Meta opened a school on 1 September. Two hundred early-stage Indian consumer brands, three months, no retainer, and continuing access to venture capital and industry mentors through Meta Startup School. An advertising business has stopped waiting for demand to arrive and begun manufacturing the firms that will buy it. Call it patronage with a login.

The admissions rule is the confession. A business qualifies only if it is “neither be directly managed by Meta nor working with an agency”, and is beginning to use Meta’s platforms to scale revenue. An agency is the party that argues with a platform about the platform’s own numbers. Meta has screened out the one participant with a reason to check the arithmetic.

Having excluded brands that have an agency, Meta then supplies one. Participants are paired with agency partners for expert guidance “without retainer fees” for three months, after which the guidance presumably acquires a price. The platform has arranged introductions between two hundred founders and the firms that will bill them, and kept the goodwill for itself.

The scale explains the generosity. Search and social platforms, led by Google and Meta, took 64% of India’s digital advertising revenue in 2025, from a market that grew 26% to ₹94,700 crore. Meta says small and medium enterprises make up 92% of its Indian advertiser base, a figure reported by Storyboard18. A company holding that much of a market does not need new customers. It needs the ones it has to spend like bigger ones.

A curriculum written by the seller has one weakness. Its central number does not survive a control group. Meta’s own researchers examined 663 large-scale experiments on the platform using more than 5,000 user-level features, richer data than most advertisers can access. The median experimental lift on lower-funnel outcomes, the purchases a consumer brand lives on, was 5%. The two standard observational methods put it at 24% and 64%. Two of the three authors, Brett Gordon and Florian Zettelmeyer, are Northwestern faculty with no stake in Meta’s ad business; only Robert Moakler drew a Meta salary, which makes their shared conclusion in Marketing Science, that an ad campaign’s causal effect cannot be reliably estimated, harder to dismiss, not easier.

A brand that can only prove it works inside one channel has not built a business; it has built an account.

Meta has run a version of this for years. By April 2022 its VC Brand Incubator Initiative had tied up with fifteen venture funds and trained more than 450 small businesses in India. What Startup School adds is a shorter clock and closer contact: three months, not an open-ended mentorship, with three named funds sitting in the room.

The venture firms are what makes the arrangement durable. Fireside Ventures, DSG Consumer Partners and V3 Ventures sit inside the programme as mentors, reading the same growth figures Meta supplied. A brand trained inside one reporting system states its payback period in that system’s terms, having no other version to state. Capital then moves toward whichever brands the measurement flatters.

Everyone in that room will be reading the same number. Only one of them made it.