Proof in the field

Meta didn't just drive their D2C. It lifted their Quick Commerce too.

When demand shifted nearly 50% from owned D2C to quick commerce, the question was not which channel was selling, but which was actually causing growth. A synthetic experiment across matched cities found the hidden halo.

When you're growing 2-3x year-on-year and answering to investors, every rupee needs a reason.

CMO and Founding Team, Sweet Karam Coffee

New stores opened the doors. Media filled them.

A fast-growing fashion label scaled from D2C into hundreds of offline stores, and needed to know how much growth its media caused, not just where sales landed. Two independent models, orders and footfall, ranked every campaign by the footfall it drove.

Performance had the credit. Brand built the demand.

A scaled wellness brand poured budget into brand media, YouTube, TV, Meta, OTT and Print, while last-click credited performance. A geo-isolated incrementality test measured what each channel actually caused, ranking them by real lift, not clicks.

Organic traffic got the credit. Brand ads created the demand.

A fashion e-commerce marketplace ran Meta brand upper-funnel ads in test cities and kept comparable cities dark. A geo-isolated Synthetic Control model rebuilt the organic traffic those cities would have seen with no ads, isolating the growth the campaign truly caused.

TV moved the percentage. Digital moved the money.

Three geo synthetic-control experiments, two TV flights and one digital brand flight, measured what each medium actually caused on the owned store and on marketplace. The biggest percentage lift was not the biggest return.

You cannot click a CTV ad. You can still measure one.

Click attribution cannot see a living-room screen. So CTV ran in two states and nowhere else, a synthetic twin rebuilt what those states would have sold with no CTV, and a mix model set the result against every other channel.

The ads manager said 2.07. The model said 5.40.

A leading hair and skin care brand planned its ₹30 crore a year across Meta and Google from the numbers each platform reported about itself. A mix model measured what each channel actually caused, then what each product's ads did to the others.