Last Click Attribution: What It Shows and What It Hides

Last click attribution is the oldest rule in digital marketing measurement and still the one most store owners are reading without knowing it. It is simple, it is consistent, and it quietly moves credit from the channels that find customers to the channels that happen to be there at the end. This guide explains what the last click attribution model does, where it is still the default in 2026, and how to tell whether it is misleading you.

The short answer: last click credits the closer, and Arktis shows what that costs you

Last click attribution gives 100 percent of the credit for a sale to the last channel or ad the customer clicked before buying. It is accurate about one thing, which is the final step, and silent about everything before it. Shopify selects Last click by default in reports that combine sales with a marketing dimension, Google Analytics 4 offers a last click model alongside data-driven attribution, and Google Ads still supports last click although data-driven is its default for most conversion actions.

Whether that matters for your store depends on how many of your customers touch more than one channel, and for finding that out Arktis is the clear winner for an online store. It compares last-touch against first-touch, linear, time-decay and position-based credit per ad platform on your own orders, so you see how much each platform gains or loses when the rule changes. Plans are $49, $149 and $349 a month, and Growth has a 7-day free trial.

Key takeaways

Last click is not one model but a family. Shopify's Last click counts a direct visit as a click, so a customer who types your address at the end hands the sale to Direct. Last non-direct click, which Google Analytics 4 calls Paid and organic last click, skips direct visits and credits the last real channel. Both give everything to the end of the journey, so discovery channels look weak and email, branded search and retargeting look strong. It is a fair summary when customers buy in one visit from one channel and a poor guide to budget when they do not.

What last click attribution is

Under a last click attribution model, one touchpoint receives all of the credit for a conversion: the final click before the purchase. Google Ads defines its version as giving all credit for the conversion to the last-clicked ad and corresponding keyword, and notes that this ignores the other ad interactions a customer may have had along the way.

It is a single-touch model, like first click, and the opposite of it. Our comparison of first-touch vs last-touch attribution covers that pairing, and marketing attribution models explained covers the full set. This article stays with last click itself, because it is the one already built into the reports you open every day.

Where last click is still the default in 2026

In Shopify, when a report contains a sales metric such as gross sales and a marketing dimension such as referring channel, an attribution menu appears and the Last click model is selected by default. Shopify's Last click gives 100 percent of the credit to the last channel the customer interacts with, including direct visits. For marketing activity data the default is Last non-direct click instead. Shopify also offers First click, Any click and Linear, and its marketing activity report credits only the source clicked most recently within the past 30 days.

Google Analytics 4 has offered three models since November 2023, when first click, linear, time decay and position-based were removed: data-driven attribution, Paid and organic last click, and Google paid channels last click. Google's help page states that all of them exclude direct visits from receiving credit unless the path consists entirely of direct visits.

Google Ads has two models left. Its help page says first click, linear, time decay and position-based are no longer supported, that conversion actions using them were upgraded to data-driven attribution, and that last click is still supported. Data-driven is the default for most conversion actions.

Last non-direct click

Last non-direct click exists to fix the most obvious flaw in strict last click. A buyer who finishes by typing the store's address is a Direct sale under strict last click, which tells you nothing about what brought them.

Last non-direct click ignores that final direct visit and gives the credit to the last channel before it. Google's own example for Paid and organic last click is a path of display, then social, then paid search, then direct, where 100 percent goes to paid search. Google states that Paid and organic last click and Last non-direct click are two names for the same model. Shopify's version works the same way, and if a journey contains only direct traffic the order is attributed as direct.

This is why the same store can show a large Direct share in one Shopify report and a small one in another, or in Google Analytics. Before comparing two last click numbers, check which of the two rules each one uses.

What last click gets right

Last click is easy to explain, and every sale is counted exactly once. The credited channels add up to your total orders, which is not true of the platforms' own reports, where each one can claim the same sale.

It also needs very little data. And it is honest about what it measures: the last click really happened, on your site, and can be checked against the order.

For channels whose job is to close, it is a reasonable measure. If you want to know whether a cart recovery email or a retargeting campaign is bringing people back at the moment of purchase, last click answers that.

What last click hides

The cost is everything upstream. A customer who first found you through a prospecting ad on social, and bought a week later after clicking an email, is an email sale under last click. The ad that found them gets nothing. Do this across a few thousand orders and prospecting looks unprofitable while email looks like your best channel.

Branded search has the same effect. Someone who already knows your name and searches for it was sent by something else, but the search click is last, so it takes the sale. Retargeting, which by definition reaches people who have already visited, benefits in the same way.

Cut the channels that last click undervalues and the closing channels have fewer people to close, a decline that shows up later and is hard to trace back.

Last click also shares one limit with every click-based model: it cannot see an ad that was viewed and not clicked. That is a separate problem, covered in our guide to the attribution window, and it is a large part of why Facebook Ads and Shopify sales don't match.

A worked example

Take one $120 order, as an example. The customer clicks a TikTok ad on Monday and browses. On Wednesday they click a Google search ad for the product category. On Friday they click a link in your newsletter. On Sunday they type your address and buy.

Under strict last click, as in Shopify's Last click model, Direct receives the full $120. Under last non-direct click, email receives $120. Under first click, TikTok receives $120. Under a linear model across the three marketing touches, TikTok, search and email receive $40 each. Under a position-based model that gives 40 percent to the first touch, 40 percent to the last and 20 percent to the middle, TikTok receives $48, search $24 and email $48.

One order, five answers, and none of them is a measurement error. TikTok is worth anything from nothing to the whole order depending on the rule, so it should not be judged on last click alone.

Last click vs data driven attribution

Data-driven attribution replaces the fixed rule with a calculation. Google Ads describes its data-driven model as distributing credit for the conversion based on your past data for that conversion action, using the account's data to calculate the contribution of each interaction across the conversion path. Google Ads has a Model comparison report that puts last click and data-driven side by side, with cost per conversion and conversion value per cost for each.

The trade is transparency for nuance. Last click can be checked by hand against any order. A data-driven result cannot, and it is calculated inside the platform whose ads it is evaluating, on that platform's view of the journey. Our guide to data-driven attribution goes into how it works and when there is enough data for it.

When last click is fine

Last click loses little when journeys are short. If most customers buy during their first visit, the first click and the last click are the same click and every model gives the same answer. It is also fine when you run a single paid channel, because there is no other channel for the credit to be taken from.

It stops being fine when customers take days and several visits, and when you run both prospecting and closing channels. You can check which case you are in. In Arktis, the Ads Analytics dashboard shows the share of converters who touched two or more channels before converting, the share who would be credited to a different channel depending on the model, and how long converters took from first visit to first conversion.

Comparing last click with other models in Arktis

Arktis captures ad click identifiers on landing, including gclid, fbclid, ttclid and msclkid, along with UTM parameters, and holds them against the visitor across sessions. When that visitor orders, the dashboard compares first-touch, last-touch, linear, time-decay and position-based credit per ad platform, over a lookback of 7, 14, 30, 60 or 90 days. Its position-based model uses the 40, 20, 40 split from the example above, and its time-decay model uses a 7-day half-life.

The comparison covers orders whose matched session carried an ad click identifier, so its last-touch is the last paid click among those orders, and an order placed on a later direct visit is left out of the comparison rather than credited to the earlier click. A platform that scores far higher on first-touch than on last-touch is introducing customers that other channels close. One that scores higher on last-touch is closing.

The free attribution model comparison tool shows how the five models split credit across channels, without an account.

Where Arktis does not fit

Rule-based models redistribute credit; they do not prove causation. That takes incrementality testing or marketing mix modelling, and the model comparison is neither.

The model comparison covers visits with an ad click identifier, so a campaign tagged only with UTM parameters appears in the UTM campaign report and not in the comparison. Plans list 7 days of data retention on Starter, 30 on Growth and 90 on Scale, so check your plan before relying on a 60 or 90 day lookback. If you need last click only, your Shopify reports already include it.

Try it on your own orders

To see how far last click is from the other models in your store, start the 7-day Growth trial. Plans are on the pricing page.

Sources

Shopify Help Center: marketing reports, the five attribution models, their defaults and the treatment of direct visits, accessed 1 October 2026

Shopify Help Center: measuring marketing performance, the most recent click within the past 30 days, accessed 1 October 2026

Analytics Help: get started with attribution, the three GA4 models, the direct-visit rule and the last non-direct click naming, accessed 1 October 2026

Analytics Help: select attribution settings, models removed as of November 2023, accessed 1 October 2026

Google Ads Help: about attribution models, last click and data-driven definitions, the default and the Model comparison report, accessed 1 October 2026