‹ The Operator's Edge

L3 · Paid Course · Level 3 · Difficulty 6–7

The Operator's Loop

The weekly cycle that separates a button-pusher from an operator.

3 modules18 lessons
Scroll ↓

Module

Truth & Attribution

Making the numbers honest before you act on them.

6 lessons ↓

Truth & Attribution

Why platform ROAS double-counts

Every platform claims the same sale — added up, they overclaim reality.

Meta: ₹1.9L
Google: ₹1.3L
TikTok: ₹0.9L
Shopify truth
₹2.7L

Platforms claim ₹4.1L. The store says ₹2.7L. The store is right.

01Each channel takes credit for shared conversions.
02Sum of platform revenue > actual revenue.
03Name the inflation before you trust the report.

Truth & Attribution

Shopify as ground truth

The store knows what actually sold — anchor everything to it.

Meta: ₹1.9L
Google: ₹1.3L
TikTok: ₹0.9L
Shopify truth
₹2.7L

Platforms claim ₹4.1L. The store says ₹2.7L. The store is right.

01Orders are reality; platforms are estimates.
02Allocate store revenue back to sources via sessions.
03Ground truth ends the cross-platform argument.

Truth & Attribution

Building a True ROAS per campaign

Re-attribute real revenue down to the campaign that earned it.

Meta: ₹1.9L
Google: ₹1.3L
TikTok: ₹0.9L
Shopify truth
₹2.7L

Platforms claim ₹4.1L. The store says ₹2.7L. The store is right.

01UTM sessions bridge ads to orders.
02Distribute store revenue, not platform claims.
03True ROAS changes which campaigns you keep.

Truth & Attribution

The inflation ratio and how to brief a client

One ratio explains the whole platform-vs-reality gap to a non-expert.

Score
0
01Inflation = platform revenue / true revenue.
02Lead the report with it, calmly.
03It builds trust instead of breaking it.

Truth & Attribution

Post-iOS / cookieless modeling

When you can't see everything, you model the gap honestly.

● Frequency ↑● CTR ↓

Fatigue shows in the leading signals before ROAS drops. Refresh before the cliff.

01Accept partial signal; stop pretending.
02Use holdouts and trends, not just last-click.
03Decisions over precision.

Truth & Attribution

Anomaly detection: normal ranges + health flags

Define what 'normal' looks like so abnormal raises its hand.

● Frequency ↑● CTR ↓

Fatigue shows in the leading signals before ROAS drops. Refresh before the cliff.

01Set per-metric normal bands.
02Flag deviations automatically.
03Triage flags, don't watch dashboards.

Module

Creative Testing as a System

Turning creative from guesswork into a compounding machine.

6 lessons ↓

Creative Testing as a System

Concept → angle → execution

Test the biggest lever first: concept, then angle, then execution.

Test concept → angle → execution. One winner graduates; the rest fund learning.

01Concept = the core promise.
02Angle = the way in.
03Execution = the polish — test it last.

Creative Testing as a System

Statistical honesty

Most 'winners' are noise wearing a costume.

ROAS
0.0×
CAC
0
AOV
0
MER
0.0×

Four headline numbers — each true, each with a blind spot.

01Small samples lie confidently.
02Wait for volume before you crown anything.
03Kill on signal, not on a single good day.

Creative Testing as a System

The testing budget and cadence that compounds

Reserve a fixed slice for testing, run it relentlessly.

Read truth
Test
Scale
Cut
Report

Same loop, every week, every brand. Consistency compounds.

01A standing testing budget, not leftovers.
02Weekly cadence beats sporadic sprints.
03Winners graduate; losers fund learning.

Creative Testing as a System

Reading creative fatigue before it costs you

Fatigue shows in the leading signals before the ROAS drops.

● Frequency ↑● CTR ↓

Fatigue shows in the leading signals before ROAS drops. Refresh before the cliff.

01Rising frequency, falling CTR are the tells.
02Refresh before the cliff, not after.
03Rotate angles, not just colours.

Creative Testing as a System

A creative knowledge base that feeds the next launch

Every test should make the next test smarter.

Test concept → angle → execution. One winner graduates; the rest fund learning.

01Log what won and why, structured.
02Feed winners back as priors.
03Compounding insight is the real moat.

Creative Testing as a System

AI creative pipeline: one product → 10 scored ads

Generate, score, and stage ten ready-to-launch ads from one SKU.

Test concept → angle → execution. One winner graduates; the rest fund learning.

01Reference-rooted templates from real winners.
02A vision model gates quality before spend.
03Ten paused, scored ads — you pick, you launch.

Module

Scaling & Optimization

Thresholds, not vibes — the disciplined growth cycle.

6 lessons ↓

Scaling & Optimization

The scale decision: thresholds, not vibes

Define the numbers that earn more budget before you feel tempted.

+20%
+20%
+20%
+20%
+20%
+20%

Scale in steps the algorithm can absorb — not by doubling overnight.

01Pre-commit your scale criteria.
02Emotion scales losers; rules scale winners.
03Write the threshold, then obey it.

Scaling & Optimization

Budget moves that don't kill learning

How you raise budget matters as much as whether you do.

+20%
+20%
+20%
+20%
+20%
+20%

Scale in steps the algorithm can absorb — not by doubling overnight.

01Gradual steps keep the algorithm stable.
02Duplicate to preserve a proven ad set.
03Avoid edit-storms that reset everything.

Scaling & Optimization

Audience overlap & self-competition

Your campaigns can bid against each other — find it and stop it.

Campaign ACampaign B

Hover — where they overlap, your CPMs inflate.

01Overlap inflates your own CPMs.
02Measure it; don't assume separation.
03Consolidate fragmented ad sets.

Scaling & Optimization

Pausing underperformers below break-even

Cutting losers is half the job — do it on the number, not the hunch.

Platforms claim4.1L
Meta
Google
TikTok
Actually landed (blended)2.7L

Add up what each platform claims and you over-count reality by 52%.

01Break-even is the line, margin is the judge.
02Give enough signal, then act decisively.
03Freed budget funds the winners.

Scaling & Optimization

The weekly optimization cycle, end to end

One repeatable cycle: read truth, test, scale, cut, report.

Read truth
Test
Scale
Cut
Report

Same loop, every week, every brand. Consistency compounds.

01Same loop, every week, every brand.
02Checklist beats memory.
03Consistency compounds; heroics don't.

Scaling & Optimization

The monthly review a client actually trusts

A report that leads with truth and decisions, not vanity charts.

ROAS
0.0×
CAC
0
AOV
0
MER
0.0×

Four headline numbers — each true, each with a blind spot.

01Blended results first, platform second.
02What changed, why, and what's next.
03No internal jargon, no cost-of-goods drama.

End of The Operator's Loop

That's the tier.

Anish Jajodia · Boring Commerce