If ads do not drive sales, what should you diagnose first?

Audience fit, the offer, page experience and sales follow-up can affect progression from advertising to sales. Missing attribution makes the source of a completed sale harder to identify; it does not establish that no sale occurred. Define the observation, then compare possible explanations.

Do reached accounts have a real buying trigger?

Does the message explain the problem, difference and next step?

Do page, form and CTA reduce decision friction?

Who assesses the enquiry, when do they respond and how is the next step tracked?

Are the enquiry, opportunity and completed sale linked? Treat missing source information separately from missing sales.

Reducing, increasing or pausing spend are not excluded at the outset. Assess the goal, available evidence, sales capacity and reversibility together. For each intervention, define the explanation being tested and the observation that would change the next decision.

In Gordon and colleagues’ comparison with Facebook field experiments, observational measurement methods did not consistently reproduce experimental results. This supports distinguishing recorded associations from causal effects, not abandoning reporting.

Hornpiper’s application: platform records, CRM opportunities and customer-reported sources answer different questions. Rather than treating them as equally conclusive, we explain what each can establish for the decision. Controlled experiments may be appropriate; data volume, the sales cycle and uncertainty inform their design.