Demand and conversion
Attribution
The rule that decides which touchpoint gets credit for a sale — a modelling choice, never a fact.
Last-click credits the final touch and systematically overstates branded search and retargeting, which are the channels people pass through on the way to a purchase they had already decided on. First-click overstates discovery. Linear and time-decay split credit by rule rather than by evidence. Data-driven models infer it, which is better and still a model.
Every platform reports on its own contribution using its own model, so the sum of all platforms' claimed conversions routinely exceeds the number of orders the business actually took. That is not fraud; it is each one counting a touch it genuinely had.
The only reliable answer to “did this channel cause sales” is an experiment: turn it off in a matched geography, or hold out a share of the audience, and measure the difference in total orders. It is slower and less flattering than the dashboard, and it is the only method that survives a change in the platform's model.
Two platforms each claiming 400 conversions in a month with 620 orders are not lying — they are double-counting shared journeys. Reconcile against the order count in the company's own system before any budget decision, and treat platform figures as directional only.
Not quantities. Methods, artefacts and failure modes — the absence of a formula here is the point rather than an omission.
Is an input to
- ROAS
- Revenue attributed to advertising divided by the advertising spend — a channel efficiency ratio, not a profitability one.
An error in this figure does not stay in it. It reaches one number above it, and each of those is quoted as though it were measured.
The term is a reading taken at a point in a sequence that is already running. These are the sequences, with the place each of them jams.
- From an enquiry arriving to somebody answering it
- A message from a person who might buy, travelling from the form they filled to somebody who can actually answer what they asked. 6 steps, one inbox per channel, and a person per inbox is the arrangement that fails while looking reasonable.
Knowing the definition is not the same as being able to check the figure. These are the procedures that do the second thing.
- Testing a growth number you were given
- “Conversion is up 6.4×. Revenue grew 40%. Abandoned carts down 65%.” · 30 minutes, 5 questions.
A definition is free. Being answerable for the figure it produces is the part that is bought, and this term is a working part of the engagements below.
- CRM, websites and omnichannel
- One customer, one record. Every metric downstream depends on it.
- Client-base reactivation
- The cheapest customer already gave you their contact details.
- Intelligence and decision layers
- A metric with two definitions has none.
The definitions are the easy part. Whether the figure on your dashboard was computed this way is a different question, and usually the more expensive one.