₹62 Crore Managed Across
Eight Ad Accounts And 18 Markets.
EdTech Company 1, an education and professional-training business, operates through multiple subsidiaries. I run paid acquisition across all of them — six Google Ads accounts and two Meta accounts, from India through the Gulf to North America and Australia.
lifetime
delivered
per lead
Most performance marketers manage one account. This is a group structure — separate ad accounts for separate subsidiaries and separate billing currencies, each with its own conversion setup, budget ceiling and market. The work isn't running campaigns; it's deciding which of eight accounts a rupee belongs in.
| Account | Platform | Mandate | Spend | Leads |
|---|---|---|---|---|
| Flagship | Core acquisition — languages, SAP, Six Sigma, DBA | ₹54.4 Cr | 1,428,610 | |
| SAP specialisations | 13 dedicated SAP module campaigns | ₹2.17 Cr | 49,920 | |
| International | Google USD | USA, language & TEFL — billed in dollars | ₹2.53 Cr | 15,499 |
| Meta International Account | Meta | International DBA — Gulf, Mauritius | ₹2.21 Cr | 39,088 |
| Meta Growth Account | Meta | Franchise, language & social demand | ₹1.37 Cr | 49,354 |
| Review network ×3 | Brand defence across 5 review microsites | ₹18.6 L | — |
Scale is the easy half. The harder question is whether efficiency holds as budget grows. Here is a five-month window inside the flagship account — a ₹1 crore-a-month operation — starting from its least efficient month on record.
Five consecutive monthly declines — while cost per click rose 8.9%
₹465 → ₹438
11.17% → 12.93%
five-month window
five-month window
The Problem
Before touching a bid, I mapped where efficiency actually sat. Cost per lead ranged 25× across the account — ₹47 to ₹1,185. That isn't variance, it's a routing problem. 32 of 55 campaigns were running above blended CPL while absorbing 47% of spend. Twelve efficient campaigns were capped by budget. And with Target CPA running on 53 of 55 campaigns, the bidding was only ever going to be as good as the conversion signal underneath it — which had gaps.
The Intervention
- Keywords — match types and intent tiers restructured so budget follows purchase intent, not search volume
- Location targeting — split by state-level economics; regional and national demand price differently and shouldn't share a bid
- Landing pages — rebuilt for the highest-spend verticals, where a point of conversion rate is worth the most
- Tracking — conversion measurement corrected first, so automated bidding optimised toward real leads
- Campaign optimisation — budget systematically reallocated from the ₹1,100 CPL tail toward the ₹331 CPL core
| Campaign | Spend | Apr CPL | Aug CPL | Change |
|---|---|---|---|---|
| Language Campaign — French | ₹40.6 L | ₹526 | ₹430 | −18.4% |
| Language Campaign — Japanese | ₹17.2 L | ₹474 | ₹398 | −15.9% |
| SAP Campaign — Generic | ₹32.7 L | ₹392 | ₹331 | −15.5% |
| Six Sigma Campaign — Generic | ₹26.2 L | ₹737 | ₹637 | −13.6% |
At eight accounts, allocation beats optimisation. The largest gains don't come from improving a campaign — they come from noticing an account is the wrong home for the budget it holds.
Measurement comes before bidding. Automated bidding is a signal amplifier — feed it an unreliable conversion signal and it will scale the wrong thing efficiently.
Budget caps on efficient campaigns are the most expensive error in a large account. Twelve capped campaigns held 29% of spend — the account was throttling its own winners.
A market is not a translation. Gulf, US and Indian demand price differently, convert differently and search differently. Running 18 markets means 18 cost structures, not one campaign duplicated.
A 25× CPL spread is a routing problem, not a creative problem. The fix is moving money, not making more ads.
Falling CPL against rising CPC is the only efficiency that counts. Clicks got 8.9% more expensive over these five months. The gain came from the funnel, not the market.
Spending at this scale and not sure where the waste is?
Let’s TalkFigures compiled from the client’s own Google Ads and Meta Ads account exports across eight accounts, covering all available reporting history to 31 August 2026. The five-month efficiency window is 1 April – 30 August 2026, measured April vs August. The international account bills in USD and is converted at ₹83/USD; at any higher rate the figure understates. Market count is derived from campaign-level targeting in campaign names. Figures marked Modelled are derived, not observed — engagement click volume evaluated at the pre-engagement conversion rate. All amounts INR unless stated.