Explaining a move away from multi-touch attribution
Someone in the room approved the tool you are replacing. Three arguments that work, one that never does, and how to structure it so nobody has to be wrong.
Three platforms, three answers, one budget to defend. Your GA4 export already holds the answer, and the first finding is free and never leaves your browser. Get your first finding free
By Joris van Huët, Founder & CEOUpdated 3 min read
Run the numbers for your store: the free customer journey credit calculator.
Somebody in the room approved the tool you want to replace, so the argument has to be about what changed rather than about what was wrong. That framing is not diplomacy, it is accurate: multi-touch attribution worked considerably better in 2019 than it does now, and the reason is external.
Three arguments that land
| Argument | The evidence to bring |
|---|---|
| Coverage has eroded | Your own consent rate and unassigned traffic share, now against two years ago |
| The claims do not reconcile | Platform-claimed revenue against actual orders for one window |
| The method cannot answer the question asked | A channel where credit is high and the counterfactual is obviously weak |
The first is the strongest because it is measured on your own data and requires no theory. Pull your unassigned traffic share and your consent rate for this quarter and the same quarter two years ago.
The argument that does not land
"Causal inference is more rigorous." True and useless in a board meeting, because it asks a room of non-specialists to adjudicate a methodological claim they cannot evaluate. It also sounds like the pitch the previous vendor made.
Replace it with the reconciliation. Platform claims summing above actual revenue is a fact anyone can check, needs no statistics, and makes the point about credit allocation without requiring anybody to take a position on method. It is set out in the claim ratio.
The slide structure
Four slides. What we used and why it made sense. What changed externally, on our own data. What we propose, in one sentence, including its limits. What we will check in ninety days to know whether it was right.
The fourth slide is the one that gets approval, because it converts a permanent decision into a reversible one. Name the specific number you will report back on.
Pre-empt the two questions
"Will the numbers get worse?" Yes, in the sense that credit will be lower for channels that were being over-credited, and thresholds need resetting. Say this before you are asked, because discovering it in month two looks like concealment.
"How do we know the new number is right?" Every number in the read checks against two files, your GA4 export and the read, recomputed in your browser. The mechanics are in how to prove marketing incrementality in budget meetings.
Making it cheap to try
The lowest-friction version of this proposal is a parallel read on a window already reported by the existing tool: a €99 one-time upload of a Google Analytics export, refundable within 30 days, no questions asked. Bringing the comparison rather than the proposal changes the meeting entirely.
The tone that works
Not "we were wrong". Something closer to: the ground moved, here is the measurement on our own data, here is what I propose, here is when I will report back. Nobody has to lose an argument for that to be approved.
Related answers
Frequently asked questions
How do I justify replacing our attribution tool to leadership?
Lead with coverage erosion measured on your own data: your unassigned traffic share and consent rate now against two years ago. It requires no methodological argument and reframes the decision as responding to an external change rather than correcting a past mistake.Will our marketing numbers get worse after switching?
Credit will fall for channels that were being over-credited, so the headline figures usually drop and thresholds need resetting. Say this before the migration rather than after, because discovering it in month two looks like something was concealed.What makes a new attribution number more trustworthy?
Not a claim of accuracy, which nobody can verify. Every number in the read checks against two files, your GA4 export and the read, recomputed in your browser.
Go deeper: Incrementality testing, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- AnalyticsAnalytics is the systematic computational analysis of data. It reveals customer behavior and measures campaign performance.
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Causal InferenceCausal inference is the set of methods for estimating how much an outcome changes because of a cause, rather than alongside it. It works by comparing what happened with an estimate of what would have happened without the cause.
- CounterfactualA counterfactual is what would have happened under conditions that did not occur, such as a customer's outcome had they not received the campaign they did receive. It is never observed, so every effect estimate has to stand in for it with a control group, comparable units or a forecast.
- Google AnalyticsGoogle Analytics is a web analytics service that tracks and reports website traffic.
- IncrementalityIncrementality is the part of an outcome, such as orders or revenue, that a marketing activity caused: the extra results compared with what would have happened without it. It is a causal effect, not the number of conversions credited to a channel.
- Multi-Touch AttributionMulti-Touch Attribution assigns credit to multiple marketing touchpoints across the customer journey. It provides a complete view of channel impact on conversions.
- RevenueRevenue is the total income generated by the sale of goods or services related to a company's primary operations.