CPC vs CPI: Clicks or App Installs
In short: Both price a step in a mobile acquisition funnel, but CPC prices the tap on the ad and CPI prices the completed app install that sometimes follows it. CPI depends on app store friction, device compatibility, and install attribution from the ad network or measurement partner, none of which the ad auction controls. A campaign can hold a flat CPC while CPI swings, because everything between the click and the completed install sits outside the auction entirely. Use CPC to judge the ad's pull; use CPI to judge the full click-to-install path, and remember neither one tells you whether the person who installed ever opened the app again.
By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026
CPC
Total ad spend divided by clicks, giving the average price you paid for each click over the reporting period.
It tells you what the auction charged on average, not what a click was worth. A rising figure can mean stronger competition, weaker relevance, or simply a shift into more expensive placements and audiences. The common misreading is treating cheaper as better: low-priced clicks from poorly matched traffic routinely produce a worse cost per conversion than expensive ones.
Full definitionCPI
Ad spend divided by app installs, the average cost of one install attributed to your advertising by the store or measurement partner.
It prices the download and nothing after it. Installs are the cheapest event in a mobile funnel, so campaigns optimized toward this metric reliably find people who install once and never open the app again. Attribution is also contested between networks and SDKs. The common misreading is judging campaigns on it when retention and post-install revenue say the opposite.
Full definitionSide by side.
The differences that actually change what happens in your account.
| CPC | CPI | |
|---|---|---|
| What determines the number | Auction competition, ad relevance, bid strategy. | CPC plus app store conversion rate, device eligibility, and attribution accuracy from the store or SDK. |
| How it's attributed | Directly by the ad platform's click tracking. | By the app store or a third-party mobile measurement partner, which can disagree with the ad platform's own count. |
| What a rising number tells you | Auction got more competitive or ad relevance dropped. | Could be a worse auction feeding in, a worse app store listing, or attribution disputes - hard to tell from the number alone. |
| What it says about user quality | Nothing - a click carries no quality signal. | Nothing either - it prices a download, not retention or in-app revenue. |
| Typical optimization risk | Chasing cheap clicks from low-intent placements. | Chasing cheap installs from users who open the app once and never return. |
| Where it's measured | Every ad platform's core reporting. | App-install campaign reports, cross-checked against the app store's own install console when available. |
| Failure mode | Reading a cheap click as a win regardless of what happens after. | Judging a campaign a success on install volume while retention and post-install revenue tell the opposite story. |
What actually separates them.
CPC is fully determined inside the ad auction; CPI adds everything that happens between the click and a completed, attributed install.
Install attribution runs through the app store or a mobile measurement partner rather than the ad platform alone, so CPI can differ between what the ad platform reports and what the app's own install console shows.
Installs are the cheapest, easiest event in a mobile funnel to generate, so optimizing purely to CPI reliably finds users who install once and churn immediately.
CPC responds same-day to auction and creative changes; CPI additionally lags behind app store review times, download speed, and device compatibility issues outside your control.
Lowering CPC through broader placement generally lowers CPI too, since more clicks usually produce more installs, but it tends to raise the share of installs that never return, a problem invisible in either metric.
Which one should you use?
Use CPC when
- You are diagnosing whether the ad or auction is the source of a cost problem before the install step is even reached.
- The campaign doesn't yet have install tracking or SDK attribution wired up.
- You are comparing ad creative or targeting efficiency purely on click cost.
- You want a metric unaffected by app store listing quality or device compatibility issues.
Use CPI when
- You need to report the cost of one install to justify user-acquisition spend.
- You are testing app store listing changes, like screenshots, description, or icon, and want a metric that reflects the full click-to-install path.
- You are comparing UA campaigns across networks where install, not click, is the deliverable.
- You suspect the funnel problem sits after the click, in the store listing or device eligibility, and want a number that would move if that's the issue.
Common questions.
Why does my CPI differ between the ad platform's report and my app store console?
Because installs are typically attributed by a mobile measurement partner or the store itself, using its own attribution window and matching logic, which can disagree with the ad platform's self-reported install count. Treat the app store or measurement partner number as closer to ground truth and use the ad platform's figure mainly for same-day pacing.
Is a low CPI a good sign?
Not on its own. Installs are the easiest, cheapest event in a mobile funnel, so a low CPI can just mean the campaign is finding people who install and never open the app again. Pair it with a retention or post-install revenue metric before calling a UA campaign a success.
Why did my CPC stay flat while my CPI went up?
That points to a problem after the click - a slower app store conversion rate, a device compatibility issue, or a change in attribution matching - rather than a problem with the ad or auction. Check the app store listing and device eligibility settings before touching bids or creative.
Or stop choosing between them.
AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.