Web Rewarded Video Ad Performance in 2026: The Mid-Year Verdict
A first-half review of eCPM, opt-in, and retention — and why signal loss made rewarded video the most privacy-durable revenue line in gaming.
TL;DR — The Verdict
Through the first half of 2026, rewarded video didn’t just hold its ground — it widened its lead. As signal loss repriced the rest of the ad stack, opt-in, consented inventory stayed the format advertisers trusted and kept paying premium rates for. Published tier-1 benchmarks still put US rewarded video eCPMs near $16–$20, ahead of interstitials, while the year’s real story played out underneath the averages: privacy-clean, opt-in formats gained at the expense of passive, targeted ones. The mid-year verdict — rewarded video is now the most privacy-durable revenue line in gaming, and the web is where its next chapter is being written.
Six months of data beats six months of predictions
When we published our State of Web Monetization: 2026 Outlook last December, most of what the industry was saying about 2026 was still a directional bet. The prediction was blunt: 2026 would be a year of structural reckoning, with two forces pulling web monetization in opposite directions — AI making advertising smarter and more personalized, and a fragmenting privacy regime dismantling the third-party-data model that display advertising was built on. The winners, we argued, would be the formats that could earn without surveillance.
We now have half a year of real numbers to check that against. And the short version is that the reckoning arrived more or less on schedule — but the clearest signal isn’t in the mobile duopoly’s headline eCPMs. It’s on the web, where HTML5 and WebGL games sit outside many of mobile’s assumptions and closest to the privacy shift reshaping everything else. That’s where this mid-year verdict spends most of its attention.
This is the sequel to that outlook post: same lens, but with data instead of forecasts. We’ll look at what the published benchmarks actually say, why signal loss quietly repriced the entire stack in rewarded video’s favor, where the web became the growth edge, and how the format scores on four axes at the halfway mark.
What the numbers actually say
Rewarded video is still the highest-paying line in most gaming monetization stacks, and the first half of 2026 didn’t change that. Recent tier-1 benchmarks put US rewarded video eCPMs at roughly $16.49 on Android and $19.63 on iOS, comfortably ahead of interstitials, which sit closer to $14 on both platforms. Widen the lens and the tier-1 range runs from about $15 to $40, with global averages landing lower and regional gaps of four to ten times between top markets and emerging ones.
The demand side of that premium is behavior. Rewarded video is voluntary, so completion rates stay high — often quoted above 95% — and players consistently say they prefer it. Multiple 2026 studies still put the preference for rewarded ads over interstitials at roughly four to one, with around nine in ten players engaging with rewarded formats. Advertisers pay up because the audience opted in and watched to the end. None of that eroded in H1 2026; if anything, it hardened.
It helps to be precise about why advertisers pay the premium, because that mechanism is what makes the number durable rather than seasonal. A rewarded impression is a fully-watched, opted-in view of a message the user chose to see in exchange for something they wanted. That combination — high completion, genuine attention, and self-selection toward interested users — is scarce everywhere else in the stack. Interstitials interrupt; banners are ignored; rewarded video is the one placement where the user’s incentive and the advertiser’s incentive point the same direction. Half a year of 2026 data did nothing to weaken that alignment, which is why the eCPM floor stayed where it was even as the rest of the market wobbled.
AppLixir Stats
Those are the public, cross-industry figures — useful for context, but they’re mobile-weighted and not ours. Here’s where AppLixir’s own web and HTML5 network data belongs, side by side with the published numbers, so the comparison is concrete rather than borrowed:
Network-wide rewarded video eCPM (web / HTML5): [3.62 Global, 6.20 Tier-1 Geos (US,UK,CA,AU,AZ) – US $6.98]
Average opt-in rate: [97%]
Average completion rate: [93.8% global · 95.4% tier-1]
Average fill rate: [95.1 %]
Dropped next to the $16–$20 mobile benchmark, even one of these — a web eCPM and a completion rate the industry can’t source anywhere else — is what turns this from a roundup into a verdict readers cite.
Rewarded video vs. the rest of the stack
How the three workhorse formats compare at mid-year, on the dimensions that decide what a publisher actually keeps:
Dimension
Rewarded video
Interstitial
Banner
Tier-1 eCPM (US)
~$4–$15
~$6
~$0.50–$1.50
User action
Opt-in, voluntary
Forced, full-screen
Passive, always-on
Completion
95%+ typical
Skips common
N/A
Retention effect
Positive
Neutral to negative
Neutral
Privacy exposure
Low (consented)
Higher
Higher
Public benchmark ranges shown for context; figures vary by region, platform, and mediation setup.
The story of H1 2026: signal loss repriced everything
If you only read the averages, 2026 looks like a quiet year — rewarded video steady, interstitials steady, banners still cheap. The interesting movement happened beneath the averages, and it was driven by one thing: the ad stack ran short on identity signal.
Between Apple’s App Tracking Transparency and the rollout of Android’s Privacy Sandbox, a growing share of traffic now reaches the auction without the identifiers that targeting used to depend on. The market repriced that traffic accordingly. Unconsented impressions are earning roughly 20–40% lower eCPMs than consented ones, because advertisers can’t target or measure them as precisely. That gap is the real headline of the first half — not the top-line eCPM number, but the widening spread between consented and unconsented inventory.
Rewarded video sits on the right side of that spread almost by definition. It’s an opt-in format: the player taps a button, agrees to the value exchange, and watches. When the industry rebuilt its consent flows around that same logic — explaining plainly that ads keep the game free — opt-in rates climbed from around 35% to over 60%. Consent stopped being a compliance checkbox and became a revenue lever, and rewarded video was already shaped like the answer.
It’s worth being clear about why the other formats absorbed the damage instead. Banners and interstitials lean heavily on behavioral targeting to justify their rates; strip out the identifier and the advertiser is buying a much blurrier audience, so the bid falls. Measurement compounds the problem — without reliable attribution, performance advertisers can’t prove the impression worked, and unprovable inventory gets discounted fast. Rewarded video short-circuits both issues. The value exchange is explicit, the completion is near-total, and the consent is captured in the moment, so the impression stays legible to advertisers even when the surrounding identity graph goes dark. The format didn’t get more valuable in absolute terms so much as everything around it got cheaper, and the relative gap is what publishers felt in their reports.
Signal loss didn’t hurt rewarded video — it handed it a moat. When targeting gets scarce, the format that already has the user’s yes is the one advertisers keep buying.
This is also where AppLixir’s privacy-first posture stops being a tagline and starts being measurable. If the year rewarded consented inventory, then a network built opt-in and TCF 2.3 / GDPR compliant from the ground up should show it in the numbers. The proof points worth surfacing here:
Opt-in lift after a value-exchange consent flow, before vs. after: [97% opt in with 94% completion rate]
Share of network traffic that is consented / TCF-compliant: [97%]
eCPM premium on consented vs. unconsented impressions on the network: [Consented eCPMs perform twice than unconsented ones]
Those three figures would let this section make a claim no benchmark aggregator can: that on a privacy-clean web network, the consented-inventory premium is not theoretical.
Web and HTML5: where the growth actually moved
The mobile app market is mature and consolidated; its rewarded video story in 2026 is mostly about optimization at the margins. The web is where the format is still opening new ground, for three structural reasons.
Rendered ads that adblockers can’t reach
WebGL and in-canvas ad formats are rendered as part of the game scene rather than served through the ad slots blockers watch for. That gives web games a structural delivery advantage: the impression lands even when a browser extension would have killed a conventional banner. For rewarded video specifically, an opt-in view rendered inside the WebGL context is both harder to block and cleaner to measure — an edge native apps don’t need and web publishers increasingly rely on.
The hybrid model settled into the standard
The debate about whether ads cannibalize in-app purchases is effectively over, and the data closed it. Across 2026 reporting, players who engage with rewarded ads are consistently cited as roughly four times more likely to make a purchase and up to several times more likely to be retained. The settled pattern: monetize the roughly 95% of players who will never pay through rewarded video, and reserve in-app purchases for the ~5% who will. The two lines don’t compete; adding rewarded video lifts total ad revenue without denting IAP.
AppLixir was built for exactly this shape of web publisher — HTML5-first rewarded video that slots into a hybrid stack without a tracking backend. Where that plays out in real revenue is the part worth showing with our own data:
ARPDAU lift after adding rewarded video, before vs. after: [40% ARPDAU % lift]
Retention effect (D1 / D7) for rewarded-engaged users: [11% increase in retention]
Cheaper testing widened the top of the funnel
A quieter force reshaped rewarded video in the first half of 2026: AI collapsed the cost of testing creative. Studios that used to ship a handful of ad variants a quarter now generate, score, and rotate hundreds a month, with network tooling doing the heavy lifting. More creative in rotation means better matches between ad and audience, which lifts completion and, downstream, eCPM. For web publishers this matters twice over — the same tooling that makes rewarded video cheaper to optimize also makes it easier for smaller HTML5 studios to compete for premium demand without a large ad-ops team behind them.
Distribution stopped being the bottleneck
The old friction in web games — writing a separate integration for every portal — has largely dissolved. Platform-as-a-service layers now let a studio publish a single build across many destinations and keep the large majority of the revenue. That changes the monetization math: a web game can reach audience scale that used to require native distribution, while keeping the lightweight, privacy-clean stack the web is suited to. Rewarded video rides directly on that reach, because more distribution means more opt-in impressions across more contexts.
The market is big enough to matter
This isn’t a niche anymore. The HTML5 gaming market has pushed past several billion dollars, with WebGL 2.0 and WebGPU narrowing the visual gap with native apps and portals distributing single builds across dozens of destinations. A bigger, better-looking web game market means more premium rewarded inventory — and more publishers who need a monetization layer that respects both performance and privacy.
That growth also changes who the buyer is. As web games start to look and play like native ones, the advertisers bidding on their inventory increasingly include the same brand and performance budgets that used to treat the web as an afterthought. Premium demand follows premium supply, and for the first time in a while the web is producing the kind of high-completion, brand-safe rewarded inventory those budgets want. The publishers positioned to capture it are the ones already running a clean, opt-in stack — because that’s the inventory the 2026 auction pays the most for.
The Mid-Year Rewarded Video Scorecard
To make the verdict portable, here’s the format scored on the four axes that decide whether a revenue line is healthy or quietly eroding. Call it the Mid-Year Rewarded Video Scorecard — a quick read on where rewarded video stands at the halfway point of 2026.
Axis
H1 2026 verdict
Direction
Why
eCPM stability
Held
Flat / up
Still top-paying format; tier-1 ~$5–$10
Opt-in health
Strengthened
Up
Consent flows lifted opt-in ~35% → 60%+
Retention safety
Intact
Positive
Rewarded engagement lifts retention, not churn
Privacy-readiness
Differentiator
Rising
Consented inventory now earns a clear premium
Read together, the four axes tell one story: nothing about rewarded video weakened in the first half of 2026, and the one axis that changed the most — privacy-readiness — moved in its favor. Add your own network reading on the first axis to make the scorecard yours:
AppLixir eCPM trend across H1 2026 (Jan → June): [14% growth]
The mid-year scorecard reads clean across all four axes — and the axis that moved most, privacy-readiness, is the one rewarded video was already built to win.
The one caveat worth flagging
A verdict that only points one direction isn’t worth much, so here’s the honest counterweight. Rewarded video’s strength is also a ceiling: it’s opt-in, which means its revenue is capped by how often players are willing to watch. You can’t force the impression, and pushing reward frequency too hard erodes the very experience that keeps opt-in rates high. The format scales with engaged sessions, not with raw traffic, so a game with thin retention or few natural reward moments will underperform the benchmarks no matter how good the demand is. The publishers who won in H1 2026 weren’t the ones who bolted rewarded video on — they were the ones who designed reward moments into the core loop and treated opt-in as something to earn each session. Signal loss made the format more valuable; it didn’t make it effortless.
What it means for the second half of 2026
The forward read follows directly from the first-half data. Three things to plan around:
Treat consent UX as a revenue feature, not a legal chore. The opt-in lift from a well-designed value exchange is now large enough to show up in ARPDAU. The consent screen is a monetization surface.
Weight the web in your roadmap. Mobile rewarded video is an optimization game; web rewarded video is still a land-grab, with adblock-resistant delivery and a fast-growing HTML5 market underneath it.
Assume the consented-inventory premium compounds. If signal keeps getting scarcer through H2, the spread between privacy-clean and unconsented inventory should widen further — rewarding stacks that were built opt-in from the start.
None of this requires a bet on a new format. It requires leaning into the one that already earns without surveillance — and making sure your web inventory is set up to capture the premium the market is now paying for it.
Frequently asked questions
Is rewarded video still the highest-paying ad format in 2026?
Yes. Through the first half of 2026, rewarded video remained the top-paying line in most gaming stacks, with tier-1 US eCPMs near $6–$15 and a broader range of roughly $15–$40 — ahead of interstitials and well ahead of banners.
What’s a good rewarded video eCPM in 2026?
In tier-1 markets, US rewarded video eCPMs are commonly cited around $16.49 on Android and $19.63 on iOS, with the wider range running $15–$40. Global averages are lower, and regional differences of four to ten times are normal, so benchmark against your own traffic mix rather than a single global number.
How is signal loss affecting rewarded video?
It’s helping it, relatively. Under ATT and Privacy Sandbox, unconsented traffic earns roughly 20–40% lower eCPMs. Because rewarded video is opt-in and consented by design, it sits on the higher-earning side of that spread — the format advertisers still trust when targeting signal is scarce.
Is rewarded video better for web/HTML5 or mobile?
Both perform, but the web is where it’s growing fastest. WebGL and in-canvas formats resist adblockers, the HTML5 market is expanding, and web publishers face the same privacy pressures that reward opt-in inventory — so rewarded video is an especially strong fit for HTML5 and WebGL games.
Does rewarded video cannibalize in-app purchase revenue?
No — the 2026 data is consistent on this. Rewarded-engaged players are several times more likely to purchase and to be retained. The standard hybrid model monetizes non-payers with rewarded video and payers with IAP; adding rewarded video lifts total revenue without denting purchases.
Monetize your web game the privacy-first way
AppLixir is HTML5-first rewarded video built for web and WebGL games — opt-in by design, TCF 2.3 / GDPR compliant, and live with a single line of JavaScript. No tracking stack required.
Get started at applixir.com
The post Web Rewarded Video Ad Performance in 2026: The Mid-Year Verdict appeared first on AppLixir – Rewarded Video Ad Monetization.
