Scaling Meta ads from 5L to 50L without losing your CAC

Most founder-led brands hit a wall around 5L a month on Meta. CAC is clean, ROAS looks healthy, and then somebody says — let’s push to 50L. Six weeks later CAC has doubled, the agency is defensive, and the founder is staring at a dashboard wondering what broke. Almost always, nothing broke. The structure was just built for 5L, not 50L. Here’s how we re-architect it.

The 5L structure is not the 50L structure

At 5L, you can get away with three campaigns, ten ad sets, and a creative refresh once a fortnight. At 50L, that exact same setup will collapse. Auction overlap eats you alive, the algorithm runs out of fresh signal, and your best ad goes from a 1.8 ROAS to 0.6 in nine days. The first move is consolidation — Advantage+ Shopping or fewer, broader campaigns with high-volume budget that Meta can actually optimise inside.

Creative volume becomes the rate-limiter

At small spend, one good ad carries you. At 50L, you need 8 to 12 new creatives a week or you’re re-serving the same hooks to people who’ve seen them four times. Founder-led brands have an unfair advantage here — founder-led UGC, behind-the-scenes clips, founder talking to camera. We script and shoot in batches of 30 every two weeks, and we ship before they feel polished.

• Hook variations. Same ad, four different first three seconds.
• Format splits. Reel, carousel, static, founder-talk-to-camera.
• Angle rotation. Pain, aspiration, social proof, founder story, before-after.

Bid strategy: stop using cost cap when you scale

Cost cap looks safe at 5L. At 50L it actively starves your delivery. The algorithm spends an hour finding a cheap conversion, then sits idle for three. We move scaled accounts to highest-volume bidding with a daily budget that’s 3-4x average daily spend, then manage CAC with creative rotation and audience exclusion — not with the bid cap.

Audience exhaustion is the silent killer

Most India-based D2C and service brands have a usable audience of maybe 8-15 lakh people. At 5L of spend you’re seeing 15% of them. At 50L you’re seeing all of them, three times. CAC doesn’t double because the ads got worse — it doubles because you ran out of cheap impressions. The fix is geo-stacking, layered lookalikes, and seriously, more languages. Hindi creative in Tier 2 unlocks audiences that English-only campaigns will never touch.

The 70/20/10 budget split

At scale we run 70% of budget on the proven, broad campaign that already converts. 20% goes to a controlled test campaign with new creative, new hooks, new audiences. 10% goes to retargeting and bottom-funnel. Founders who flip this and put 50% on tests blow through CAC budgets in week one. Founders who put 100% on the winner watch ROAS decay within a month. The split is boring and it works.

Measure CAC, not ROAS, at scale

ROAS is a vanity metric for paid social once you’re past 20L a month. Returning customers, organic, brand search — they all dirty the attribution. We track new-customer CAC against contribution margin, weekly, blended across Meta plus the small Google brand campaign that catches the searches Meta drove. If new-customer CAC is under your contribution margin minus a 25% buffer, you scale. If not, you fix creative before you fix budget.

When to slow down

Three signals: frequency above 3.2 on your top campaign, week-on- week CAC up more than 15%, or click-through rate dropping below 1%. Any one of these and you pause new spend, ship four new creatives, and let the algorithm reset. Founders who push through these signals always end up cutting spend by half a month later.

How we help at The Nerdish Mic

We run Meta ads end-to-end for founder-led Indian brands — D2C, jewellery, services, real estate. Strategy, creative, scripts, media buying, and the weekly CAC review the founder actually wants to read. If you’re stuck under 10L a month and you can see 50L on the horizon, talk to us before you raise the budget. Half our clients come to us after they’ve already tried scaling once and got hurt.