01
Creative breadth
10 static images so the test never relied on one visual angle, letting the system find stronger combinations.
CASE STUDY 02 · SUPPLEMENT BRAND · MOROCCO
Validating a new product with 10 images and $164.76.
I wanted to validate whether a new product concept could generate efficient purchase intent through a controlled creative test: 10 static images in one flexible ad set, optimized for purchases, judged on full-funnel economics rather than CTR.
Full case study PDF (4 pages)Product: nicotine-free oral pouch (snus-style format) positioned around caffeine and vitamins. No health or medical claims are evaluated here. Brand withheld for confidentiality.
SPEND$164.76test investment
PURCHASES76website purchases
CPA$2.17cost per purchase
ROAS10.80xreported purchase ROAS
ATTRIBUTED VALUE$1,779spend × ROAS
IMPRESSIONS192,563115,543 people reached
LANDING VIEWS4,601qualified visits
FREQUENCY1.67xlow repetition
TESTING SYSTEM
Why a flexible 10-image test.
Fast creative signal without fragmenting budget across ten isolated ad sets: the purchase objective stays central and Meta distributes the visuals against the same conversion goal.
02
One purchase test cell
The creative pool tested inside one flexible ad set: less fragmentation, one read of the full system’s economics.
03
Read the full funnel
CPM, CTR, landing arrival, conversion, CPA and ROAS together. A click alone decides nothing.
04
Scale blocked by compliance
The test earned the right to scale; the manufacturer had to stop this product version before the planned phase.
FUNNEL ANALYSIS
Where the performance came from.
WIDTHS ON A LOG SCALE · CLICK A STAGE
- CPM (cost / 1,000 impressions)
- $0.86
- Link CTR
- 2.92%
- Click → landing page
- 81.90%
- Landing page → purchase
- 1.65%
- Cost / link click (derived)
- $0.029
- Cost / LPV (derived)
- $0.036
Clear scaling headroom remained. Frequency was just 1.67x, with no meaningful saturation. At a $2.17 CPA and 10.80x ROAS, the test justified scaling.
COMMERCIAL OUTCOME
The test had earned the right to scale.
It was stopped by a manufacturer-level legal / compliance issue with this product version — not by media performance. The campaign ended in validation, before the planned scale phase, with meaningful headroom for a compliant version.
- CPA $2.17
- ROAS 10.80x
- Frequency 1.67x
EVIDENCE
Meta Ads Manager results.
The source screenshots for this case study’s metrics. Ad-set label anonymized. Click to enlarge.