How Advertising Performance Metrics Differ Across Web, App, and CTV   

Think about three different rooms in a house: the kitchen, the garage, and the living room. You'd use a different tool to check if each one is "clean." A kitchen gets checked for crumbs on the counter. A garage gets checked for whether the car actually fits. A living room gets checked for whether it's comfortable to sit in. Same house, three completely different checklists.

Web, mobile app, and CTV (the ads you see on your smart TV) all sell ad impressions, but they're measured with different rulers, because the "room" is different every time.

We laid out the core metrics every publisher should know in our guide on what advertising performance actually is. This post picks up from there and answers a more specific question: which metrics matter most on web, which ones change shape inside an app, and which ones only really exist because of CTV.

Web: the original battlefield, and the noisiest one   

Web display advertising is the oldest and most crowded of the three, and its metrics reflect that. The two numbers that matter most here are viewability and fill rate.

Viewability asks a simple question: did a human actually have a chance to see this ad? The industry standard is 50% of the ad's pixels visible for at least one second for display, or two seconds for video. This matters so much on web because a huge number of ad slots sit below the fold, where nobody scrolls, or load in tabs nobody's looking at.

Fill rate matters just as much, because web display inventory is enormous and spread across a chaotic mix of exchanges, SSPs, and waterfalls. A typical healthy web setup lands somewhere around 70 to 85 percent fill. Below that, you're likely leaving real money on the table from unfilled requests.

CTR (click-through rate) still shows up on web, but it's genuinely small. Standard display banners often sit under 0.5 percent. That's normal, not a red flag, because most web display isn't a direct-response format, it's a branding and reach format.

 Web, in short: 

• Viewability is the make-or-break metric, since so many slots sit unseen below the fold

• Fill rate should sit around 70–85% on a healthy setup

• CTR is naturally low (under 0.5% is normal), so don't panic over it

• Inventory is fragmented across many exchanges, which is exactly why fill rate swings so much

App: fewer eyeballs wasted, but a different kind of trust problem   

Inside a mobile app, the whole measurement picture shifts. There's no scrolling past an ad below the fold the way there is on a webpage, so viewability is less of a battle. Instead, the two metrics that take center stage are fill rate by format and completion rate for video.

Format matters enormously in-app, more than almost anywhere else. Rewarded video, where a user chooses to watch an ad in exchange for something (extra lives, in-app currency, unlocked content), regularly clears 90 percent fill and commands some of the highest eCPMs of any format, because the user opted in and completion rates are naturally high. Interstitials, which appear between app screens, sit in the middle. Banner ads, wedged at the top or bottom of a screen, sit at the bottom of both fill and eCPM, simply because there's less real estate and less user attention on them.

There's also an attribution layer that web doesn't really deal with. Apps rely on mobile measurement partners and device-level identifiers rather than browser cookies, so "did this ad lead to an install or an action" gets tracked through an entirely different pipe than "did this ad lead to a website visit." If you're used to thinking in web-style conversion tracking, this is the part that trips people up first.

 App, in short: 

• Format is king: rewarded video > interstitials > banners, in both fill rate and eCPM

• Rewarded video often clears 90%+ fill because the user opted in

• Viewability matters less here — there's no "below the fold" inside an app screen

• Attribution runs through mobile measurement partners and device IDs, not cookies

CTV: unskippable, full-screen, and measured on a completely different curve   

CTV is the newest of the three, and it borrows almost nothing from web or app measurement. Because most CTV ads are full-screen and unskippable, viewability is often treated as close to 100 percent by definition, which makes it a much less useful signal than it is on web.

Instead, the star metric on CTV is video completion rate, or VCR: the percentage of viewers who watch the ad all the way through. Because skipping usually isn't an option, completion rates on CTV run high compared to other video formats, with anything above roughly 95 percent considered strong engagement, and anything under 85 percent usually pointing at creative fatigue or a mismatched audience.

Frequency also matters more on CTV than almost anywhere else, because it's measured at the household level rather than the individual level. The same ad hitting the same living room too many times in one evening burns budget fast, so frequency capping gets tracked closely.

And CTV pricing reflects all of this. eCPMs on CTV routinely run several times higher than open web display, because advertisers are effectively buying TV-style attention with digital-style targeting and reporting attached.

 CTV, in short: 

• Viewability is almost always near 100%, so it stops being a useful diagnostic

• Video completion rate (VCR) is the real signal: 95%+ is strong, under 85% flags a problem

• Frequency is tracked per household, not per person, since one TV serves a whole room

• eCPMs run several times higher than web, because it's TV-level attention with digital tracking

Putting the three side by side   

If you only remember one thing from this post, make it this: viewability is the web metric, format-driven fill and completion rate are the app metrics, and household-level completion and frequency are the CTV metrics. They're not competing versions of the same idea, they're genuinely different questions being asked of three different environments.

That also means you can't manage all three the same way, and most publishers don't have the bandwidth to build a separate optimization playbook for each one. This is exactly the gap MagicBid was built around: dedicated Web, In-App, and CTV Monetization products, each built around that environment's own rules, instead of one generic setup stretched thin across all three.

A lot of that complexity gets absorbed by MagicBid's AI-driven ad tag, which reads over 100 real-time signals — including user behaviour, advertiser demand, and page speed — to route each impression to the right buyer automatically.

 FAQ   

1. Why is viewability less useful on CTV than on web?

Because most CTV ads run full-screen and can't be skipped, they're already close to 100 percent viewable by definition. On web, where ads sit below the fold or in background tabs, viewability is doing real diagnostic work.

2. What's the most important app metric to watch?

Fill rate and eCPM broken out by ad format, since rewarded video, interstitials, and banners behave completely differently from each other, even on the exact same app.

3. Is a high CTV eCPM automatically better than a high web eCPM?

Not automatically, but it often reflects reality: CTV draws from bigger, brand-focused TV budgets and delivers stronger completion and attention, so the premium is usually earned rather than inflated.

4. Does CTR still matter on CTV?

If you’re not making the most of your ad space, you’re leaving money on the table. MagicBid helps web, app, and CTV publishers maximize revenue with smarter ad placement and optimization tools.

• Web Monetization: Get better ad visibility, higher engagement, and more revenue from every impression.

• In-App Monetization: Connect with premium advertisers to effortlessly boost fill rates and eCPMs.

• CTV Monetization: Deliver high-quality, tailored ad experiences that keep viewers engaged and advertisers paying more.

With MagicBid’s advanced ad tech and expert support, you can turn your traffic into higher earnings without the guesswork. Connect with us now to get a free ad revenue evaluation.