An Employee Is Not an Audience: What Breaks When a Licensed Game Catalogue Goes Behind the Firewall

Putting a licensed HTML5 game catalogue on a company intranet breaks in three predictable places: the ad stack, the leaderboard, and the licence's domain lock. None of them show up in the demo, because the demo runs on the open web against a consumer audience. Behind a firewall, the audience is employees — a category that European data protection law treats very differently from players.

This comes up more than you would think. An internal comms team wants a break-room arcade. A safety trainer wants a puzzle at the end of the module. A sales director wants a Friday tournament on the intranet homepage. Somebody remembers that licensed casual games cost a fraction of a bespoke build, and the request lands on whoever runs the portal. It is a reasonable request. It is also the deployment context where a standard catalogue licence fits worst, and almost nobody checks before signing.

🎯 The Pitch HR Hears — and the Number Underneath It

The macro case is real and easy to quote. Gallup's State of the Global Workplace 2025 found the global share of engaged employees fell from 23% in 2023 to 21% in 2024, with managers dropping from 30% to 27%; Gallup put the cost of that one-year drop at US$438 billion in lost productivity. It was only the second annual decline in twelve years. Europe has now posted the lowest engagement of any world region five years running.

So there is budget moving toward anything labelled engagement. But look at how that budget is actually distributed, because the gap is the whole story. The Business Research Company's Gamification Market Report 2026 sizes the global gamification market at $26.91 billion in 2025 rising to $34.43 billion in 2026, a 28% CAGR. Inkwood Research, meanwhile, sizes the US gamified corporate training platforms market at $801.75 million in 2026, reaching $3.47 billion by 2034 at a 20.09% CAGR.

Those two figures are measuring different things and the methodologies are not comparable — but the shape is instructive. Corporate training platforms are a small, slower-growing slice of a market whose bulk sits in consumer loyalty, marketing and retail. Growth-rate estimates across published reports run from roughly 20% to 28% depending on scope definition, which is wide enough that you should treat any single figure as directional rather than a plan input.

The practical read: the tooling ecosystem is optimised for consumer engagement. When you drop a consumer-optimised game build inside an employer-employee relationship, the parts that make it economical on the open web become the parts that create liability.

🔌 The Ad Stack Comes Out First, and Not for a Legal Reason

Most licensed HTML5 titles ship with an ad SDK wired into the game loop — an interstitial between rounds, a rewarded placement on the continue button. On a public portal that is the revenue model. On an intranet it is dead weight before anyone gets to the legal question.

Contextual ad serving needs a crawlable, publicly resolvable URL. An internal hostname has none. There is no page for a crawler to classify, no public inventory record, no meaningful auction. The placements either fail to fill or serve house defaults, and either way you have added third-party network requests to a page inside your corporate perimeter for zero revenue.

Then the network security review notices those requests, and you are explaining to an infrastructure team why the employee portal is beaconing to ad exchanges. That conversation ends the project more often than any lawyer does.

So the requirement is an ad-free build, and that is a licensing question, not a configuration toggle. Ask before you sign:

• Is an ad-free variant of each title available, or only ad-enabled?
• Does removing the ad calls void the warranty, or is it a supported build?
• Does the pricing change when the ad revenue share disappears? It usually should — you are buying the game outright rather than sharing upside.
• Who ships the fix when a title breaks after an ad SDK is stripped out? See our note on licence terms and pricing structures for how this is normally apportioned.

⚖️ At Work, Consent Is the Wrong Lawful Basis

Here is the part that catches teams out. On a consumer portal, you collect a consent signal through a CMP and the ad stack proceeds. Inside an employment relationship, that signal is worth much less than you think.

The Article 29 Working Party's Opinion 2/2017 on data processing at work — carried forward by the European Data Protection Board — holds that the dependency built into the employer-employee relationship means employees are rarely able to refuse consent without fear of detriment. GDPR Recital 43 makes the same point about a clear imbalance of power between controller and data subject. The practical consequence is that consent is almost never a valid lawful basis at work, and employers fall back on legitimate interest under Article 6(1)(f), which requires a documented, defensible balancing test.

Run that test on a break-room arcade and it goes fine. Run it on behavioural ad targeting of your own staff and it does not. "We serve personalised advertising to our employees because it subsidises the game licence" is not a balance that survives contact with a data protection officer.

The same logic reaches further than advertising. Any per-employee gameplay telemetry — session length, time of day, frequency, drop-off — is personal data processed in an employment context. You need a lawful basis for it, a retention period, and an answer to the question of who inside the company can see it. If your honest answer is "our analytics vendor collects it by default and nobody looks at it," you have a problem you have not priced.

🏛️ In Germany, a Leaderboard Is a "Technical Device"

If you operate anywhere with a works council, the leaderboard is the single riskiest feature in the whole build.

Section 87(1)(6) of the German Works Constitution Act (BetrVG) gives the works council co-determination over the introduction and use of technical equipment intended to monitor employee behaviour or performance. The Federal Labour Court has read that expansively for decades: as German employment counsel summarise it, objective suitability to store performance data is sufficient, the employer's intent to monitor is irrelevant, and manual entry and storage alone can qualify — which sweeps in basically any IT system that captures user and access data.

A named leaderboard is a system that ranks identified employees by measured performance and stores the result. Whatever you intended, that is what it does.

The standard is genuinely unsettled

Be fair about the state of the law rather than alarmist. The Federal Administrative Court has taken a narrower line, requiring actual monitoring pressure — a realistic likelihood that behaviour or performance data is collected and stored in meaningful quantity — instead of mere technical capacity. German counsel note that it remains unclear which standard the employment courts will settle on, and the standing advice is to keep IT framework agreements in place or ensure sufficient anonymisation until it is clarified.

Anonymisation is the cheap route and it costs you almost nothing that matters. Rank teams instead of people. Show a personal best that only the player can see. Publish a floor-wide aggregate. You keep the competitive pull and you never build the artefact that triggers the argument.

🚫 February 2025 Closed the Webcam Door

Some engagement vendors still pitch sentiment or attention detection layered over interactive content. In the EU, that specific pitch became illegal at work.

Article 5(1)(f) of the EU AI Act — applicable since 2 February 2025 — prohibits "the use of AI systems to infer emotions of a natural person in the areas of workplace and education institutions, except where the use of the AI system is intended to be put in place or into the market for medical or safety reasons". Emotion inference from biometric signals — face, voice, gait, physiological data — is out. Penalties for prohibited practices reach €35 million or 7% of global annual turnover, whichever is higher.

No standard casual game does this. But "engagement analytics" platforms wrapping a game catalogue sometimes do, and the safety-or-medical carve-out is narrow. If a vendor's deck mentions inferring mood, focus or frustration from a camera or microphone in a workplace deployment, that is a hard stop in the EU, not a negotiation.

📄 Your Licence Was Drafted for a Public Website

Read the scope clause on a standard non-exclusive catalogue licence and you will usually find it defined by domain. One site, one domain, sometimes a named subdomain. Many builds ship site-locked so they simply refuse to run off the licensed host.

An intranet deployment collides with that in ways nobody drafted for:

• The host is not a public domain. An internal hostname, a private IP, or a path inside an SSO-gated portal may not match anything the licence names or the site-lock accepts.
• Users are authenticated, not anonymous. Licences written around visitors and page views translate badly to a fixed population of named employees.
• Metering has no meaning here. If pricing is tied to impressions or sessions and the games sit behind a login used by 4,000 people, both sides are guessing.
• Offline and air-gapped copies. Factory floors, retail back-offices and field devices often have no internet path at all. Serving from a local server is redistribution unless the licence explicitly allows it.
• Territory. A multinational intranet reaches every country you operate in. A licence limited by territory does not.

Ask for internal-use language in writing before you commit. If you need to serve builds from your own infrastructure rather than a licensor-hosted embed, that is usually a source-code or self-hosted arrangement, priced differently from a hosted catalogue feed.

If you're an agency building this for a client

You need sublicensing rights, and you need them named. A licence granting you the right to deploy on your domain does not let you install a catalogue on a client's intranet. This is the most common way agencies end up technically in breach on a project everyone considered low-risk, and it is entirely avoidable by asking one question during procurement.

🧯 The Four Ways This Actually Fails

Having watched enough of these, the failure modes are consistent:

1. Nobody removes the ad stack. Security review kills the project three weeks before launch, and the licence is already paid for.
2. The leaderboard ships named by default because that is how the game was built, and the works council question surfaces after go-live instead of before.
3. It is treated as training. An arcade catalogue teaches nothing about your compliance policy. A game beside a module can lift completion; a game instead of a module is a completion metric with no learning behind it.
4. Nobody owns it after week two. Internal engagement projects have no growth loop and no acquisition channel. Play collapses unless someone schedules the tournaments, rotates the titles and mentions it in comms. That is a recurring internal-comms job, not a launch task.

🛠️ What a Behind-the-Firewall Build Looks Like

Strip it back to what survives every constraint above and you get something fairly simple, which is the point:

• Ad-free builds, licensed as such, with no third-party network calls leaving the perimeter.
• No individual accounts. Local storage for personal bests. No server-side profile keyed to an employee ID.
• Team or aggregate scoring only. Competitive enough to be fun, anonymous enough to be uncontroversial.
• Aggregate telemetry. Total plays and title popularity, to decide what to rotate. Nothing per-person.
• Genuinely voluntary, and framed that way in writing. Anything with a prize attached that managers encourage participation in starts drifting toward measured performance.
• Offline-capable where the site is offline. Warehouses and stores are exactly where break-room games land best and connectivity is worst.
• A rotation schedule and a named owner in internal comms.

That is a modest, achievable deployment. It is also the version that gets approved.

🧭 Where to Start

Before you request budget, do the boring thing first: get three answers in writing from the licensor. Is there an ad-free build of every title you want? Does the scope clause cover a non-public internal host, and can it be amended if not? If an agency is involved, does the grant include sublicensing to the end client? A licensor who cannot answer those in a week is telling you something useful about how the rest of the relationship will go.

Then design the deployment around anonymity from the first wireframe rather than retrofitting it after a works council raises §87(1)(6). Retrofitting anonymity into a build that already ships named leaderboards means either editing the game or dropping the feature entirely — and if you did not license source code, editing is not on the table.

Internal deployments are one of the narrower uses of a licensed catalogue, but they are not exotic. HTML5 titles run in whatever browser the corporate image ships, need no install, and survive locked-down devices better than anything native. The constraints are legal and contractual, not technical — which means they are solvable in procurement, at a stage where solving them is cheap.

Forestry Games licenses HTML5 and Android titles for deployments including white-label portals and internal use cases, and questions about ad-free builds and licence scope for a specific deployment context are worth raising before signature rather than after. Whoever you license from, ask them early — the answer determines what you can actually ship.