Measuring UGC ROI for FMCG launches

User-generated content for FMCG launches in India is where the gap between "everyone says it works" and "we can prove what worked" is widest. The creative volume is high (dozens of variants per campaign is common), the production cost per variant is low (a single UGC creator shoots ten spots in a day), the reach is broad (FMCG campaigns fire across performance social, ConnectedTV, and in-app video), and the revenue surface is fragmented (kirana, modern trade, quick commerce, direct-to-consumer, and traditional e-commerce each have different data stacks). The result is that most FMCG UGC campaigns get an ROI number attached after the fact that is a blended guess, presented as a point estimate.

This playbook is the framework vexo.club uses to produce ROI numbers on FMCG UGC launches that a brand's finance partner can defend in an annual review. It is specific to Indian FMCG with launch budgets between ₹50 lakh and ₹5 crore and assumes the brand has a working relationship with its e-commerce platforms (Amazon India, Flipkart, BigBasket, Blinkit, Zepto, Instamart) and at least one modern-trade chain.

What UGC is actually being measured against

The most common mistake in FMCG UGC measurement is comparing campaign performance against a zero baseline — as if the alternative was "no advertising at all." That comparison inflates UGC ROI by attributing to UGC every rupee the brand would have earned from whatever creative it would have run instead. The correct comparator is the brand's existing agency-produced creative on the same surface.

Concretely: if the UGC variants ran on Meta's Reels placement and drove ₹1.2 crore of attributed sales on ₹40 lakh of media, the ROI of UGC is not "₹1.2 crore attributed / ₹40 lakh spent = 3x." The ROI of UGC is "₹1.2 crore attributed on the UGC variants minus ₹X crore attributed to the agency variants running at the same time on the same surface with the same audience and spend, divided by the incremental media that went to UGC over agency." That number is the defensible figure. It is smaller than the gross ratio, often by a lot, and it is the number the brand's CFO will trust.

The creative-test matrix

Setting up the test is where most campaigns fail. A properly-designed UGC-vs-agency creative test has the following shape, committed to in writing before any shoot happens.

• At least two UGC variants. A single UGC variant that happens to underperform makes the category look bad; two variants produce a comparison and a floor.
• At least two agency variants. Same reason. The agency creative is the control; a sample size of one creates false winners.
• At least two audience cohorts. A launch that is going to female 25–40 urban is different from the same launch going to female 25–40 Tier-2. Test both; the UGC winner may differ by cohort.
• Equal spend allocation per variant. Meta's automatic allocation will kill the under-performing variants within hours and produce a skewed spend-share report. Force equal allocation across variants for the first 72 hours so every variant gets a readable sample.
• Minimum per-variant sample size before reading. 100 attributed conversions is the floor; 300 is comfortable. Reading earlier produces false winners from noise.

The matrix is simple to state and hard to enforce inside a brand's existing media agency. The request to force equal allocation usually produces pushback ("but we're leaving money on the table"). The pushback is wrong in the narrow case — equal allocation during the learning phase pays for itself in the subsequent two quarters of better creative direction. Every vexo.club FMCG UGC engagement has this rule in writing.

Joining pixel attribution to retail sell-through

FMCG revenue does not live in a single system. For a launch in India, the typical split is 30–50 percent modern trade (Reliance Retail, DMart, Spencer's), 20–30 percent e-commerce (Amazon, Flipkart), 10–20 percent quick commerce (Blinkit, Zepto, Instamart), and the remainder in kirana, direct-to-consumer, and other channels. Meta's pixel attribution captures the e-commerce and direct-to-consumer share; everything else is invisible to it.

The data-join that makes UGC ROI measurable on the full sales surface:

• E-commerce. Brand-share dashboards on Amazon Marketing Cloud, Flipkart Ads Console, and BigBasket Seller Central provide campaign-window sales data. The UGC variants' utm parameters tie back to these dashboards via the platforms' native attribution.
• Quick commerce. Blinkit, Zepto, and Instamart provide campaign-window SKU-level sell-through reports to brand accounts. The reports are daily; stitch them to the campaign's daily spend curve manually.
• Modern trade. Nielsen RMS or Kantar Worldpanel syndicated data arrives weekly-to-monthly and has a lag of four to six weeks. Use the pre-campaign four weeks as baseline, the campaign four weeks as test, and the post-campaign four weeks as halo.
• Kirana and general trade. The weakest data surface. The brand's regional sales team's weekly primary-sales numbers are the best available signal; pair them with a field ride-along once during the campaign to validate.

The stitched read is a weekly table where each row is a sales channel and each column is campaign week. The UGC ROI calculation uses the full table, not the pixel slice.

The incrementality read

Creative-test matrices answer "which variant won." Incrementality reads answer "what would have happened without this campaign at all." Both are needed; the first without the second inflates ROI by the amount of baseline brand demand that would have converted regardless.

Two incrementality methods work on Indian FMCG launches.

• Geo-holdout. Exclude a matched set of pincodes or cities from the campaign for the launch window. Typically 10–15 percent of the eligible geography, chosen to match on demographics, pre-campaign brand sell-through, and retail distribution. Compare sell-through delta in exposed vs holdout geographies. This is the gold standard for incrementality measurement in India and is the method Meta, Google, and Nielsen all endorse for FMCG.
• Audience-holdout (Meta-side). Meta's Conversion Lift studies exclude a random-sample test cell from campaign exposure and compare conversion rates in exposed vs excluded cells. Usable for brands that sell through Meta-attributable surfaces (primarily e-commerce); less useful for brands where modern trade dominates.

Geo-holdout is worth the operational complexity on launches over ₹2 crore; below that the study cost outweighs the incremental insight. For sub-₹2-crore launches, a pre-post comparison on syndicated retail data (four weeks pre, four weeks campaign, four weeks post) is the practical substitute.

Reporting — tier winners, not averages

The last place FMCG UGC measurement goes wrong is the report. "Our UGC campaign delivered 3.8x blended ROI" is a number that describes nothing specific and cannot be repeated. The honest reporting tier structure is:

• Top UGC variants. Which UGC variants beat the agency variants, on which audiences, by what margin. With confidence interval.
• Bottom UGC variants. Which UGC variants underperformed both agency variants and the brand's own organic conversion baseline. These are the expensive lesson and the most important report line for future briefs.
• Audience-specific reads. The winner on urban females 25–40 is often not the winner on Tier-2 females 18–35. Report both.
• Channel-specific reads. The UGC winner on Reels is not always the UGC winner on ConnectedTV. Report by surface.
• Incrementality summary. The holdout-adjusted revenue lift and the blended-attribution revenue lift, reported as two numbers not averaged.

The campaign's "ROI" number, if a single figure is mandated, is the incrementality-adjusted revenue minus the campaign spend, divided by the campaign spend. The un-adjusted number is reported in the same report but clearly labelled as the pre-incrementality attribution read.

What vexo.club bills for this

FMCG UGC measurement work breaks into three line items on a single engagement: test design (pre-campaign; matrix, sample sizes, incrementality method selection), measurement (pixel joins, retail sell-through stitch, holdout execution), and reporting (post-campaign; tier-by-tier creative read, incrementality summary, recommendations for next launch). Rate guidance is published at vexo.club/benchmarks quarterly.

How to use this playbook

If you run FMCG in-house with agency media, use this as a negotiation document with the media agency. The creative-test matrix is the single biggest leverage point; without it every subsequent measurement is downstream of a flawed test. If you want vexo.club to run the measurement layer alongside your existing media plan, or to audit an already-wrapped launch for hidden insights, start a brief.

Published 22 April 2026. Reviewed quarterly.