59 leads generated. A campaign refined through testing.
How Rare Blue Moon used Meta Ads testing to generate measurable lead volume, identify a promising lead-capture approach, and establish a clearer baseline for future optimization.
Edge AI wanted to generate leads for strategy-call opportunities through paid social. The campaign needed to identify a workable combination of ad format, audience targeting, and lead capture—while balancing lead volume with the need to learn more about lead intent.
A test-and-learn approach.
Instead of treating the first campaign setup as the final answer, RBM adjusted the approach as performance data came in.
The outcome.
August cost per lead was $49.23, below the projected $50–$150 target range. One strong Instant Form week delivered 16 leads at $24.99 each. Campaign tests helped clarify which lead-capture approaches warranted further evaluation.
Against target.
August cost per lead was $0.77 below the projected target floor of $50. This is a modest but positive efficiency result against the stated benchmark.
What the process shows.
Test with a purpose: each change explored a specific campaign question. Change based on evidence: early results informed changes to lead capture and targeting. Keep learning visible: the campaign produced learnings and benchmarks for future testing.
Insights.
Cost efficiency: August's $49.23 cost per lead is a useful benchmark for future tests, not a guarantee of future performance. Engagement and traffic: a $1.13 link CPC and 6.85% link CTR in August indicate the ads attracted clicks at a cost within the projected target, and 533 landing-page views add context. Lead-capture learning: the Instant Form phase included a week with 16 leads at $24.99 each, a promising signal worth validating with controlled tests and lead-quality tracking. Measurement gap: booked strategy-call totals were not available in Meta reporting.
What the numbers show
August active campaign, Aug 1–31. The full engagement and the August report describe different windows, so they are kept separate.
A test-and-learn approach
Establish a baseline
The campaign began with a landing-page ad format designed to drive strategy-call bookings. The first week produced no conversions, so the approach was changed rather than continuing unchanged.
Test native Instant Forms
RBM shifted to Meta Instant Forms and tested the medspa category. This phase produced a stronger run of conversions, including one week with 16 leads at $24.99 per lead.
Explore lead intent
A higher-intent form was introduced to explore whether additional qualification could help balance lead volume with intent. This was a test direction, not a confirmed improvement in lead quality.
Expand audience testing
Broader targeting was tested to evaluate reach. Significant changes reset Meta’s learning phase, which can bring short-term volatility as delivery recalibrates.
Use performance signals
Cost per lead, cost per click, click-through rate, and landing-page views helped build a more informed picture of campaign efficiency.
Recommended next steps
- Validate Instant Forms: run structured tests comparing form types while tracking both cost per lead and downstream lead quality.
- Track leads through bookings: connect Meta lead data to CRM or appointment records so booked calls and lead-to-booking rate can be measured.
- Change fewer variables at once: use clear test windows and avoid making multiple major changes simultaneously when practical.
- Optimize for business outcomes: once enough data is available, evaluate cost per qualified lead and booked call alongside volume and platform metrics.
The takeaway
Good marketing is not just about launching ads. It is about learning what the data can support—and using those insights to make the next decision better.
Measurement & publishing
notes.
- Booked strategy-call totals and downstream conversion data were not available in Meta reporting.
- This case study reports lead-generation and traffic metrics only; it does not claim confirmed appointments, revenue, or customer acquisition.
- Campaign edits and learning-phase resets can affect delivery, so results reflect a testing period, not a controlled experiment isolating any single change.
- Full engagement: Aug 10–Sep 22, 2026. August report: Aug 1–31, 2026.