The Hero Wars Playbook: The Browser First Billion-Dollar Franchise

 

A field teardown of the web-first operating system behind Hero Wars’ ~$750M browser business — the untaxed revenue channel studios like InnoGames, Wooga, GameForge, Goodgame Studios and Plarium are built to own, and how any HTML5, WebGL, or Phaser game wires in the same rewarded-video layer.

 

TL;DR

•     Everyone files Hero Wars under “mobile hit.” Its browser version, Dominion Era, is a separate ~$750M business — born on the web in 2016, still live-op’d in 2025. It doesn’t compete with the app; it complements it.

•     Strip the art away and the game is an operating system for engagement and revenue — the Live-Ops Operating System. Its six layers: Calendar, Progression Stack, Urgency Engine, IP Injection, Habit Engine, Distribution.

•     The layer most teams skip is Distribution. On the open web it earns three edges — margin, funnel, inventory. It dodges the 30% store tax, widens acquisition beyond store search, and opens ad formats the store buries.

•     Chief among those is rewarded video — opt-in and retention-positive. On the open web it’s untaxed by store commissions and free of a mobile tracking stack.

•     Wiring it into an HTML5, WebGL, Phaser, or React game is one script tag and one callback. Grant on “complete.” Hero Wars is proof web pays; AppLixir is how your web game captures it.

•     Durability comes from the system, not any single mechanic. Copy the discipline. Then tune aggression only as far as your audience tolerates — because the Aggression Tax comes due.

The headline nobody writes

Search “Hero Wars revenue” and every result tells the same story. A ~$1.7 billion mobile RPG. 185 million installs. A decade on the grossing charts. All true. All filed under mobile. And all of it steps around the more interesting number.

Hero Wars was born on the web in 2016. It launched as a browser-and-social title, live in players’ browsers before it ever reached an app store. That version never went away. Today it runs as Dominion Era. It has crossed roughly $750 million across browser and social platforms. That’s on the order of 90 million registrations. And it is still shipping new content in 2025. A three-quarter-billion-dollar web game most of the industry doesn’t realize exists. And this isn’t coasting. The web build runs the same live-ops machine as its mobile sibling. Through 2025 it added new heroes and fresh event formats, including an in-game rock festival. That kind of continuous cadence keeps a decade-old title’s daily actives climbing rather than decaying. This is a living product, actively operated on the open web. Not a legacy URL kept alive for nostalgia.

That’s why complements is the right verb. The browser build isn’t a fallback or a downgrade. It’s a parallel engine running the same operating system. It adds roughly a third of lifetime franchise revenue — from the channel everyone else treats as a relic. Study how it does that and you have a blueprint any web game can run. So let’s reverse-engineer it.

Figures here are drawn from public sources: GDEV’s filings and press releases, GameRefinery’s live-ops teardowns, and app-intelligence data. They’re directional. Re-verify before quoting in a deck.

The Live-Ops Operating System

Strip the fantasy art away and Hero Wars resolves into six interlocking layers. We call the stack the Live-Ops Operating System. No single layer is novel. What’s rare is the discipline of running all six, in rotation, for years. And the stack is platform-agnostic. Every layer runs on the browser build exactly as it does on mobile. That’s precisely why the web version earns like it does. Each play below is one layer.

Layer
What it does on the open web

1 · Calendar
Zero dead weeks — cadence is the product, and on web you ship it with no store-review lag.

2 · Progression Stack
Nested systems that keep re-opening wallets that should have closed.

3 · Urgency Engine
Limited-window, escalating-cost events that decide the purchase for the player.

4 · IP Injection
Collaborations that reacquire lapsed users and re-monetize them in one shot.

5 · Habit Engine
A recurring pass that converts daily logins into forecastable revenue.

6 · Distribution Layer
Web and cross-platform reach that dodges the store tax and opens ad inventory.

Play 1 — Run the game as an event calendar, not a content roadmap

Open any Hero Wars schedule, mobile or browser, and you notice one thing: there’s no gap. Guardians of Outland, Piggy Bank, Secret Wealth, Emerald sales, Tournament of Titan Power, Seer’s Game — they cycle almost continuously. Engagement events and spend events interleave, so the player is never in a dead week. Community trackers even rank each day by how many events overlap. Three or more at once is the optimal day to spend. GameRefinery’s read on the longevity is blunt. The biggest driver of the grossing rank is the rising number of monetization events, not any one feature. And the count keeps climbing. New permanent modes like Legends Draft (August 2024) feed the rotation. Most teams think in roadmaps: ship a feature, move on. Hero Wars thinks in calendars. The feature set is stable, and the rotation is the product.

Ideas for Similar Games >  Build a fixed, repeating calendar that alternates engagement and spend events, with zero dead weeks. The days several events stack are your highest-intent spend windows. On the web you control the whole client, so that calendar runs without a store-review lag. Cadence beats content.

Play 2 — Stack progression systems so there’s always a new wall

A veteran who has maxed their heroes should, in theory, stop spending. Hero Wars answers by stacking systems: heroes, then Titans, then Titans wired into guild mechanics. The Tournament of Titan Power (October 2024) is the tell. It’s the first solo tournament scored purely on Titan-power gains. Titans are a separate character class tied to guild play. So chasing that leaderboard pulls a “finished” hero-collector into a whole new track — and a whole new set of purchases. Every “endgame” is just the doorway to another progression track. Each track re-opens a wallet that should have closed. Heroes gate Titans. Titans gate guild standing. Guild standing gates the next tournament.

Ideas for Similar Games > Design nested systems that cross-sell each other. Optimizing one should require engaging — and spending — in the next. Layered progression is how you monetize a player who has already “finished” your game.

Play 3 — Manufacture urgency and risk

In March 2025 Nexters shipped Seer’s Game, a roulette-style event. Each round, the player picks from several cards and tries to avoid a cursed one that ends the run. The rewards escalate the deeper they go — and so does the cost. Every attempt burns Seer’s Orbs, earned slowly through tasks or bought in bundles. Wrap that around Piggy Bank and time-boxed Emerald sales, and you get a steady drip of engineered scarcity. A static store asks players to decide to spend. A closing window decides for them — on impulse, before it shuts.

Ideas for Similar Games > Limited-time, escalating-cost events convert in ways a permanent store never will. Pull loss-aversion and scarcity deliberately — and disclose the odds and costs honestly. The honesty is not optional; see the Aggression Tax below.

Play 4 — Rent relevance through IP collaborations

In 2024 Hero Wars ran its first brand crossovers. The Lara Croft / Tomb Raider event brought themed missions and IAP bundles. It reportedly pulled in over 500,000 new players. Paramount’s Teenage Mutant Ninja Turtles followed at year-end. One event, double duty: user acquisition and monetization. Plus a built-in reason for lapsed players to log back in. For a mature title, a well-chosen collab is a relevance injection. It reactivates dormant accounts, hands the press a story, and justifies premium themed bundles. The whole trick is partner fit. The IP has to be something your lapsed users already love, or the reactivation doesn’t fire.

Ideas for Similar Games > Time-boxed IP collaborations are a UA and monetization lever in one package. Pick partners your lapsed users already recognize. Wrap the collab in a limited bundle. Treat the crossover window as a reacquisition campaign, not just content.

Play 5 — Battle pass as the habit engine

Hero’s Way, the recurring pass, isn’t built to sell a big-ticket item. Its job is to convert daily logins you already have into spend you can forecast. GameRefinery flags the pass as one of the strongest revenue drivers in the game. It spikes hardest in the final days of each season, as players scramble to finish the track before it resets. A player chasing a monthly reward ladder logs in every day. And daily logins are the raw material every other play here consumes. The pass is the cheapest, most predictable revenue in free-to-play. It monetizes habit rather than impulse.

Ideas for Similar Games > A recurring pass turns engagement you already own into forecastable revenue. Its end-of-season deadline is a reliable second spike. It’s the flywheel the other plays feed.

Play 6 — The Distribution Layer: monetize across platforms, not just the App Store

Here’s the layer most mobile-native teams never build. It’s the one this whole teardown has been driving toward. Hero Wars doesn’t live only in the app stores. Its browser-and-social build, Dominion Era, is a distinct ~$750-million business. It runs the same operating system on the open web. That single fact reframes everything. The “mobile game” is really a cross-platform franchise. It earns roughly a third of its lifetime revenue outside any app store. Distribution isn’t only a reach decision. It’s a monetization decision — and on the web it pays three ways.

Edge
What the open web gives you

Margin
A meaningful share of web and social transactions sidestep the 30% platform commission mobile teams treat as a law of physics. It isn’t a law. It’s a toll you can route around.

Funnel
Acquisition runs beyond store search and ASO — links, embeds, social, and web-game portals put your game one click from play, with no install wall.

Inventory
The web opens ad formats the app economy leaves on the table — most importantly rewarded video, where players opt in to watch an ad for in-game value.

That funnel edge compounds on web-game portals. Sites like Poki and CrazyGames put a game one click from play, across tens of millions of sessions. And the genres that thrive there — casual puzzles, hyper-casual arcade, block-fitters — pair polished gameplay with an underdeveloped ad architecture. A portal audience arrives with zero install friction. It has a high tolerance for a rewarded prompt in exchange for progress. Leaving that inventory empty is the clearest money left on the table in web gaming today.

Ideas for Similar Games > Treat distribution as a monetization decision, not just a reach one. A web build widens the funnel and lifts per-transaction margin. It also opens ad formats the store model buries — rewarded video chief among them.

The layer web unlocks: Rewarded Video Ads

Rewarded video is the ad format players actually tolerate, because they choose it. Take a familiar prompt: “Watch a short ad for an extra life, 100 gems, or a second chance.” It trades 15–30 seconds of opt-in attention for in-game value. No interruption, no forced pre-roll, no dark pattern. Done right, it’s the rare monetization surface that raises retention instead of taxing it. That makes it the natural complement to the engagement machine of Plays 1 through 5.

It also monetizes exactly the players IAP never will. Most of a free-to-play base never pays a cent. On the web, audiences skew even more casual and price-sensitive, so that share is higher still. Rewarded video turns those non-payers into revenue without asking for a credit card. And on the open web, that revenue never passes through a store commission. Web rewarded eCPMs sit in a directional $4–15 range, with rewarded placements at the premium end. For a casual/RPG audience, that can mean [AppLixir data: web rewarded eCPM + opt-in/completion for casual RPG inventory].

The craft is in where you place the prompt. The best moments are the ones where the player already wants what the ad offers. A revive after a lost run. A doubled level-completion reward. A lives or energy refill at a wall. A second spin on a near-miss reward. Each is a voluntary, well-timed offer, not an interruption. That’s why a well-placed prompt tends to lift session length and return rate, not dent them. Get the placement wrong and it feels like a paywall. Get it right and players thank you for it.

And here the web has a structural privacy advantage. Mobile rewarded inventory is tangled in device-graph tracking and store-specific identifiers. The open web is cleaner. A privacy-first placement — TCF 2.3-compliant, consent-aware, no third-party tracking stack — monetizes the same attention without the surveillance plumbing. For a browser or HTML5 game, that isn’t a compromise. It’s the native way to do it.

How you actually wire it in

Here’s the part that surprises teams bracing for an SDK slog. On the web, a rewarded layer is one script tag and one callback. Below is the AppLixir drop-in. The base pattern is identical across HTML5, Phaser, React, and Unity WebGL. Only the trigger wiring changes.

HTML5 / vanilla JS — the base pattern

<!– 1 · Load the SDK (pin the version in production) –>
<script src=”https://cdn.applixir.com/applixir.app.v6.1.0.js”></script>
<!– 2 · Player container –>
<div id=”applixir-ad-container”></div>
<!– 3 · Show an ad on a user action –>
<script>
function showAd() {
   initializeAndOpenPlayer({
     apiKey: “YOUR-API-KEY”,
     injectionElementId: “applixir-ad-container”,
     // adStatusCallbackFn receives a STATUS OBJECT: { type, ad?, error? }
     adStatusCallbackFn: (status) => {
       if (status.type === “complete”) grantReward();
     },
     adErrorCallbackFn: (e) => console.error(e.getError().data),
   });
}
</script>

The one rule that matters for correctness: grant the reward only on status.type === “complete”. The SDK fires a status object on every lifecycle event, so it’s easy to reward on the wrong one. Take allAdsCompleted. It fires at the end of any ad, and even when no ad was available. So it’s a cleanup signal, not a reward signal.

status.type
Meaning
Reward?

complete
User watched the full video
✅ Yes

skipped
User skipped the ad
❌ No

manuallyEnded
User closed it early
❌ No

allAdsCompleted
End of any ad, or no fill — clean up
❌ No

consentDeclined
User declined personalized-ads consent
❌ No

 

Two production notes

Make rewards fraud-proof — the client callback is optimistic. Use the server-side web callback as the source of truth before granting anything valuable.

Add your ads.txt — drop the entries from your dashboard to unlock full fill rates. Skipping this quietly caps your revenue.

The counter-lesson: the Aggression Tax

None of this is a license to bolt every mechanic onto your game and hope. The same event density that drives the revenue also draws “cash grab” reviews. And the franchise has a long history of misleading puzzle-style ad creatives. That drew regulatory scrutiny in Germany over a decade ago. It remains a real, recurring brand cost. Call the two together the Aggression Tax. A churn tax, paid by players who burn out on the pressure. And a reputation tax, paid at the store listing and in the press.

The Aggression Tax

Churn tax — players who burn out on constant urgency and leave, shrinking the base every future event depends on.

Reputation tax — “cash grab” reviews, misleading-ad blowback, and regulatory attention that raise acquisition costs for years.

The tax is easy to underweight, because it’s deferred. Even Hero Wars shows the strain. Through 2025, GDEV’s top-line softened year-over-year. Consumer spending tightened. The company leaned on efficiency and margin, not growth, to keep profit climbing. A machine this good still can’t out-run a saturated, price-sensitive audience forever. Aggression borrows from your future retention. The bill always arrives. And notice which layer is least taxing. Rewarded video is opt-in. It adds revenue without adding pressure — the rare surface that doesn’t feed the Aggression Tax at all.

The lesson isn’t “do all of this.” Durability comes from the system — the calendar, the layered progression, the cross-platform reach. Not from any single aggressive mechanic. Build the operating system first. Then tune the aggression to what your audience will tolerate. No further.

The meta-play

A game that out-earns titles half its age isn’t winning on a mechanic. It’s winning on operating discipline. Relentless cadence. Systems that keep re-opening the wallet. A pass that compounds habit. Distribution that refuses to accept the store tax as inevitable. They’ve run it for nearly ten years — on mobile and the open web in parallel. The core loop got them to launch. The operating system kept them on the charts.

Where it goes next is the interesting part. Three forces are converging: web-first distribution, rewarded video, and AI-assisted live-ops tuning. Together they make the operating system more of the product than the game itself. One honest footnote, though. Hero Wars didn’t need a third-party network to prove the point — they built their own web monetization stack over a decade. Most teams can’t. That’s the gap a drop-in rewarded layer closes. Hero Wars is the proof the web push pays. AppLixir is how a smaller team captures it — in an afternoon, with one script tag, no tracking stack, and no 30% toll. Treat monetization as an operating discipline. Treat the open web as a first-class channel, not a fallback. That’s who’s still charting in 2035.

The core loop gets you to launch. The operating system — run on mobile and web at once — is what keeps you on the charts.

FAQ

How much of Hero Wars’ revenue actually comes from the web?

The browser version, Dominion Era, has crossed roughly $750 million across browser and social platforms. That’s separate from the ~$1.7 billion the mobile version has earned. It’s on the order of a third of lifetime franchise revenue — from a channel most coverage ignores.

What actually keeps Hero Wars on the charts after nearly ten years?

Operating discipline, not the core combat. A relentless event calendar. Layered progression. A habit-forming pass. Cross-platform distribution. Run in rotation for almost a decade, on mobile and web in parallel. The mechanics are ordinary. The discipline of running all of them together is not.

Why is rewarded video better suited to the web than to mobile?

On the open web it sidesteps the 30% store commission. It also doesn’t depend on a mobile-style tracking stack. A privacy-first, TCF 2.3 placement monetizes opt-in attention without the surveillance plumbing. So the format that’s player-friendly everywhere is also margin-friendly and privacy-friendly on web.

How hard is it to add rewarded video to an HTML5 or Phaser game?

One script tag, one container div, one callback. The base pattern is identical across HTML5, Phaser, React, and Unity WebGL. Only the trigger wiring changes. The single rule to get right: grant the reward on status.type === “complete”. For anything valuable, use the server-side callback as the source of truth.

Could a web game like Hero Wars run on AppLixir?

Yes — that’s exactly what it’s built for. Hero Wars proves web-first monetization pays at scale. They just spent nearly a decade building their own stack to do it. Most teams can’t. AppLixir gives you the same rewarded-video layer as a drop-in: one script tag, no tracking stack, no 30% store cut.

I’m Robel Alemu, co-founder of AppLixir — the web-first rewarded video platform for HTML5, WebGL, and browser-based games. Privacy-first, TCF 2.3, no tracking stack.

Running a web or cross-platform game with an empty rewarded-video layer? That’s the conversation I’m always up for. The integration examples above are open source at github.com/applixirinc/applixir-integration.

The post The Hero Wars Playbook: The Browser First Billion-Dollar Franchise appeared first on AppLixir – Rewarded Video Ad Monetization.