← Marianne Jaen — Portfolio
Media Buying Diagnostics · Worked Exercise

Diagnose & Decide

Two funnel scenarios — an ecommerce store and a lead-gen business — worked end to end: set the target, read the actual numbers, diagnose what's really happening, and recommend an action.
Scenarios
eCommerce (Meta ads → purchase) and Lead Gen (Meta ads → booked sale)
Method
Target → Actual → Diagnosis → Action → Why, applied to 3 campaigns per scenario
My approach
Never judge a campaign on one number — trace cost and conversion metrics stage by stage until the real breakpoint shows itself, then recommend one specific action with a stated reason
Data
Illustrative example figures used to practice the method — not a live client account
What this is: before you can call a campaign "good" or "bad," you need a target to measure it against — and before you recommend keep, scale, iterate, fix, or pause, you need to know why a campaign landed where it did, not just that it did. This exercise works two funnel types end to end: calculate the breakeven target, compare three campaigns against it, diagnose the real cause using the metrics in between (not just the headline number), and recommend a specific next action for each.
Scenario A · 1 of 2

eCommerce

A store selling a single core product needs to know the most it can pay to acquire an order and still hit its profit goal — then measure three live campaigns against that number.

Average order value
₱2,400
Variable cost per order
₱1,000
Desired contribution profit
₱400 / order

Step 1 — Calculate the target

Target CPA
₱1,000
AOV − variable cost − desired profit
2,400 − 1,000 − 400 = 1,000
Target ROAS
2.40x
AOV ÷ Target CPA
2,400 ÷ 1,000 = 2.40

Step 2 — Actual performance

CampaignSpendPurchasesRevenueActual CPAActual ROASHookHoldATCWeb CVR
A₱30,00030₱72,000₱1,0002.40x27%14%11%4.1%
B₱18,00010₱24,000₱1,8001.33x39%12%7%1.9%
C₱12,0008₱19,200₱1,5001.60x34%20%14%4.5%
Target₱1,0002.40x

Step 3 — Diagnosis

Campaign AScale
CPA ₱1,000 = target · ROAS 2.40x = target

Lands exactly on target at the largest spend of the three, so it isn't a fluke of low volume — it's a proven, repeatable result. Nothing here is "broken" to fix; the only question is how much further it can go before CPA starts to creep.

Campaign BIterate
CPA ₱1,800 vs ₱1,000 target · ROAS 1.33x vs 2.40x target

This is the strongest ad of the three at the top of the funnel — best Hook (39%), best CTR, cheapest clicks (₱18 CPC) — so the creative is clearly earning attention cheaply. But ATC (7%) and Website CVR (1.9%) are the lowest of the three: people click, then don't convert. That's a post-click problem — the landing page or offer isn't backing up what the ad promised — not a creative problem. Fixing the ad won't fix a page that isn't converting.

Campaign CFix
CPA ₱1,500 vs ₱1,000 target · ROAS 1.60x vs 2.40x target

The mirror image of B: once someone lands, this converts the best of the three (Hold 20%, ATC 14%, Website CVR 4.5% — all the top scores). The problem is reaching people affordably — CPM (₱396) is the highest of the three and volume is the lowest (8 purchases). The offer and page are working; the delivery is expensive. This calls for cheaper or broader audience testing, not a creative or landing-page rewrite — and the small sample size (8 purchases) means it's worth more spend before judging it too harshly either way.

Scenario B · 2 of 2

Lead Gen

A lead-gen business needs to know the most it can pay per lead and still hit its profit goal per closed customer — using its own booked-appointment and show-up rates to translate a customer target into a lead target.

Gross profit per new customer
₱12,000
Desired contribution profit
₱4,000 / customer
Assumed funnel
Lead → Appt. 50% → Show 70% → Close 20%

Step 1 — Calculate the target

Target CAC
₱8,000
Gross profit − desired profit
12,000 − 4,000 = 8,000
Target CPL
₱560
Target CAC × (Appt. × Show × Close)
8,000 × (0.50 × 0.70 × 0.20) = 560

Step 2 — Actual performance

CampaignSpendLeadsCPLAppt.ShowCloseEst. salesActual CACEst. profit
A₱28,00050₱56050%70%20%~3.5₱8,000+₱14,000
B₱30,00040₱75070%75%25%~5.3₱5,714+₱33,000
C₱24,00060₱40025%60%10%~0.9₱26,667−₱13,200
Target₱560₱8,000

Step 3 — Diagnosis

Campaign AScale
CPL ₱560 = target · CAC ₱8,000 = target

Both the cost per lead and every downstream rate (appointment, show, close) land exactly on the assumptions used to set the target. That's the cleanest possible read: scale it and keep watching that the funnel holds as volume grows.

Campaign BScale
CPL ₱750 (34% over target) · CAC ₱5,714 (29% better than target)

By CPL alone this looks like the worst campaign — the most expensive lead of the three. But its appointment (70%), show (75%), and close (25%) rates all beat the assumed funnel, so the blended CAC comes in well under target and it produces both the most sales (~5) and the most profit (+₱33,000) of the three. This is the campaign to scale first, even though its top-line lead cost looks the least attractive — worth checking why the funnel converts this much better (a better-matched audience, or a stronger booking process) before assuming a higher CPL is automatically bad.

Campaign CPause
CPL ₱400 (best) · CAC ₱26,667 (3.3x over target) · Est. −₱13,200

The cheapest, highest-volume leads of the three — and the worst business outcome by a wide margin. Every downstream rate collapses (25% appointment, 60% show, only 10% close, all the lowest of the three), which points to a lead-quality problem, not a lead-cost problem: the targeting or ad promise is likely attracting people who aren't a real fit for what's being sold. Cheap leads that don't convert are more expensive than expensive leads that do — pause and fix qualification before spending further.

Takeaway

What this teaches

The single biggest trap in both scenarios is judging a campaign by the metric that's easiest to see first — CTR, CPC, or CPL — instead of following it all the way to CAC or ROAS.

The pattern repeats in both scenarios: in eCommerce, Campaign B had the best top-of-funnel numbers (Hook, CTR, CPC) and the worst ROAS. In Lead Gen, Campaign C had the best CPL and the worst CAC — actually losing money. In both cases, the cheapest-looking campaign up front was the one that needed to be fixed or paused, while a campaign with a "worse" headline number (Lead Gen B) was actually the one worth scaling. Cost-per-click and cost-per-lead describe the top of the funnel; only CPA/CAC and ROAS describe the business result — that's the number a keep/scale/iterate/fix/pause call has to be based on.
Quick recap

Target, bottleneck, action, why

The same four questions, answered fast, for the campaign in each scenario that needed the most attention.

eCommerce — Campaign B

Target₱1,000 CPA / 2.40x ROAS
BottleneckPost-click: lowest ATC (7%) and Website CVR (1.9%) despite the best Hook and CTR
ActionIterate the landing page / offer; keep the ad as-is
WhyHook and CTR prove the message resonates and earns cheap clicks — the drop happens after the click, which a new ad can't fix

Lead Gen — Campaign C

Target₱560 CPL / ₱8,000 CAC
BottleneckLead quality: appointment, show, and close rates are all the lowest of the three despite the cheapest CPL
ActionPause and re-diagnose targeting/qualification before spending further
WhyActual CAC (₱26,667) is 3.3x target and the campaign runs at an estimated loss — cheap leads that don't convert cost more than they save
Built as an independent diagnostic exercise, applying a structured target-setting and performance-diagnosis framework to two scenario funnels. All figures are illustrative example numbers used to practice the method — not a real client account or live campaign data.