One game loop,
four ad networks.
The brand wanted an ad people play instead of skip. We built one 20-second catch game, packaged it for four networks with their different size caps and install calls, and put a longer version with a leaderboard on the brand's site.
This is a composite scenario built from how we deliver this kind of work. The client is fictional and the figures are targets, not measured results. We will replace it with a named engagement when a client agrees to be quoted.
A packaged-snack brand in Bangladesh running a festival campaign through its media agency, with ads on Meta, Google, TikTok and YouTube
- Static festival ads were being skipped. The agency proposed a playable, then discovered that every network wants a different package: a 5 MB cap almost everywhere, a different install call on each, and separate review queues.
- The brand also wanted the game on its own website in Bangla and English, with a leaderboard, and it needed to run on mid-range phones over mobile data.
How we approached it.
One loop, tested on paper first
A 20-second catch game: snacks fall, customers tap to catch, bonus items multiply. We prototyped it in grey boxes and playtested with twelve people on their own phones before any art.
A size budget before art
A 3 MB ceiling for the ad build, with a sprite atlas, compressed audio and no framework. Every asset was weighed before it was drawn.
One codebase, four wrappers
The core loop stays identical. Thin adapters handle each network's install call and packaging, so a fix lands once.
Web version with a leaderboard
The same loop runs on the brand site at full length, in Bangla and English, with a weekly leaderboard and basic anti-cheat.
Measure what players do
Analytics events for start, catch, bonus, finish and click-through, in both the ad and the web version.
Try it yourself.
A simplified illustration with made-up data. It plays once when it scrolls into view, and you can pause it or step through.
Tap the phone to add your own taps.
Step 1 / 6 An empty shell. The size meter reads 0.0 MB.
What we built.
- HTML5 canvas game loop with Bangla and English text
- Four ad-network builds from one codebase
- Web version with weekly leaderboard and server-side score checks
- Asset pipeline with size budget checks
- Analytics dashboard for the agency
Outcomes (targets).
These are the results we would aim for. They are not measured results from a named client.
- Target2.9 MBad build sizeunder every network's cap
- Target4network builds from one codebasefixes land once
- Target71%players who finished the 20 secondsweb version, first two weeks
- Target64 smedian web sessionagainst a 20-second game
Timeline and stack.
6 weeks from brief to the first network submission
What we learned.
- Decide the size budget before the first sprite. Shrinking later costs a week.
- One loop, many wrappers. The game should never know which network it is in.
- Playtest on phones over mobile data. Office Wi-Fi hides every problem.
How we would deliver it.
Want this
for your business?
Message an engineer on WhatsApp or send a brief. We agree scope and price before work begins and build in small working steps you can test.