A simple example
A customer clicks an advert, later reads an email and then buys. The advertising and email systems may both claim credit; the store still has one order.
Credit is a model-dependent view
A platform may credit an interaction it observed while another platform credits a different interaction on the same order. Their windows, identities and rules can differ. Treat each reported total as a view with a definition, rather than as a separate pool of unique revenue.
Choose the decision before the dashboard
A weekly campaign decision, a market-entry decision and a finance reconciliation need different evidence. Use attributed performance to inspect campaign signals, order data to verify transactions, and appropriately designed experiments when the question concerns incremental impact.
What to document
- Conversion action and revenue definition.
- Reporting window, date basis and time zone.
- Attribution model and available identity coverage.
- Whether modeled or observed data is included.
- Treatment of refunds, returning buyers and cross-market orders.
A practical comparison
Place platform-attributed revenue beside store net revenue, acquisition cost and contribution. If attributed revenue rises while total contribution falls, investigate discounts, margins and customer mix before increasing spend. Do not demand perfect agreement between systems that answer different questions.
Use attribution to ask better questions
Two reporting systems can assign the same order to different touchpoints because their rules, windows and available observations differ. A customer may discover the brand through social content, return through search and buy after an email. Treating each platform’s attributed revenue as an independent sales total risks counting the same business outcome several times.
A useful review begins with the decision you need to make. Attribution can help describe observed journeys and identify patterns, but a budget decision also needs spend, margin, customer quality and the limitations of the evidence. Compare trends on consistent definitions and look for changes in total business performance. Where the decision justifies it, an appropriately designed experiment can provide a stronger test of incremental impact. Do not quietly change the attribution basis halfway through a comparison.
- Write down the reporting window and model used by each source.
- Keep platform-attributed revenue separate from the store’s total order revenue.
- Investigate disagreement before treating a preferred dashboard as the final answer.
Compare ROAS with profitability →
Frequently asked questions
What is marketing attribution?
Marketing attribution assigns credit for conversions to recorded interactions under a chosen set of rules. It helps interpret journeys, but it is different from measuring whether marketing caused additional sales.
Does attribution prove causation?
No. Attribution allocates credit according to a model. A causal claim needs an appropriate evaluation design and stronger evidence.
Why do advertising platforms disagree?
They may use different windows, rules, identities and collection scopes. Compare definitions before interpreting the disagreement as a broken setup.
Sources and further reading
Primary material for the concepts discussed here. Our practical checklists and illustrative examples are editorial recommendations. Source links open in a new tab.