// STAGE BREAKDOWN
Under the hood.
What actually happens in each stage — the tactics, the tooling, and the number we're moving. Scroll and the pipeline runs.
01
STAGE 01 · DIAGNOSTIC
Audit — find where the money leaks.
Before a single rupee moves, we run a forensic teardown of the account, the tracking layer and the funnel math. Most accounts don't have a spend problem — they have a measurement and structure problem. We map it, quantify it, and hand you the leak list whether you hire us or not.
- Pixel & CAPI event-quality (EMQ) scoring
- Account structure & learning-phase audit
- Contribution-margin & unit-economics model
- Funnel drop-off & CVR-by-source map
- Creative fatigue & frequency read
- Attribution & data-discrepancy check
output → leak_map.pdf · baseline MER · target nCAC · eta 72h
02
STAGE 02 · INSTRUMENTATION
Instrument — so you can see the truth.
You can't optimize what you can't measure. We rebuild the data layer server-side so iOS, ad-blockers and cookie loss stop lying to your ad accounts. Clean signal in means smarter optimization out — this alone routinely recovers 15–30% of "lost" conversions.
- Conversions API (server-side) + event dedup
- Enhanced conversions & first-party data hashing
- GA4 / GTM server container hardening
- UTM & naming taxonomy enforcement
- Offline / CRM conversion imports
- Consent-mode & data-integrity QA
signal → EMQ 9.0+ · match rate ▲ · dedup on · data loss ▼ ~25%
03
STAGE 03 · BUILD & LAUNCH
Launch — feed the algorithm signal.
We consolidate the account into fewer, better-fed campaigns so the algorithm escapes the learning phase fast, then pair it with a creative engine that ships and tests weekly. Ads are ~80% of performance — so we treat creative like a system, not a one-off.
- Consolidated CBO / Advantage+ structure
- 3:2:2 creative testing (concept · hook · format)
- Hook-rate & hold-rate optimization
- Statistical-significance test gating
- Offer & landing / CRO alignment
- Weekly UGC + static + video refresh cadence
cadence → 12–20 creatives/wk · learning exited · decision 95% CI
04
STAGE 04 · SCALE
Scale — on incremental profit.
Scaling is where most agencies light money on fire — chasing last-click ROAS the platform reports back to itself. We scale on marginal ROAS and prove causality with holdout tests, so every extra rupee of spend is buying real, incremental revenue — not conversions you'd have won anyway.
- Marginal ROAS & spend-elasticity curves
- Geo-holdout & conversion-lift tests
- Budget pacing & bid-strategy control
- Winner scaling / loser euthanasia rules
- Channel diversification (Meta · Google · more)
- MMM triangulation at higher spend
scale on → marginal ROAS · proof geo-lift · guardrail MER floor
05
STAGE 05 · COMPOUND
Compound — make the dollar work harder.
Acquisition is only half the equation. We stack retention, lifecycle and CRO on top so the same traffic returns more revenue over time. As cohort LTV rises, you can afford a higher CAC than your competitors — and out-bid them into the ground.
- Cohort LTV by acquisition month
- Klaviyo / lifecycle flow buildout
- Payback-window & LTV:CAC modeling
- Post-purchase & AOV / bundle testing
- Retention-adjusted bidding headroom
- New-channel expansion off proven economics
north star → LTV:CAC 3:1+ · payback < 90d · email rev ▲ 30–60%
06
STAGE 06 · REPORT & LOOP
Report — tie every rupee to money.
One report, every week, that a human can actually read — no 40-slide deck, no vanity charts. It ties spend to contribution margin and closed revenue, states what we changed and why, and feeds the learnings straight back into Stage 01. The loop closes; the account compounds.
- Single source-of-truth dashboard (live)
- Spend → contribution-margin attribution
- What changed / why / next — in plain English
- Experiment log & win/loss ledger
- Forecast vs. actual variance tracking
- Learnings piped back into the audit
delivered → every Mon · metric profit, not impressions · loop closed