody would know why.

The reverse mistake is possible as well. Looking only at first touch, the shop might cut Meta retargeting, which closed three orders that TikTok started.

Why platform numbers disagree with each other

Each ad platform runs its own attribution, and it differs from yours in several ways.

Each platform only sees its own touches

Meta knows about clicks on Meta ads. It does not know that the customer also clicked a TikTok ad. So each platform credits itself for every order it touched. In the example above, Meta touched eight orders, TikTok three and Google three. If each claims all of its own, the platforms report 14 orders between them for a week in which the shop had ten.

This is why adding up platform-reported revenue gives a total larger than your real sales. Nobody is lying. Each platform is counting within its own view.

Attribution windows differ

An attribution window is how long after an ad interaction a sale still counts. Platforms choose different lengths, and you can often change them. A sale 20 days after a click may count on one platform and fall outside the window on another.

Views count on some platforms

Some platforms credit a sale to an ad the customer saw and never clicked, if the purchase follows within a short window. Your own site analytics cannot see an ad view. It only sees visits. View-based credit is one of the largest sources of difference between a platform's report and yours.

Dates and estimates differ

Some platforms record a conversion on the day of the ad click, while your shop records it on the day of the order. Platforms also fill gaps with statistical estimates when they cannot observe a conversion directly. Both make daily numbers hard to line up.

How to read the two models together

A practical routine looks like this:

  1. Keep one source of truth for totals: the orders in your shop.
  2. Attribute those orders yourself, by first touch and by last touch, using the UTM parameters on your ad links. Both columns will add up to your real revenue.
  3. Read first touch when deciding prospecting budgets.
  4. Read last touch when judging retargeting and closing channels.
  5. Keep the platform's own number beside yours. When the two move in the same direction, trust the trend. When they diverge sharply, look for a tracking problem first.

A channel that is strong on first touch and weak on last touch is an opener. A channel with the opposite profile is a closer. Most accounts need both, and the two models together show which is which. For the budget-level view, see ROAS vs blended ROAS.

Where Tagbreeze fits

Tagbreeze records each purchase or lead with its first and last touch. Its reports show return on ad spend and cost per result by channel, campaign and ad, beside what the platform itself reports, so the differences described here are visible in one place.

In short

  • First touch shows what introduces customers. Last touch shows what closes sales.
  • Each model splits your real revenue differently, and both add up to the same total.
  • Platform numbers overlap because each platform sees only its own ads, uses its own window, and may count views.
  • Attribute your own orders both ways and keep the platform's number beside them.

Know what your ads really bring in.

It takes a few minutes. Nothing is changed on your site until you install the tag.

Add your site
Hands on a laptop beside parcels ready to ship