Written by Aylin, your former Puzzle Addict.
Cube Land has been one of the most talked-about puzzle launches of the summer. It went live globally on July 2, and within weeks, it was in the US top 50 free and top 80 grossing charts. The studio behind it, Rotatelab, is based in Gaziantep, a city in southeastern Türkiye, more than a thousand kilometers from Istanbul, where most of the country's game industry sits.
How Rotatelab got here is the bigger story. The team spent years shipping games, left the publisher model when they felt ready, spent a year and a half learning to self-publish with Sort Land, and put the whole studio behind Cube Land once its retention test came back. They did all of it from Gaziantep with a team spread across the country. Cube Land is what that path produced, and I think the path deserves as much attention as the game.
Ten years of shipping
The founders have been making games together since 2016. They founded Rotatelab in 2020 and spent about 2.5 years working with Rollic on hyper-casual titles. Over that period, their games passed 17 million lifetime installs, and one of them reached #1 on the US App Store free chart.
They have launched around 20 games over the years, and that number leaves out every prototype that died in testing. I read that number as a record of how many times this team went through the full cycle of building, testing, and killing a game. I see hit-making in this genre as a numbers game, and the studios that get there are usually the ones that kept taking shots long after most teams would have given up.
In August, Rotatelab closed a seed round co-led by Laton Ventures and Arcadia Gaming Partners.
November to March
The idea for Cube Land came up in November 2025, and development started at the end of December. The first marketability test ran in January 2026, around the time the market was moving toward pixel-style art because it brought cheaper installs. Results were good enough for two months of focused production. The retention test at the end of March held up, and after that, the whole studio moved onto the game.
About 4 months passed between the idea and a validated game. Rotatelab says it built its own AI tools covering market analysis, playable prototypes, and marketing automation, and this timeline supports that claim better than any press release could. I'll be more convinced when those tools start producing levels at the pace the game needs, which I come back to at the end.
Rotation changes the level design
Cube Land comes from the This is Blast family, and I'd place it closer to This is Blast than to Pixel Flow. There's a queue of color shooters, a small bench, and a block picture you clear by firing shooters in the right order. Pixel Flow added a conveyor belt for the cannons, while Cube Land has none. It turns the picture into a 3D voxel object that the player rotates and adds the layered structure we saw in Screwdom and Yarn Loop, where each model is cleared from the outside in.
Screwdom made a similar move when it took Screw Jam's flat board and turned it into 3D objects. With an object, a level designer can hide cubes behind other cubes, spread one color across several sides, and build inner layers that only open once the outer shell is gone. The player makes two decisions per move: which shooter to fire and which side of the object to face.
Most sort games released this year only asked for the first one. Turning the object to line up a shot also feels much better in the hand than tapping at a flat grid, and I'd put Cube Land next to Screwdom as one of the cleanest iterations this genre has seen.
The blockers are standard for the genre, including hidden and linked shooters. The layering does most of the new work, and I expect later levels to lean on it much more than on new blocker types. Sort puzzlers have never needed long mechanic curves. A big part of the pleasure is the object itself and peeling it apart layer by layer, so I think the choice of what goes on the block matters almost as much as the mechanics do.
The fail condition is the design detail I like best. Most games in this family end the level when the bench fills up. Cube Land ends it when the bench is full, and none of your shooters can reach a visible cube of their color. Before that point, the game rotates the object by itself to search for a valid shot. Players who rotate with intent still clear levels more efficiently than the auto-spin, but nobody loses because they missed a single blue cube on the far side of the model.
The difficulty sits in the queue
Rotating the object is the part of Cube Land that feels most like skill. When you look at how levels are actually lost, rotation is rarely the reason. Levels are usually lost to the shooters you can't see yet, deeper in the queue. An early pick from the wrong column leaves the colors you need buried, and a few moves later, the bench is full.
This setup suits a hybrid game well. Players act freely in the parts they enjoy, and the designers tune the difficulty in the parts players can't fully read. When a booster or an extra bench slot shows up at that moment, it feels like help, which makes it easier to buy.
It’s easier to get through the first hundred levels without much resistance. I think the easy stretch is intentional. Puzzle players who are going to pay tend to do it within their first couple of hours, so it makes sense to keep them moving through levels during that window and let interstitials take over after it.
With an ad after nearly every level in that stretch, the early game carries a lot of the ad monetization. There is a cooldown, so quick consecutive fails don't each trigger an ad, and ads start at different points in their progress, which suggests segmentation by UA source or country.
For a team that scaled this quickly, level QA matters as much as tuning. The mid-game churn curve is the number I would most like to see from this game. Pixel Flow gets the same fairness complaints, so this is partly a genre problem.
Revenue, UA and creatives
Over the last 30 days, Cube Land is at 1.3M downloads and $5.2M in IAP revenue worldwide.
That's around $173K a day from IAP alone. Downloads have come down since September, but revenue is still climbing four months after the global launch. Average DAU for the period is 308K, and players spend 46.7 minutes a day in the game, a lot for a puzzle title built on short levels.
You get about $0.56 per daily user from purchases alone, which is strong for this genre. Add ads on top, and with an interstitial after nearly every level, I'd put ads at close to half of total revenue, which takes the game to around $350K a day.
Rotatelab ran about 200 to 250 creatives over a 30-day window while scaling. That's not a high number for this level of spend, and it suggests the game converts well without heavy creative volume.
Content and clones
The meta is still light. There's a crown tournament, a Sky Lift event, and a milestone rewards track. Games in this genre that last usually end up with thousands of levels and daily, weekly, and seasonal events running on top of each other, all fed by level wins.
A collection album built on those layers would fit this game naturally, and I think live ops is where the next big jump in spend comes from. The level count has to grow with it, and if Rotatelab's AI tooling is as strong as they say, that's where we should see it first.
The clone wave is the other pressure, and I think Rotatelab got lucky on timing. When Cube Land started scaling, most copycat studios in Vietnam, China, and Türkiye were busy with another hit, and that gave Rotatelab a few weeks to scale without direct competition. Adding rotation to a blast game will take a competent team a few weeks. Matching Rotatelab's speed on 3D levels, 3D creatives, and a live ops calendar will take them a lot longer.
One Hit or the Start of Something?
I've seen plenty of puzzle breakouts fade within a few months, usually because the studio behind them had one good idea and nothing to follow it with. Or they just copied exactly what was working.
Rotatelab has close to a decade of launches behind it, a self-publishing setup built the hard way, and a great CPO, Ahmetcan, a friend of mine who lives and breathes this market.
They still need a level pipeline and live ops depth that can keep a game like this in the charts for years, and the next six months will show whether they can build that before the clones catch up.
Either way, a remote team from Gaziantep putting a game into the US top-grossing charts is good news for everyone making games in Türkiye outside the usual cities. I'm rooting for them, and hopefully we'll see more hits from them soon!
