A skincare brand running Instagram ads at Rs.480 CPA asked us to audit their creative before a budget scale-up. The ads were polished — clean studio shots, branded fonts, voiceover. They were also performing exactly like polished ads in a saturated category: scroll-past rates above 70%, CTRs under 0.6%. What followed was a nine-week UGC production sprint that ended with their CPA sitting at Rs.240. This is the mechanics of how that happened.
We are not going to attribute this result to "authenticity" as an abstract quality. The improvement was traceable to specific production and briefing decisions — some obvious in hindsight, some counterintuitive. Understanding those decisions is more useful than celebrating the number.
Starting Point: The Creative Audit
Before any filming, we ran a structured creative audit against the brand's existing Meta ad library. This is standard in our process and routinely surfaces the same pattern: brands have invested heavily in assets that look like ads and are treated as ads by the algorithm, because they are indistinguishable in structure from the competition's ads.
For this brand — a direct-to-consumer skincare company shipping primarily to metro India (Bengaluru, Delhi, Mumbai, Hyderabad) — the audit revealed three compounding issues:
- No local language variation. All creatives were in English only, despite Kannada and Hindi being the dominant home languages in their two highest-revenue cities.
- Hook failure at 0–2 seconds. Every ad opened with the product logo or pack shot. Scroll-stop data showed abandonment before the value message appeared.
- Absence of social proof formats. ASCI guidelines require testimonials to be genuine and verifiable — but nothing in their library used real customer voice at all. They had been over-correcting on compliance to the point of removing all testimonial structure, which left the creative feeling impersonal.
The audit output becomes the brief foundation. Every creator we cast is briefed against it specifically, not against a generic "talk about the product" template.
Creator Casting for a Specific Conversion Problem
The casting brief for this project was narrow. We were not looking for creators with the largest following — we were looking for creators whose own audiences resembled the brand's existing buyer persona: women aged 24–35 in Tier 1 cities, with demonstrated interest in skincare routine content, not just beauty at large.
We shortlisted eight creators from our network across Bengaluru, Mumbai, and Pune — all with Instagram followings in the 8,000–45,000 range. Micro-creators in this range typically have engagement rates 3–4 times higher than accounts above 500K, and their content style is closer to genuine recommendation than performance. For a skin-care category where purchase hesitancy is high (people are cautious about what goes on their face), that peer-recommendation register matters for conversion.
Each creator received a detailed written brief covering:
- The specific skin concern they were being asked to address (hyperpigmentation from sun exposure, which is the top concern among the brand's existing buyers in South India)
- Mandated disclosures per ASCI's Guidelines for Celebrities and Influencers — every paid collaboration video had to carry a visible #Ad or #Paid disclosure, and we built this into the brief as non-negotiable, not as a post-production afterthought
- Three "forbidden claims" the brand's legal team had flagged — no "100% results guaranteed" type language, in line with ASCI Rules 1 and 2 on unsubstantiated efficacy claims
- A suggested hook structure (problem statement in first 3 seconds, not brand name)
The Production Formats We Prioritised
We produced three distinct format types across the eight creators. Volume across formats is deliberate — it generates enough variation to run real creative tests rather than anecdotal comparisons.
Format 1: The 30-second problem-first Reel. Creator opens with a relatable skin problem shown on camera (no filter, no studio lighting), transitions to the product introduction, ends with a visual result. Filmed on the creator's phone at home. No branded end card in the first iteration — we wanted to measure whether the unbranded look affected performance (it didn't hurt; it helped).
Format 2: The routine integration video. A 45–60 second video showing the product as part of an existing morning or evening skincare routine. This format performs particularly well in Hindi because routine content in Hindi has strong organic pull on YouTube Shorts and Instagram Reels in cities like Lucknow, Jaipur, and Bhopal — cities the brand had not previously targeted but which showed up as latent demand in their Shopify analytics.
Format 3: The reaction / comparison video. Two creators who had been using the product for 4+ weeks filmed a candid "before vs. now" piece. ASCI compliance here was explicit: the comparison was against their own prior skin state, not against competitor products (comparative advertising rules under ASCI Chapter IV require claims to be verifiable, and "my skin four weeks ago" is verifiable; "better than Brand X" requires documented proof).
The brief told creators: "Film like you are showing your sister, not advertising to a stranger." That single instruction consistently produces a different body language and eye contact pattern than standard influencer performance mode.
How the Ad Account Was Structured Around the UGC
Good UGC briefed and produced badly into the ad account still underperforms. The media side of this project was structured as follows on Meta:
- Separate ad sets by format type, not by creator. This lets you learn whether "routine integration" outperforms "problem-first" as a format signal, rather than crediting one creator with what is actually a format win.
- Broad targeting with Advantage+ audience as the primary learning structure. We did not layer in interest targeting on top of UGC creative — the algorithm's look-alike modelling is strong enough that adding manual interests constrains delivery without improving CPA at this budget level (Rs.40,000–Rs.60,000 per month).
- Language-matched creative to placements. Hindi-language videos ran with Hindi primary text copy; Kannada-language videos were tagged with geo-priming toward Karnataka. Meta does not offer language-level targeting directly, but geo + language creative alignment reduces creative-audience mismatch friction.
- First-party data custom audiences excluded from prospecting ad sets from day one, so CPA figures reflected genuine new-customer acquisition cost, not retargeting economics.
The Rs.240 CPA result was measured on prospecting only — new customers, not existing buyers retargeted. That distinction matters for any brand benchmarking against this figure.
What Did Not Work (and Why It Is Worth Knowing)
Two of the eight creators underperformed significantly. In both cases the pattern was the same: the creator deviated from the brief's hook structure and opened with a brand shoutout instead of the problem statement. Their content was not bad — it performed well as organic posts — but as paid ads entering an interrupted feed moment, the branded opening gave the viewer no reason to stop scrolling.
We also tested one creator-led Stories format with a swipe-up link. Stories CPA came in 40% higher than Reels CPA across equivalent spend. This is consistent with what we see across most D2C categories in India currently — Reels (both on Instagram and repurposed to YouTube Shorts) consistently outperform Stories for new customer acquisition because the full-screen immersive format combined with audio-on defaults gives UGC content more time to establish the problem and the solution before the viewer exits.
The learning we retained: brief compliance on the hook is not optional. Future briefs now include a short video example of a compliant and a non-compliant hook, not just a text description.
The Numbers in Full Context
The headline result — CPA moving from Rs.480 to Rs.240 — happened over a 9-week window. Weeks 1–3 were production and launch. Weeks 4–6 were the learning phase where we held budget steady and resisted optimisation pressure. Weeks 7–9 were where the winning creative set was identified and budget was shifted toward it.
The total UGC production investment was approximately Rs.1.8 lakh (creator fees, brief development, and editing across all eight creators and three format types). Against a media budget of roughly Rs.1.8 lakh across the test window, the effective cost per incremental CPA improvement was well within the payback window for a brand with a 30–40% gross margin on an Rs.1,400 average order value.
The more durable outcome: the brand now has 24 UGC assets that continue running in rotation. Creative fatigue is managed by rotating new creator faces and updated seasonal briefs rather than by spinning up fresh studio production each month. The ongoing production cost per month is now roughly Rs.35,000–Rs.45,000 for refresh volume, a fraction of what equivalent studio production would cost.
If you are working through a similar CPA problem — or want to understand what a production sprint for your category would realistically look like — our team walks through this during a free consultation call. We cover brief structure, creator casting criteria, and ad account setup specific to your product and margins.