What Launching pickuma-play Games Taught Us About Traffic and Audience

Sometime late last year we did a slightly ridiculous thing: we took a calm, text-heavy editorial blog about developer tools and bolted a hub of small browser games onto the same brand. The blog is what you are reading now. The games live at play.pickuma.com — a handful of tiny, instantly-playable things you open in a tab, mess with for ninety seconds, and either close or share. On paper the two surfaces have nothing to do with each other. That was sort of the point. We wanted to know, with real numbers instead of vibes, whether a developer brand can run more than one surface, and whether the audience for one feeds the other. The short answer is yes to the first and mostly no to the second, and the gap between those two answers is the most useful thing we have learned all year.

This is a meta essay, not a launch announcement. I am going to be honest about the numbers in the way you can be honest when you are not raising money against them: in shapes and ratios rather than precise figures I would have to fabricate to make a point. The shapes are what matter anyway.

Why a content site spun up a games hub

The obvious question is why. We did not build the games to grow the blog. If you want blog traffic, the move is to write more for the blog, not to learn a second discipline from scratch. We built the games for two honest reasons.

The first was that we wanted a surface where distribution worked on a completely different clock. Editorial SEO is a slow, compounding game — you publish, you wait weeks, you watch a page slowly climb and then plateau, and the feedback loop between “I made a thing” and “the thing found an audience” is measured in months. That is a hard environment to learn distribution in, because every experiment takes a season to read. Games are the opposite. You ship a small game, and within hours you know whether anyone wants to share it. We wanted a fast loop to sharpen instincts that the slow loop kept blunting.

The second reason was that a games hub forces you to confront virality directly, with no SEO crutch. A blog post can be mediocre and still accumulate traffic if it ranks for the right query. A browser game that nobody shares is simply dead — there is no search demand for “fun thing I have never heard of.” That brutality is clarifying. It made us think about share mechanics, first-session retention, and the actual moment a person decides to send a link to a friend, in a way that years of writing never had.

Two completely different distribution machines

Once both surfaces were live, the difference in how they get found stopped being theoretical. They are not two flavors of the same channel. They are two different machines.

The blog runs on search. Someone has a problem — “is this MCP server worth installing,” “how should I disclose AI-assisted writing” — they type it into Google or ask an assistant, and a page we wrote months ago answers it. The traffic is intent-driven, patient, and almost entirely from strangers who will read one article and may never return. Growth is a function of coverage and authority accumulating over time. There is no moment of virality; there is just a slope.

The games run on share loops and discovery platforms. The two channels that actually moved the needle were itch.io and Bluesky. itch.io is a discovery surface in its own right — people browse it looking for small web games to play, which is exactly the intent we needed, and it sends a steadier trickle than you would expect from a directory listing. Bluesky was the surprise. Game crossposts there — a short clip or a GIF of the thing in motion with a direct play link — got picked up and reshared in a way our blog links never do. A blog link on social is a promise of homework; a playable game link is a thirty-second payoff someone can verify by clicking. The reshare math is completely different.

And underneath all of it, the share loop. Every game link carries UTM parameters so that when a player shares it and a friend clicks, we can see the path. That is the whole engine: play, enjoy, share, friend plays, friend shares. When it works, a single seed visit can fan out into a small tree. When it does not — which is most of the time, honestly — the game just sits there. The difference between a game that loops and one that does not is almost never traffic volume at the top; it is the share rate in the middle, and the share rate is a property of the game, not the marketing.

What actually crossed over (spoiler: almost nothing)

Here is the part that surprised us least in hindsight and most in the moment. We assumed there would be meaningful audience overlap. Same brand, cross-links in the footer, a shared sense of taste — surely some blog readers would wander over to the games and some game players would discover the blog. We instrumented the cross-links specifically to measure this.

The crossover was nearly zero. Not literally zero, but small enough that you would be a fool to plan around it. Blog readers came for a specific answer and left when they had it; a footer link to browser games read as noise, not invitation. Game players were in a play mindset and had no interest in long-form writing about developer tooling — the link to the blog was, from their side, equally noise. The two audiences are not segments of one population. They are two different populations who happen to share a logo.

The non-overlap is not a failure. It reframed the project. We stopped treating the games as a funnel into the blog and started treating both as independent surfaces of one developer brand, each with its own audience, its own distribution machine, and its own definition of success. That framing is much closer to how an indie studio with several products thinks, and much further from the content-marketing playbook we had half-consciously imported.

Retention and virality pull in opposite directions

The two surfaces also taught us opposite lessons about keeping people, and holding both lessons at once is uncomfortable.

The blog’s version of retention is the newsletter and the slow return. A reader who liked one article might subscribe, and months later read another. Return visits are rare and that is fine — the model assumes most readers are one-and-done strangers, and the value is in serving each of them well enough to rank. There is essentially no in-session retention to optimize; there is a page, and then there is the back button.

The games are the inverse. In-session retention is everything, because it directly drives the share. A game that holds someone for two minutes produces a dramatically higher share rate than one that loses them in twenty seconds, and the relationship is not linear — there is a threshold, somewhere around “they got good enough to want to show someone,” past which sharing becomes likely and below which it almost never happens. So on the games side we obsess over the first session in a way the blog never demands. But long-term retention on the games is genuinely worse than the blog — people play a thing, share it, and rarely come back to that specific game. The games win on virality and lose on durability; the blog wins on durability and has no virality at all.

Holding both surfaces forced us to stop talking about “retention” as one thing. There is in-session retention that fuels sharing, and there is long-term retention that fuels compounding, and a given surface is usually built for one or the other. Pretending a single metric covers both is how you optimize the wrong loop.

How the surfaces compare, honestly

Laid side by side, the contrast is clean enough to tabulate. These are directional, not precise — the point is the shape of each machine.

The blog is a slope; the games are a series of spikes. Neither is better. They are answers to different questions, and the mistake we almost made was treating one as a worse version of the other.

The broader thesis: distribution beats product

If there is one thing both surfaces shouted in unison, it is the unglamorous thesis that distribution beats product. On the blog, two articles of roughly equal quality can perform an order of magnitude apart purely because one matched a real search query and one did not. On the games side it is even starker: we have shipped games we were proud of that went nowhere because they had no share mechanic, and simpler ones that traveled because the share loop was tight. Quality is necessary and nowhere near sufficient. The channel, the loop, the moment of sharing — that is where outcomes are decided.

Running two surfaces also made the case that a developer brand can credibly operate more than one. The cost is real — a second surface is a second discipline, a second set of metrics, a second thing to keep alive — and you should not do it expecting the audiences to merge. But the cross-discipline learning is genuine. The games taught the blog to think about the first session and the moment of sharing. The blog taught the games to respect the compounding patience of search. Neither surface would have learned its counterpart’s lesson on its own, and that, far more than any shared traffic, is the return on running both.

Who should run a second surface

If you are a writer hoping a side project will funnel readers to your main thing, do not bother — the crossover will disappoint you, and you will have split your attention for nothing. If you are running a content site purely for traffic, stay focused; a second surface dilutes the one discipline you are trying to master.

But if you have a developer brand and you are genuinely curious how distribution behaves on a faster clock than SEO allows, a second surface is one of the best teachers available. Build it as an independent thing with its own success metric, instrument the share loop and the cross-links so you can measure crossover honestly, and treat any audience overlap as a pleasant accident rather than the plan. The justification is the learning, and the learning is real. Just do not lie to yourself about which audience is reading which thing.

FAQ