Every dashboard claims the win. Follow the evidence, separate reported credit from causal lift, and make the call you can defend.
Case 1 of 4
The dashboards disagree. What can you report?
Three ad platforms take credit for orders. The store’s order book is the business record. First, establish what actually happened.
54Meta-reported purchases
47Search-reported purchases
68Analytics purchases
72unique paid store orders
The path report looks persuasive. Is it causal?
Among observed order journeys, 52% ended with a paid-search visit. Paid social appeared earlier in 34% of journeys. The groups can overlap.
52%search was last observed
34%social appeared earlier
72orders in this report
Unknownwhat would happen without ads
Design a fair test of paid social prospecting.
The question is whether prospecting ads create additional purchases—not whether a platform can find people who buy.
One questionincremental purchases
Same windowtreatment and comparison
Unique ordersconsistent outcome
Randomizebefore exposure
Read the test result—and decide what to say.
A randomized test ran to its planned end. The estimate is positive, but its uncertainty matters as much as the point estimate.
5.2%exposed purchase rate
4.8%holdout purchase rate
+0.4 ppestimated absolute lift
−0.7 to +1.5 ppsimulated 95% interval
Evidence note
Evidence score
0
out of 100
Your case report
The measurement rule
Attribution allocates reported credit. Incrementality estimates what changed because of advertising. Use each for the question it can answer, and be candid about uncertainty.
Want clearer evidence from a real campaign?
Bring us the measurement question and we’ll help map the right next step.
Fictional educational simulation using synthetic data. Scores are not a forecast or guarantee; real experiments require appropriate design, sufficient data, and careful interpretation.
Your privacy choices
We use cookies to run this site and, with your permission, to understand traffic and measure marketing. Choose what you're comfortable with below.