Skip to content

← Writing

engineering

Wordle and the Beauty of Tiny Web Apps

· Jerwin Arnado

Archive note: this is a backdated post, written years later while rebuilding this site. It’s dated to the moment it covers, but the hindsight is real.

The biggest game in the world right now is a static web page. Wordle — five letters, six guesses, one puzzle per day — went from about 90 players in November to millions in January, and this week the New York Times bought it for a price “in the low seven figures.”

No app store. No account. No ads, no tracking, no push notifications, no monetization whatsoever. Josh Wardle built it as a gift for his partner, in plain JavaScript, and put it on the web. As someone who builds web software for a living, I find this the most encouraging thing to happen to the industry in years — so let’s appreciate why it works, as engineering.

Design decisions disguised as simplicity

  1. One puzzle a day is the growth engine. Everything in modern product design says maximize engagement: endless levels, streaks, energy timers. Wordle does the opposite — it ends. That scarcity makes it a shared daily ritual instead of a private compulsion: everyone on Earth is solving the same puzzle, which makes it conversation. The restraint isn’t a missing feature; it’s the entire distribution strategy.
  2. The emoji grid is genius-tier virality. 🟩🟨⬛ — the share format spoils nothing, requires no screenshot, works in any chat app, and is simultaneously a brag, an invitation, and an ad. The most effective social integration of the year is a string of Unicode squares. No SDK involved.
  3. State lives in localStorage. No backend, no login, no database of users. Your streak is on your device. This is why it scaled to millions without a single outage story — the “server” is a CDN handing out the same static files to everyone. There is profound architecture wisdom in choosing to have nothing that can fall over.
  4. It respects the player. Loads instantly, works on any phone, takes three minutes, then lets you leave. After a year of writing about platforms engineered to never let you go, a hit game with a built-in exit feels almost radical.

The counter-evidence to everything we’re told

The app stores are already flooding with predatory Wordle clones — paywalled, ad-stuffed, IAP-riddled — which makes the original’s success read even more clearly: people can tell the difference, and they reward the thing that doesn’t exploit them. The moat wasn’t technology (any of us could build Wordle in a weekend — the word list is the hard part). The moat was goodwill.

There’s also a quiet lesson for those of us who default to heavy stacks. Wordle needed no framework, no build pipeline, no Kubernetes. The web platform — HTML, CSS, vanilla JS, localStorage — was sufficient. Most software is smaller than its architecture. I say this as someone who lives happily in Laravel and Vue: the right tool is sometimes almost nothing.

The wish

Wardle built a small, kind thing, refused to ruin it, and got paid anyway. May the takeaway the industry absorbs be that — and not “acquire viral games faster.” The NYT says Wordle will “initially” remain free. Enjoy the era of the unspoiled version; tiny software like this is proof the personal web still has room to win.

Today’s was guessed in four. 🟨⬛🟩🟩🟩