Training · FAMA Elite, Season 3
Training cases, labeled as training.
Diagnoses, audits and briefs from the Facebook Ads Mastermind Academy PH, on real brands or practice data. My client results are in the case files.
Featured · FAMA Elite training case · Performance diagnosis
ROAS fell 43% in one month. The ads weren't the problem.
In this FAMA Elite training case, an ecommerce account's ROAS fell from 1.96x to 1.12x between January and February while CPA jumped 108.7%. Before touching any ads, I ran the FAST check (Funnel, Attention, Spend, Tracking).
CTR had actually risen, which pointed away from creative fatigue, and tracking was clean. The break sat in one place: Click to Add to Cart collapsed 52% while Add to Cart to Purchase held steady. That pointed off the ad account and onto the store, where best sellers were unpublished and prices raised on Feb 2. Daily ROAS went from 3.45x on Feb 1 to 0.60x on Feb 2.
The evidence
- Spend was cut 62.9%, but purchases fell 82.2%. Efficiency broke, not just volume.
- AOV rose 19%, which rules out a demand collapse.
- CPM rose only 12%, which cannot explain a 109% jump in CPA.
- Likely root cause: the Chinese New Year manufacturer holiday stopped restocking.
- On Feb 8 and 9, a single purchase absorbed each day's entire spend.
| Stage | January | February | Change |
|---|---|---|---|
| Click through rate (link) | 4.09% | 4.28% | Up 4.6% |
| Click to Add to Cart | 11.98% | 5.74% | Down 52% |
| Add to Cart to Purchase | 11.91% | 12.79% | Up 7.4% |
| Tracking | OK | OK | No issues |
Action plan
- Republish the best sellers
- Roll back or offset the price increase
- Auto pause any ad set that hits 3x target CPA with zero purchases in a day
- Ramp budget back gradually
- Track Click to ATC daily as an early warning metric
The lesson: the fix was on the website, not in Ads Manager. Pausing good creative would have made it worse.
FAMA Elite training case
The cheapest lead was the most expensive customer
In this FAMA Elite training case, I set targets from the scenario's own economics first, then judged every campaign against them. Lead gen target: CAC at or under ₱8,000 and CPL at or under ₱560, from ₱12,000 gross profit per customer.
| Campaign | CPL | Lead to sale | Est. CAC | Decision |
|---|---|---|---|---|
| A | ₱560 | 7.0% | ₱8,000 | Keep |
| B · most expensive lead | ₱750 | 13.125% | ₱5,714 | Scale |
| C · cheapest lead | ₱400 | 1.5% | ₱26,667 | Pause |
Campaign C had the best CTR, CPC and CPL, and it destroyed the most value: 3.3x over target CAC. Judging by CPL alone would have scaled the worst campaign.
The ecommerce scenario showed the same trap. Target: CPA at or under ₱1,000 and ROAS at 2.40x or higher, from a ₱2,400 AOV. The campaign with the best CTR (2.3%) and hook rate (39%) lost ₱400 on every order, because its landing page wasn't converting the cheap traffic.
FAMA Elite training case · Team audit
Ready to test, not ready to scale
Meta Ads readiness audit for a B2B marketing and AI training company, done as a pod. Offer score: 29/35 on the 7C framework.
- Confirmed through live network inspection that the pixel fired before cookie consent on pricing, and was missing from the homepage
- Spotted Meta's Conversions API Gateway running, and recommended event_id deduplication
- Flagged an unqualified "50x ROI" testimonial as a policy risk before it reached an ad
- Strong outcome and credibility, weakest on cost justification. Top fix: a clear price vs value comparison near the call to action
FAMA Elite training case
Research that refuses to guess
Full market research pack for a mature skin cosmetics DTC brand.
- Every finding tagged Observed, Pattern, Inference, or Gap
- Three independent scam risk tools scored the domain 10 to 37 out of 100
- The brand's own pages contradicted each other: 5 shipping countries vs US only, and 50% vs 60% off on the same price
- Found an unsupported "recommended by dermatologists" claim, next to a competitor's named 12 week, 30 person study
- Spotted a 30 day guarantee that excluded opened product, which likely blocks the most common reason to return makeup: a wrong shade (labeled as inference)
- Rejected a founder name match after verifying it belonged to an unrelated business
FAMA Elite training case
Two directions, one shoot ready brief
Same cosmetics brand, taken from research to production.
Step 1: two competing directions
- Built one direction by hand and one with an AI prompt
- Caught a statistic the AI invented ("hundreds of complaints") and replaced it with what the research actually supported
- Scored six hooks on a 25 point rubric. The top hooks scored 23/25
- Picked the direction backed by three independent sources over the one backed by one
Step 2: the 9 step brief
- Concept: "Check Us Out. We Dare You." turning the buyer's trust objection into the hook
- Angle: "We know you're going to Google us. So here's the tour, unfiltered."
- Delivered three hooks, one shared body script, and a full shot list
- Marked the static brief as conditional instead of faking a missing reference ad
FAMA Elite training case
Reading the market before writing a word
Live ads teardown of a Philippine baby skincare brand using the 3M's.
- Desire: parents wanting peace of mind, not just "gentle skincare"
- Awareness: Solution Aware. Sophistication: level 2 to 3
- Recommended dermatologist testing proof to stand out in a crowding market