Performance marketing for D2C brands

What this is

Performance marketing at The Shizz means paid media on Meta, Google, Amazon and quick commerce, planned and bought against one number: profitable orders at checkout. It runs in-house alongside creative, social, strategy and CRO, so the people deciding where the spend goes are the same people making the ads and fixing the funnel the spend lands in.

The work is limited to D2C brands in FMCG, F&B, nutrition and consumer goods selling in India. Over six years that has meant 160+ brands, ₹150 Cr+ of managed ad spend and ₹450 Cr+ of attributed revenue at a 3.8× average ROAS.

How it actually runs

1. Instrument before you spend

The first job on a new account is a foundation, not a campaign: events firing correctly, purchases attributed to the right source, and performance readable by product and by audience. That was the opening move on Pure Whites, and it is the reason a brand can tell a demand problem from a shop problem. A brand that cannot see its own drop-off keeps buying traffic to solve a problem the traffic never had.

2. Decide which products carry acquisition

Hero SKUs are the products that carry higher order value and bring people back, not simply the ones that sell most often. Those are rarely the same product, and building spend around the wrong one caps revenue permanently. On Zama Organics every SKU was treated as its own mini-brand until ketchup and ghee earned the acquisition load. On Aazol, Pro Nature and Brawny Bear the same concentration is what turned scattered budget into readable data.

3. Test at a size that produces a readable answer

A test only teaches you something if one variable moves at a time and the budget behind it is large enough to clear the noise floor. On 1970 Shop the first weeks ran as a structured testing sprint, sized to produce a result rather than spread thin to feel busy, and spend moved into winners as they emerged rather than across the whole set.

4. Keep a manual layer you can read

Broad automated campaign types are good at finding buyers and bad at explaining how. On Aazol the fix was a 60:40 split: 60% of budget on manual cohort targeting for control and readable data, 40% left on Advantage+ for reach. That keeps the efficiency of automation while restoring the ability to learn from it.

5. Segment by cohort, not by blob

First-timers, cart-abandoners and category switchers need different sentences. A first-timer needs to understand the product, an abandoner needs the one objection that stopped them removed, and a switcher needs a reason this beats what is already in the kitchen. My Pahadi Dukaan split first-time scrollers, repeat buyers and fence-sitters the same way, and traction returned almost immediately.

6. Let Meta create demand and Google catch it

On Pure Whites, Google caught the buyer already searching for ghee or cold-pressed oil by name while Meta created the demand that sent them searching. On Aazol, Google had been sitting almost untouched until fresh keyword maps and conversion-built landing pages brought it into the mix. On Bon Fiction, Google went from unused to a top sales driver, which is the clearest evidence that the earlier failure was structural rather than a verdict on the channel.

7. Add marketplace and quick commerce where the category earns it

Lal Sweets is the clearest example: Meta to Blinkit collab ads brought cost per purchase to ₹19 with ROAS touching 10×, while Amazon was worked through better storefront visuals and PPC. The mechanism is that the ad creates the craving and the platform behind it can satisfy it in the same session, which closes the gap where most food advertising loses the sale.

8. Put retention underneath acquisition

When retention holds, every new customer stacks on top of existing revenue instead of replacing someone who has quietly churned. Pure Whites now takes about a third of revenue from repeat orders, which pulls down the blended cost of every sale. On My Pahadi Dukaan the retention flows ran underneath acquisition the whole way, so the loyal base kept compounding while new buyers arrived.

What we run