How to Measure Ad Creative Performance in India (2026 Benchmarks)

Short answer: Judge creatives in sequence — thumb-stop ratio, then hold rate, then CTR, then CPA, then ROAS net of RTO. You need 20–30 conversions per creative before a scaling decision, and 95% confidence before you trust it.

The winner that lost money#

A creative comes back with a 3% CTR. Forty conversions. ₹80,000 revenue on ₹25,000 spend — a 3.2x ROAS. Everyone agrees it is the winner and it gets scaled.

Then the RTO data lands. Nearly half those orders came back, and the real ROAS after RTO falls under the brand's break-even.

The creative did not win. It bought cheap clicks from an audience that does not pay, and every rupee poured into scaling it made the loss bigger.

This is the specific way Indian D2C creative testing goes wrong, and no front-end metric will warn you. Measurement has to run all the way to delivered purchases or it is measuring the wrong thing.

Which metrics decide a winner?#

Read them in this order:

1. Thumb-stop ratio / hook rate — does it stop the scroll in three seconds?
2. Hold rate — does it keep attention through the body?
3. CTR — does it earn the click?
4. CPA / CPP — does the traffic convert?
5. ROAS — does it clear profitability after returns and RTO?

Front-end or back-end metrics?#

Both, for different jobs:

• Front-end engagement (TSR, hold rate, CTR) tells you whether the creative works as an ad
• Back-end outcomes (CPA, ROAS, conversion rate) tell you whether it works as a business driver

Practically: use TSR and hold rate to filter weak hooks early, CTR to judge message resonance, and CPA and ROAS to decide scaling or killing. Conversion rate matters but is heavily influenced by landing page and offer — it is not a clean read on the creative.

Judging video creatives#

These are more diagnostic than CTR alone for video, because they show whether the story holds attention rather than just whether the thumbnail worked.

Engagement or delivered purchases?#

Front-end engagement is for diagnosis and iteration. Delivered purchases, net of RTO, are for go/no-go and scaling.

A creative with strong CTR but poor post-click conversion or high RTO is not a winner in India.

How much data before you call it?#

Impressions at Indian CPMs#

At typical Indian CPMs of ₹80–₹140:

• 10,000–20,000 impressions per creative for stable TSR, hold rate and CTR
• 3,000–5,000 impressions gives directional signal only — not scaling confidence

Conversions needed#

• 20–30 conversions per creative before a scaling decision
• 10–15 may be acceptable for high-AOV categories above ₹2,000, if CPA and ROAS are clearly in line
• Below that, results are noise

Statistical confidence#

• 95% is the standard for scaling decisions
• 90% is enough for directional signals — "this angle looks promising, test more variations"

In practice this means enough conversions that the observed lift is unlikely to be random.

Kill and keep rules#

CTR benchmarks by category#

Kill or pause when:

• CTR is below 50% of category benchmark after 10,000+ impressions and reasonable spend
• CTR is acceptable but TSR is under 25% — the hook is weak, iterate the first three seconds

ROAS thresholds by category#

Blended, post-RTO:

Kill when ROAS sits below your break-even — calculated on contribution margin after RTO and returns — across 3–5 days and 20+ conversions. If ROAS is more than 20% below target and CPA is above target, pause and rework the offer or creative.

High CTR, high RTO#

A creative can post an excellent CTR and still be unsustainable once delivery is counted. Work the gap between reported and delivered ROAS through our real ROAS after RTO guide before you decide.

Do not scale it. Use it as a diagnostic instead: high RTO with high CTR usually means the wrong audience, a misleading promise, or a COD-heavy segment. Fix targeting, offer or messaging before touching the budget.

The diagnostic funnel#

Read metrics in sequence and each failure points at a specific fix:

Setting up the system#

1. Testing cadence#

• 3–5 new concepts per week
• Each concept: 2–4 hook variations × 1–2 formats
• Use ABO initially to force spend across creatives, then shift winners to CBO

2. Kill and scale rules, written down#

Pause if:

• TSR under 25% after 10,000 impressions
• CTR under 50% of category benchmark after 10,000 impressions
• CPA above 1.3× target once the test budget is spent — do not wait for 20 conversions
• ROAS below break-even after 20 conversions

Scale if:

• TSR at 30–40%+ and CTR at or above category benchmark
• CPA at or under target, ROAS at or above target, for 3–5 days
• 20–30+ conversions at 95% confidence against control

The budget mechanics of that scale step — how much to raise, how often — sit in the test → learn → scale loop.

3. Track creative win rate#

Creative win rate = winners ÷ creatives tested

Measure yours against the India D2C creative win-rate benchmark. Log hook type, format, angle and language on every test — that log is what moves the number.

Worked example: evaluating a video creative#

Impressions:   20,000
3s plays:      7,000
15s plays:     3,000
Link clicks:   400
Conversions:   25
Spend:         ₹20,000
Revenue:       ₹60,000

Hook rate = 7,000 ÷ 20,000  = 35%   → Good (above 30%)
Hold rate = 3,000 ÷ 7,000   = 43%   → Good (40–50% band)
CTR       = 400 ÷ 20,000    = 2%    → Good for most D2C categories
CPA       = ₹20,000 ÷ 25    = ₹800
ROAS      = ₹60,000 ÷ ₹20,000 = 3x

With a 2.5x break-even ROAS and a ₹900 target CPA, this is a winner and can be scaled.

India benchmarks summary#

FAQ#

Which metric matters most when judging an ad creative in India? None alone — read them in sequence. Thumb-stop ratio filters weak hooks, hold rate catches pacing problems, CTR judges the message, and CPA and ROAS net of RTO decide whether to scale.

What is a good hook rate for Indian D2C video ads? Thirty to forty-five percent for UGC-style video, 20–30% for studio. Below 25% the hook is the problem, regardless of how the rest of the video performs.

How many conversions before scaling a creative? Twenty to thirty per creative. High-AOV categories above ₹2,000 can act on 10–15 if CPA and ROAS are clearly in line. Anything less is noise.

How many impressions does a creative need before you judge it? Ten to twenty thousand for stable TSR, hold rate and CTR at Indian CPMs of ₹80–₹140. Three to five thousand gives direction only.

When should you kill an ad for low CTR? When CTR falls below half your category benchmark after 10,000+ impressions. Beauty UGC benchmarks at 1.6–2.8%, so kill below roughly 0.8–1.4%.

Should you keep a creative with strong CTR but high RTO? Not as a scaler. Once RTO is netted off, a strong gross ROAS can land below break-even. Treat it as a signal about audience or promise, and fix targeting or messaging first.

What ROAS should an Indian D2C brand target? Post-RTO: 2.2–3.5x beauty and skincare, 1.8–3x apparel, 2–3.2x food and beverage, 2.5–4x wellness, 3x+ for high-ticket home and electronics.

What is a good creative win rate? Log winners ÷ creatives tested every month and compare it against the India D2C benchmark. Brands running a structured testing system with a logged creative library land well above the average.

---

About the author#

Rohit Sharma is a performance marketer at Profitbox360, specializing in paid media, D2C growth, creative strategy, and conversion optimization. They write about practical marketing frameworks, campaign benchmarks, and strategies for scaling customer acquisition.