2468 Block Merge

Why we built it

Most of our work is for clients. This one is ours, and that was the point. Running our own live title means we carry the same problems our clients carry: store ranking, retention that decays, ad revenue that depends on players not resenting the ads. It is a different kind of knowledge from shipping and walking away.

The merge puzzle category is crowded, so the goal was never to invent a new genre. It was to build a version that respects the player’s time and earns its revenue over months rather than days.

How we built it

Three ways to play, one economy

The game has three modes, Number Shoot, Number Merge and Color Paths. They share one gem economy, daily rewards and quests, so progress in one mode carries into the others.

Retention is a systems problem

Puzzle players judge a game on whether the tenth session is as good as the first. That is decided by the board rules, the reward pacing and the difficulty curve, not by how the game looks. So the systems were designed first and the presentation was fitted around them.

An economy that does not break

Merge mechanics are easy to break. One reward loop that is slightly too generous and progression flattens within a week, which players feel long before they can explain it. We modelled the progression before building it, because finding the stall point in a spreadsheet costs an afternoon and finding it in a live game costs a release cycle.

The near miss

The target we worked to was that a player should lose feeling like they were one move away, not feeling cheated. In a merge game that is mostly about how the board fills and what options remain when it is nearly full. Getting that right took more iteration than any other part of the build.

Monetisation that does not cost retention

The game is ad supported. Every placement was judged against whether players kept coming back, not against what it earned that day. A rewarded placement that lifts revenue this week and costs you day-seven retention is a bad trade, and it is an easy one to make by accident when you only watch the revenue chart.

Tuning without shipping a build

Difficulty and reward values are driven by remote config rather than baked into the binary. That turns a two-week feedback loop, where every change waits on store review, into a same-day one. For a live puzzle game that is the difference between reacting to your players and guessing.

Where it stands today

As of 2 October 2026 the game has passed 10,000 downloads on Google Play and holds 4.7 stars. It is live, it is ours, and we are still working on it.

It is Android only today. The store link at the top of this page goes straight to the listing.

What it gives our clients

Running our own title is why we can talk about store listing experiments, ad placement and retention curves from experience rather than from a blog post. When we tell a client that a monetisation decision will cost them retention, it is usually because we have already made that mistake on our own game.