India CPI Benchmarks by App Category

The gap between the cheapest and dearest app category in India is roughly thirtyfold on median cost per install.

Which means any benchmark that does not tell you the category is close to useless. "The average app CPI in India is X" is a sentence that describes no actual app.

Here is what we can see across our own customer base, split by category, with the sample sizes and the method stated so you can judge how much weight to give it.

What does a mobile app install cost in India?

Cost per install is total advertising spend divided by the installs it produced. In India it varies by more than an order of magnitude across app categories, which makes a category-blind CPI benchmark close to meaningless.

Across roughly 4.7 million installs at 48 India-based apps running paid campaigns on Linkrunner between May and July 2026, with around Rs12 crore of connected ad spend:

• Median cost per install: just under Rs20
• Middle half of apps: roughly Rs8 to just under Rs40
• Top decile: above Rs180
• About one in eight apps: paying more than Rs100 per install

Methodology, stated up front. Projects included are those with a connected ad network account and INR as base currency, over the three months to 31 July 2026. Blended CPI is connected ad spend divided by total installs, which includes organic installs, so these figures understate paid-only CPI. Figures are medians across projects unless stated otherwise.

That last point matters and we will come back to it.

CPI by app category

Linkrunner does not tag projects by vertical, so we classified them by hand from app and company names. Of the 48 projects, 31 could be classified with reasonable confidence, covering 87 per cent of installs. Four categories had at least five projects and are reported below.

Three categories are missing because the sample was too thin to publish. Content and Entertainment, Health and Fitness, and Booking and Marketplace each had only two or three projects. Reporting a median from three apps would be a number with the appearance of authority and none of the substance. A further 17 projects could not be classified from the name alone and are excluded from this table, though they remain in the overall distribution above.

Commerce and D2C is the outlier by a wide margin, and it is worth sitting with why. These apps compete for the same users as the largest advertisers in the country, they sell considered purchases rather than free sign-ups, and as the activation column shows, fewer than half the people who install go on to do anything.

Why CPI alone will mislead you

Look again at Fintech and EdTech in that table.

Fintech has a lower median CPI than EdTech, roughly Rs9 against Rs22. On a CPI-only dashboard, fintech acquisition looks like the better deal by a factor of two.

Now look at the right-hand column. Cost per activated user runs around Rs75 for fintech and around Rs15 for edtech. The ranking inverts, and not marginally.

Cost per activated user is total ad spend divided by users who completed the app's defined activation event, rather than by users who merely installed. It is the number that says whether you bought customers or downloads.

The pattern holds across the whole sample. Splitting all 48 apps into quartiles by activation rate:

Weak activation means paying close to three times your CPI for a real user. Strong activation means paying about one and a fifth. Ten of the 48 apps sat below 50 per cent activation and held around a fifth of all the spend in the sample.

Your category sets a rough floor on what you pay. Your activation rate determines what you actually get for it, and unlike your category, it is something you can change.

How to use these numbers without misusing them

Treat them as bands, not targets. Six projects per category is enough to show a pattern and not enough to set a benchmark anyone should be held to. If your CPI sits outside these ranges, that is a prompt to investigate, not evidence of failure.

Remember the organic denominator. Because blended CPI divides connected ad spend by all installs, an app with strong organic growth will show a flatteringly low CPI. If you want to compare like for like against these figures, calculate yours the same way rather than using your paid-only number. Our post on how paid installs cannibalise organic growth covers why separating the two is harder than it sounds.

Do not compare activation rates across apps too literally. Every app defines its own activation event. A 45 per cent rate at an app counting completed purchases and an 85 per cent rate at an app counting signups are not measuring the same thing. Compare against your own history first.

Pair CPI with a payback view. Cost tells you half the story. Our benchmarks on install-to-first-revenue timing cover how long it takes for that spend to come back, which is the other half. The click-to-install timing benchmarks cover the top of the same funnel.

What to do if you are above the band

A high CPI is not automatically a problem, and diagnosing it as one is how teams cut their best campaigns. Work through this order:

1. Check what is in your denominator. If you have recently lost organic volume, your blended CPI rises without anything changing in your campaigns.
2. Check activation before cost. If your CPI is high and your activation is high, you may simply be buying expensive good traffic, which is often correct. If both are poor, that is the real problem and cost is a symptom.
3. Check your category expectation. Commerce and D2C apps paying Rs150 are inside the normal range. Social apps paying Rs150 are not.
4. Check the auction, not just the account. Costs move seasonally in India, and the festive window from late September through Diwali repricing everything is the clearest example.
5. Then look at campaign structure and creative. Our guide to reducing CAC without cutting quality and the levers for optimising CPI both start here, which is why they are step five rather than step one.

Working through this needs installs, activation events and spend joined by campaign in one place. If you want to see where your app sits against these bands, request a demo and we will run the comparison on your data. The CPI benchmark explorer is a faster starting point if you want a rough read first.

FAQ

What is a good CPI for a mobile app in India?

Across 48 India apps we measured between May and July 2026, the median was just under Rs20 with the middle half between roughly Rs8 and Rs40. But category matters enormously: social apps sat around Rs8 while commerce and D2C apps sat around Rs235. A single national average is not a useful target for any individual app.

Which app category has the highest cost per install?

In our sample, commerce and D2C by a wide margin, at a median of roughly Rs235 against Rs8 to Rs22 for social, fintech and edtech. These apps compete against the largest advertisers in the country for the same users and sell considered purchases rather than free sign-ups.

Why is commerce app CPI so much higher than social app CPI?

Auction competition and intent. Commerce apps bid against marketplaces with very large budgets, and they are asking for a purchase rather than a sign-up. Their activation rate was also the lowest in our sample at around 45 per cent, so the real gap in cost per customer is wider still.

Does CPI include organic installs?

In these figures, yes. Blended CPI divides connected ad spend by all installs, attributed and organic. This understates paid-only CPI, and it means apps with strong organic growth will look cheaper than their paid campaigns actually are. Calculate yours the same way if you want to compare against these bands.

Is a low CPI always better?

No. Ad networks optimise towards the event you give them, so optimising for installs finds people most likely to tap install and least likely to commit further. A falling CPI and a falling activation rate are frequently the same change described two ways. Judge campaigns on cost per activated user and use CPI to diagnose the auction.