Measuring UGC Campaign Success Beyond Views

The most painful meeting I've ever sat in was a D2C board review where the marketing lead walked in with a deck full of Reel view counts, and the CFO walked out asking, very politely, whether influencer marketing was a budget line or a hobby. That meeting ended a program that was, in hindsight, actually working — the team just couldn't prove it in a language the finance side respected.

UGC works. It also gets killed in review cycles because the measurement stack most brands use is designed for brand-building, not for defending a quarterly spend. This post is the four-metric framework we use at Vireelu — and that I've watched multiple brands use to successfully defend (and grow) their creator budgets.

Why views are the wrong starting point

Views measure reach that happened. They do not measure reach that mattered, audience that fit, or revenue that moved. A 2M-view Reel on a 9-year-old's account with a beauty brand seed is worth approximately zero. A 40K-view Reel on a 25-year-old dermatologist's account is worth a great deal.

The deeper problem: views are an Instagram metric, and your CFO doesn't care about Instagram. Your CFO cares about whether the rupee you spent showed up on the P&L. The four metrics below bridge the two.

Metric 1 — CPM (cost per mille)

CPM is the floor. It's the "did this reach anyone at all" metric. Calculate:

CPM = (total campaign spend) / (total impressions across all posted content) × 1000

For seeding, include product COGS + shipping + platform fee in the numerator. A healthy seeded CPM in our 40-campaign cohort runs ₹45–₹120 for mid-tier creators, versus ₹180–₹400 for paid deals at comparable reach.

The CPM conversation is the cheapest one to have with a CFO because it's familiar — it's the same metric Meta and YouTube report. Anchoring the discussion here before you bring in the UGC-specific metrics gives you a shared vocabulary.

Metric 2 — CPE (cost per engagement)

CPM tells you it reached people. CPE tells you those people cared. Calculate:

CPE = (total campaign spend) / (total likes + comments + shares + saves)

Saves are the metric we weight most heavily internally. A save is a signal that the viewer intends to come back — for a commerce brand, it is the single strongest purchase-intent micro-behavior Instagram exposes.

From our cohort, healthy CPE for seeded campaigns sits at ₹8–₹22 per engagement, with the best-performing beauty and food categories dipping under ₹10. For comparison, the same brand's paid ads against a cold audience routinely land CPE in the ₹35–₹80 range.

Metric 3 — Attributed GMV

This is the metric your CFO actually wants to see, and it is also the hardest to get right. Two approaches:

Approach A — Promo codes (simple, under-counts)

Give each creator a unique discount code. Measure GMV through that code. This works, it's clean, and it dramatically under-counts because the majority of people who watch a UGC post and buy do not remember to use the code. Our internal benchmark: promo-code attribution captures roughly 22–34% of true UGC-driven GMV.

Approach B — Post-purchase survey (better, still imperfect)

Add a "how did you hear about us" dropdown on the order confirmation page. Offer a small incentive (free sample on next order) for honest answers. This captures roughly 55–72% of true attribution in our cohort, because the response is collected at the moment of highest recall — three seconds after purchase.

Most mature D2C brands we work with run both. The promo-code number is your hard floor. The survey number is your defensible directional truth. When you present to finance, present the range, not a single figure.

Attribution in UGC is never a single number. It is a range, bounded by a pessimistic and optimistic model. Presenting UGC ROI as a single-point estimate is the credibility kill; presenting it as a tight range is what makes finance trust you.

Metric 4 — Retention lift

The fourth metric is the one nobody runs and the one that almost always makes the seeding spend sing. Look at the 90-day repeat-purchase rate of customers who first bought through a UGC-attributed channel (via promo code or survey) vs. customers who first bought through paid performance ads in the same month.

Across our cohort, UGC-acquired customers repeat at 1.4× to 2.1× the rate of performance-ad-acquired customers over 90 days, for comparable AOV. This is the number that, once surfaced, usually shuts down the "UGC is top-of-funnel fluff" debate permanently — because it shows the LTV implication.

The mechanism is probably self-evident: a customer who bought because a creator they trust vouched for the product starts the relationship with higher belief, gives the product more benefit-of-the-doubt on first use, and is more likely to come back.

Putting the four metrics in one dashboard

The way we've seen this land best is one weekly dashboard per campaign with exactly four rows:

| Metric | Target | This week | Trend | | --- | --- | --- | --- | | CPM | ≤ ₹120 | — | — | | CPE | ≤ ₹22 | — | — | | Attributed GMV (range) | ≥ 1.5× spend | — | — | | 90-day retention lift | ≥ 1.4× vs paid | — | — |

No views column. No engagement-rate column. Those are inputs to the metrics above, not metrics in their own right. Keep the dashboard boring on purpose. A boring dashboard is a dashboard the CFO will actually read.

The metric we deliberately do not track

Every quarter, at least one brand asks us to track "brand sentiment" or "share of voice." We push back. Both metrics are expensive to track well, impossible to track cheaply, and rarely correlate with revenue in the D2C context. Leave them to the CPG giants. For D2C brands from ₹2Cr to ₹200Cr ARR, the four metrics above are genuinely sufficient.

What this means for your next campaign

If your current UGC reporting is a screenshot deck of "best Reels this month," you are one skeptical CFO away from losing the budget line. Move to the four-metric dashboard in the next 30 days, even imperfectly. The act of writing the numbers down — even with attribution uncertainty baked in — raises the credibility of the program by an order of magnitude. Views are a story. These four are a case.