> ## Content Index
> Fetch the complete content index at: https://conversion.club/llms.txt
> Use this file to discover other available public pages before exploring further.

# Which attribution model should a multichannel store use?
- URL: https://conversion.club/playbook/multichannel-marketing-attribution-model/
- Published: 2026-10-29T09:14:24.000Z
- Updated: 2026-10-29T09:14:24.000Z
- Description: Pick a stable reporting model, then compare first click and post-purchase discovery responses. Use controlled tests to resolve disputed budget decisions.
- Author: Jeremie Werner
- Tags: Measure performance, #playbook, Analytics, Acquisition, GA4

No attribution model gives a complete account of why someone purchased. Choose one for routine reporting, keep it stable, and compare it monthly with first-click attribution and a post-purchase discovery question. Differences suggest which channels introduce the brand and which close observable journeys.

## The benchmark

Marketing attribution allocates credit for orders among preceding touchpoints. GA4 and Shopify offer different views. This table retains the source's October 2026 documentation snapshot; check current report availability and labels in your account.

| Feature                   | GA4                                                                       | Shopify                                                                       |
| ------------------------- | ------------------------------------------------------------------------- | ----------------------------------------------------------------------------- |
| Models cited              | Data-driven; paid-and-organic last click; Google paid channels last click | Last non-direct click by default; last click; first click; any click; linear  |
| Direct visits             | Generally no credit unless the path is entirely direct                    | Excluded by the default model; included in last click, first click and linear |
| Comparison location cited | Advertising → Attribution → Attribution models                            | Growth/channel report → attribution-model selector                            |

GA4 removed first click, linear, time decay and position-based attribution in November 2023\. Shopify can provide the first-click comparison described here. Its “any click” model gives full credit to every clicked channel, so totals can exceed actual orders; use it to inspect an individual channel, not add up revenue.

Coverage comes before model choice. At a fashion brand in Le Club, native Shopify reports connected only 44 of every 100 orders to a tracked web session. The model comparison therefore described a limited observable subset, not all demand.

## The rule

Each model answers a different question. Last click identifies the final observed channel, first click the first observed one, and data-driven attribution allocates credit algorithmically across recorded paths. None proves causal impact.

Name the chosen model in the [monthly report](https://conversion.club/playbook/what-should-go-in-the-monthly-report-to-leadership/) and keep it stable. Compare other views to identify channels whose apparent value changes materially. A channel strong at first click but weak at last click may introduce journeys another channel finishes. Cutting it could reduce future demand, though the comparison alone does not prove that prediction.

Some influences generate no tracked click: word of mouth, podcasts, store windows or an AI recommendation read without visiting. Track observable assistant referrals through the [GA4 method](https://conversion.club/playbook/how-to-measure-ai-driven-revenue-in-ga4/), and ask buyers about the rest.

## The steps

1. **Fix the reference model.** The source suggests Shopify last non-direct click when [Shopify is the revenue reference](https://conversion.club/playbook/shopify-or-ga4-which-number-should-you-show-leadership/), or data-driven attribution for a team operating in GA4\. State the scope and coverage alongside it.
2. **Compare GA4 models.** Over 90 days, compare data-driven and last-click credit by main channel group. Note the largest shifts, keeping the date range and conversion definition consistent.
3. **Compare Shopify entry and closing roles.** Record each channel's first-click share and last-non-direct-click share. Divide the former by the latter. Ratios above 1.5 or below 0.7 are the source's indicative heuristics for introducing or closing journeys, not validated universal cutoffs. Recalibrate against your history; a zero or tiny denominator needs separate handling rather than a misleading ratio.
4. **Ask one discovery question.** On the [confirmation page](https://conversion.club/playbook/what-should-go-on-your-order-confirmation-page/), ask “How did you first hear about the brand?” Offer actual channels plus word of mouth, press, AI assistant and Other with text. Randomize choices where practical to reduce position bias. Survey answers have recall and response bias, so show sample coverage.
5. **Triangulate monthly.** Maintain four columns: channel, last-non-direct-click share, first-click share and survey share. A channel that grows substantially across these views merits discussion before a budget cut. Do not add the shares across different measurement systems.
6. **Test the disputed channel.** Use a suitable geographic or time-based [incrementality test](https://conversion.club/playbook/how-do-i-know-if-my-ads-drive-sales-i-wouldn-t-have-made-anyway/) and compare total sales with a credible control. At one outdoor brand, paid channels gained 1,100 orders while organic channels lost 1,400\. Attribution reports credited the paid gains; the aggregate movement suggested displacement worth investigating, rather than independently proving its cause.

## The pitfall

Never add the conversions claimed by separate ad platforms. Their windows and rules overlap, often crediting the same order. The sum can exceed actual sales and imply an [acquisition-cost improvement](https://conversion.club/playbook/how-much-does-my-acquisition-cost-drop-if-my-conversion-rate-rises-10/) that did not occur.

Changing models to defend a channel is another danger. GA4 reporting-attribution settings can change historical credit in affected reports. When a change is justified, record the date and scope in the [operating log](https://conversion.club/playbook/how-do-i-keep-track-of-what-moved-my-numbers/) and explain it in the next report. Preserve the earlier reporting basis so the team understands why a comparison moved.

## What to tell your team

> “Instagram accounts for 8% of sales at last non-direct click, 21% at first click and 17% of survey responses. That suggests an introducing role. We will keep the reference model stable and use a regional test next quarter before changing budget.” The figures are fictional; the hypothesis still requires testing.

## Sources

- [Google Analytics: attribution introduction](https://support.google.com/analytics/answer/10596866?hl=en&ref=conversion.club).
- [Shopify: marketing-performance measurement](https://help.shopify.com/en/manual/promoting-marketing/analyze-marketing/marketing-performance?ref=conversion.club).
- [Google Analytics: attribution-model comparison](https://support.google.com/analytics/answer/10596865?hl=en&ref=conversion.club).
- [Google Analytics: attribution settings](https://support.google.com/analytics/answer/10597962?hl=en&ref=conversion.club).