Google Crawled All 500 of Your Licensed Game Pages. The Ones That Rank Are the Category Pages.

When you license HTML5 games by the hundred, page count is not traffic. Search rewards pages you build around a catalogue, not the per-title pages it fills.

The launch plan almost always assumes the arithmetic runs forwards. Five hundred licensed titles become five hundred game pages, five hundred game pages become five hundred entries in Google's index, and the long tail does the rest. Then six months pass, and Search Console shows most of those URLs parked under Crawled — currently not indexed. The sessions that did arrive landed on eleven category pages somebody built in an afternoon.

That is not a defect in the catalogue. It is what happens when the thing you bought is a licence to run software and the thing search ranks is a page. Those are different products, and nobody in the licence conversation is accountable for the second one.

🔎 Google Counts Pages It Can Tell Apart

Start with what Google actually says, because the wording matters more than the folklore around it. Its spam policies for Google Web Search define scaled content abuse as "when many pages are generated for the primary purpose of manipulating search rankings and not helping users," and note that the practice is "typically focused on creating large amounts of unoriginal content that provides little or no value to users."

A portal running licensed games is not the target of that policy. Your pages help users — there is a playable game on them, and intent is the hinge. Worth saying plainly, because the version of this argument that circulates online is that licensed catalogues get penalised, and that is not what the policy describes.

The problem is quieter and far more common than a penalty. Google does not have to demote you to cost you the traffic; it only has to decline to index. Index selection is a judgement about whether a page adds anything to what is already indexed, and it is made per URL, silently, with no notification and no appeal. Five hundred pages carrying large amounts of unoriginal content is precisely the shape that judgement goes against — not because anyone thinks you are a spammer, but because the pages genuinely do not differ from each other or from the same titles on other portals.

The practical translation: treat indexation as an outcome you have to earn per page, not a default you get for publishing. Most catalogue launches never measure it, which is why the shortfall shows up as a traffic mystery in month six rather than a diagnosis in week two.

🖼️ The Game Is in an iframe. The Indexable Part Is the Sentence Underneath It.

A licensed HTML5 build reaches your page one of two ways. Either you host the files on your own domain, or you embed the licensor's hosted URL in an iframe. Operators argue about these on latency and control grounds. The search consequence rarely comes up, and it is the more permanent of the two.

A crawler does not play your game. It reads the wrapper. With an iframe, that separation is formal: Google's own guidance has long been to avoid iframes for content you want credited to the page, and to provide text-based links to anything inside one so it can be crawled directly. Google can fetch iframed content, but it treats it as belonging to the source URL, not to the parent page. If you iframe a licensor's build, the strongest content on your page belongs, as far as attribution goes, to their domain.

So whichever route you take, the entire ranking substance of a game page is the text you wrapped around it: the title string, the description, the tags, the breadcrumb, and anything you added. On a typical licensed catalogue page that is about forty words. You are asking Google to rank forty words of someone else's copy against pages that have been accumulating links and engagement for a decade.

📄 The Description Was Written Once and Licensed Many Times

Here is the part operators discover late. The one-line description in your catalogue manifest was not written for you. It was written once, by the licensor, and it ships with the title to every licensee who takes it. So do the tags. So does the canonical title string, often down to the capitalisation.

If a title is non-exclusive — and most of a catalogue is — that paragraph is live on every other portal running the same game. Google is not choosing between your page and nothing. It is choosing among a set of near-identical pages, and when pages are equivalent it falls back on site-level signals: authority, history, engagement, links. A portal that launched last quarter loses that comparison every time.

The fix is not clever. It is to write. But writing five hundred original descriptions is real editorial cost, so triage instead of pretending you will do all of them:

• Rank the catalogue by expected demand before you write a word. Search volume for the title name, plus whatever you know about play rates from a pilot. The distribution is brutally uneven and that is useful — it tells you where the writing budget goes.
• Write properly for the top tier. Ten to fifteen per cent of titles, given real copy: what the controls actually are, how it behaves on a mid-range Android phone, how long a session runs, who it suits and who it does not, which other titles in your catalogue it sits next to. None of that is in the manifest, because only an operator who played it can write it.
• Make a deliberate decision about the rest. Thin-but-indexed, or excluded from the index and reachable through internal links and site search only. Choosing to noindex a chunk of a catalogue you paid for feels wrong and is frequently correct: those URLs were never going to rank, and they dilute how the rest of the site is assessed.

The test to apply to any page you keep indexed is the inverse of Google's wording. Does this page contain something original that provides value a user could not get from the same game on another portal? If the honest answer is no, it is not an indexing candidate.

📉 The Click You Modelled Got Smaller, and the Sources Disagree About How Much

Any traffic forecast written before 2024 is now working from the wrong click-through assumptions. The measured numbers are worth knowing precisely, including where they conflict.

The Pew Research Center study published in July 2025 tracked 68,879 Google searches from 900 US adults during March 2025. Where an AI summary appeared, users clicked a traditional search result on 8% of visits. Where none appeared, 15% — close to double. Links inside the AI summary itself were clicked on 1% of visits. And 26% of visits with a summary ended the browsing session entirely, against 16% without.

Similarweb's tracking over roughly the same period put the share of Google searches ending without any click at 69% in May 2025, up from 56% a year earlier, before AI Overviews rolled out broadly.

Now the disagreement. Semrush analysed more than 10 million keywords across 2025, reported via Search Engine Land, and found AI Overview presence itself was volatile: 6.5% of queries in January 2025, close to 25% by July, then back under 16% by November. On zero-click behaviour it found the opposite direction to Similarweb — zero-click rates on the keywords it compared fell slightly, from 33.75% to 31.53%, once AI Overviews appeared.

Those studies measure different keyword sets by different methods, so read a range rather than a number: a given impression is worth somewhere between modestly and roughly half what it used to be, and AI Overview coverage moves month to month. Anyone quoting a single confident figure has picked a study.

Why this lands unevenly on a games portal

The Semrush data has a second finding that matters more to this business than the headline. The intent mix of queries that trigger AI Overviews shifted hard through 2025: informational queries fell from about 91% of triggers in January to roughly 57% by October, while commercial rose to about 18%, transactional to about 14%, and navigational climbed from under 1% to over 10% by November.

Game demand skews navigational and transactional. People search a title by name, or a genre plus "play", or a title plus a modifier. That is the intent class AI Overviews have historically covered least — and the one growing fastest. Which gives you a two-speed exposure worth planning around:

• Your informational content — the "best two-player browser games" style listicles most portals rely on for top-of-funnel — is the most exposed right now.
• Your per-title navigational demand is less exposed today but trending the wrong way, and it is the demand a licensed catalogue is structurally worst at capturing anyway, for every reason in the previous two sections.

If your acquisition plan is a content calendar of listicles plus five hundred thin title pages, both halves are aimed at the softest ground.

🗂️ Build the Twelve Pages That Can Win, Not the Five Hundred That Can't

The collection page is the asset a licensed catalogue can genuinely own — the one page where the original contribution is real. The selection is yours, the ordering is yours, the judgement about which eleven titles belong together is yours, and no other licensee running the same games made the same call. It also matches head demand, absorbs internal link equity from everything beneath it, and survives a title being pulled at renewal. A concrete order of work:

1. Pull query demand first, then choose collections. Not the licensor's genre column — that taxonomy was built for catalogue administration, not for the way people search.
2. One collection per demand cluster with real volume, capped at what you can write properly. Twelve strong collections beat ninety templated ones, and ninety templated ones recreate the thin-page problem one level up.
3. Put original copy above the grid. Two or three hundred words explaining the selection and who it is for. A page that is only a grid of thumbnails is as unoriginal as the title pages it links to.
4. Link deliberately in both directions. Every game page links up to the collections it belongs to; collections link across to related collections. This is how the pages you chose to index get discovered and re-crawled.
5. Control faceted URLs before launch, not after. Filters crossed with sort orders and tags generate tens of thousands of near-duplicate URLs from a five-hundred-title catalogue. Make crawlable only the facet combinations you would write copy for; block the rest at parameter level.
6. Report indexation as a KPI in its own right. Indexed URLs as a share of published URLs, tracked weekly. It is the earliest signal you have, and it moves months before sessions do.

🔑 Settle These Four Things Before You License HTML5 Games

Most of what determines your search ceiling is fixed at licence signature, by people discussing formats and territories rather than indexation. Four questions to raise while you still have leverage:

• Whose domain does the build run on? Self-hosted builds mean your URL, your page speed, your control over what a crawler sees. An iframe pointed at the licensor means a third-party dependency sitting inside your most commercially important pages.
• May you rewrite and extend the supplied metadata? Usually yes for generic titles. Not always for branded ones, where names, capitalisation and approved descriptions can be contractual rather than editorial. Establish which titles are which before you plan the editorial work.
• Are the supplied descriptions shared with other licensees? For non-exclusive titles, assume yes — but ask, because the answer tells you how much original writing your launch actually needs, and it is a fair question that a serious licensor will answer directly.
• Can you set your own URLs, canonicals and index directives freely? You need the right to exclude part of a catalogue from the index and to restructure paths as your taxonomy changes. Confirm it rather than assuming it, particularly on branded strands.

One more if you sell in several markets: a regional licence caps your international targeting. There is no point building hreflang for territories the licence does not cover.

🎯 What You Get When You License HTML5 Games From a Direct Licensor

Working with a direct licensor rather than a one-off asset sale changes what you can negotiate on all four of those points, because there is a counterparty who owns the library rather than a checkout page. Forestry Games has licensed games since 2017 and maintains a catalogue of 1,049 titles spanning HTML5 and Android APK builds, with titles also published on Google Play and the Apple App Store — so hosting arrangements, build formats and metadata rights are one conversation against one library rather than five hundred separate purchases.

Scope is set to the use case: hosted play or self-hosted builds on your own domain, source where applicable, branding and white-label portal delivery, plus distribution and monetization support rather than a file handover and an invoice. If search is going to be an acquisition channel, the sizing question is which titles carry demand you can realistically rank for — so browse the catalogue and ask for a licence scope against the surfaces you actually run. The HTML5 games and Android games sections are a reasonable place to see how titles group before you commit to a taxonomy of your own.

🧸 Licensing Branded Games for Campaigns, Portals and Events

Branded titles behave differently in this discussion, and it is worth planning for rather than discovering. Character names, approved descriptions and artwork are contractual, which constrains exactly the editorial freedom the sections above depend on — but branded titles also carry their own search demand, which generic catalogue filler does not. Forestry Games works with branded IP and has brand partnerships including Disney, Nickelodeon, Cartoon Network and Warner Bros, and businesses can license branded game content through it for campaigns, portals, events and apps.

If a branded strand sits inside a wider catalogue plan, settle the copy and naming rules at scoping time so your editorial schedule reflects what is actually rewritable. Ask for a licence scope covering the branded titles alongside the general catalogue, or request a portal demo to see how both sets sit inside one storefront.

✅ Run the Index Test Before You Sign, Not After You Launch

There is a cheap version of all of this you can run this week. Take three portals already running the catalogue you are considering. Search an exact phrase from a title's supplied description in quotes and count how many domains return it. Then take twenty title pages from one of those portals and check how many are in Google's index at all. Ten minutes of work gives you a defensible estimate of how much original content your launch has to produce before a single licensed title earns a session.

Bring that number to the licence conversation. It changes what you negotiate — metadata rights, hosting arrangements, which titles you take — and it changes the launch plan from "publish five hundred pages" to "earn twelve, then extend". The catalogue is the inventory. The pages are the product, and they are the part nobody licenses to you.