Choosing a mobile game people actually search for: why I started with Word Search
The first rule in my plan is distribution before polish: a good game nobody finds earns nothing. So I turned the usual process around. Before writing any code, I ask what people already type into the Play Store search box, who those people are, and whether a new app has any chance of showing up for it.
The first answer was Word Search.
Why word search
- Search volume. "Word search" is one of the biggest evergreen puzzle searches on Google Play, and it never goes out of fashion.
- The audience. Word search players skew 35 to 70, a bit more female than male, and heavily American. They're tolerant of ads, play long relaxed sessions (which is good for banner time), and they buy "remove ads".
- Unlimited content. Grids are generated, so the only authored content is word lists. A level never runs out.
- Low risk. A word search is a known shape. Nobody needs a tutorial.
There's a catch, and I knew it going in: the head term is crowded. The top results for "word search" have tens of millions of installs. A brand-new app won't take one of those slots. I made the bet anyway, for two reasons.
The multiplier: every language is its own market
People don't search "word search" in Spanish. They search sopa de letras. In Brazil it's caça-palavras, in France mots mêlés, in Germany Buchstabensalat or Wortsuche, in Italy parole intrecciate, in the Netherlands woordzoeker.
Each of those is a separate search market with its own competition, and Google Play lets one app have a separate listing (title, description, screenshots) per language. So one game becomes fifteen shop windows:
- en-US: Word Search - Word Find Puzzle
- es: Sopa de Letras en Español
- pt-BR: Caça-Palavras em Português
- fr-FR: Mots Mêlés en Français (and fr-CA: Mots Cachés en Français)
- de: Wortsuche - Buchstabensalat
- it: Parole Intrecciate in Italiano
- nl: Woordzoeker in het Nederlands
- pl: Wykreślanka po polsku
- and Danish, Swedish, Norwegian and Finnish versions with their own genre words
The game itself ships a native word pack for each language, not translated English words. A Spanish player gets Spanish themes with Spanish words, accents handled properly. That matters for reviews, and reviews matter for ranking.
The engine bet
The second reason is the bigger one. Word Search isn't really the product. The product is the engine underneath it: a TypeScript game wrapped for Android with Capacitor, with ads (banner, interstitial, rewarded), a remove-ads purchase, analytics, consent, daily puzzles and streaks, reminders, and remote config. All of it is built once.
Every later game in this blog starts as a copy of that engine with the rules, content and art swapped out. The first fork, Bible Word Search, took about two hours. When the plumbing is done, a new niche costs an afternoon.
How I judge whether a niche is open
Search volume isn't public on Google Play, so I use a rough proxy: look at the top ten results for a search and ask two questions.
- How big are the leaders? The median install count of the top ten tells you whether people want this thing at all.
- Is there room for a newcomer? How many of those ten have fewer than 100,000 installs, or are less than a year old? If small, young apps are in the top ten, Google is willing to rank new apps for that term.
By that measure "word search" is closed: huge demand, no small apps in the top ten. So are solitaire, sudoku, crossword and jigsaw. Others were wide open, like FreeCell and Minesweeper, each with tiny apps in the top ten despite medians in the millions. Those became later games.
Was it the right call?
Partly. Word Search went live, the ads work, and it was the first game to earn real money: eight ad impressions from a player in Spain, worth less than a cent, which I was absurdly happy about. It's still one of the steadier earners in the word group.
It also hasn't cracked the top results for "word search", and I don't expect it to soon. A new app has to earn installs, ratings and reviews before it ranks for a head term. The lesson I took forward: pick at least some titles a new app can actually win, the long-tail phrases where the top ten includes apps your size.
Next week: how the grid generator works, and why "every word appears exactly once" was harder than it sounds.