Plan an acquisition pilot you can evaluate honestly

Write the pilot agreement

This is a planning worksheet, not our first campaign-results case study. No customer performance or advertising spend is claimed here.

• Hypothesis: Which customer need and offer are being tested?
• Channel: What placement and account eligibility are verified?
• Budget: What is the total approved cap and who can change it?
• Duration: What is the intended observation period and review cadence?
• Qualification: Which requests count as suitable for the service?
• Stop conditions: What triggers a pause, such as a broken destination or misleading offer?

Use the metrics actually available in the chosen ad account. OpenAI and X document their own advertising tools; a report field should not be assumed identical across them. OpenAI: Ads in ChatGPT, the basics; X: Ads Manager.

Work through the economics without invented results

Illustrative math only — not a budget recommendation or a real campaign result.

If those two customers each bought only a $50 setup, setup revenue would be $100 before taxes and costs. The pilot would not pay for its media spend from setup revenue alone. Optional future care is not guaranteed revenue; evaluate retained revenue and delivery cost from actual cohorts when evidence exists. For a different business, a planning break-even calculation is total pilot cost divided by contribution per accepted sale, rounded up. Include labor, fulfillment and payment costs in that contribution; gross revenue alone overstates what is available to repay acquisition spend.

Publish the limitations with the result

Retain the time window, offer version, campaign settings and attribution coverage. Separate media spend from production, agency and delivery costs. A zero-customer outcome is still a result; spend per accepted customer is not calculable when there are no accepted customers.

Before publishing a case study, obtain permission for customer details, verify the underlying records and include unsuccessful or inconclusive outcomes. Do not substitute sample numbers for missing evidence.

The next step is a reviewed campaign brief with a tested destination and conversion chain. The owner approves the actual budget and launch. A profile setup or fit-check request does not authorize advertising spend.