This is always a complex question, rarely with one simple answer, so I'll try to put in a way that makes sense.

Publishing is a hit-driven business: most profit always comes from a small amount of currently operating games. Hypercasual publishers have a lot of somewhat-earning games with low margins but it wasn't our case (also Pareto principle works with HC too and 80% of profits are generated by 20% of games). In our case, one true hit might return all investments or at least earn enough money to live up to the next and possibly greater hit.

So, the actual question is: — Could we have made a hit? In fact, we almost did: all in all there were 6 ROAS-positive soft launches, none of which didn’t happen as a hit.

— But is this lot or little? Not sure, but I know at least some profitable competitors that have somewhat close numbers and also know a few closed/unprofitable all of those had less. So I guess it's not that bad.

— Another point of view: how close we were to a hit?

Statically speaking our chances to get a hit were already 70% (by the complement of independent probabilities: 1-(1-0.175)^6 ). Not that bad, huh? Here comes the temptation to say: — “Wow, we did it all right, just wasn't lucky” — but we won't take it for an answer :)

Instead, here are the things that could have changed the outcome for the best:

First, our chances of test-to-hit weren't that high, at least for some of those cases — because the opportunity to launch and scale them appeared when it was already too late.

For example, in one case, after a few months since launch, we came to a point where conventional channels stopped performing. It was a complex midcore game, so there were more ways to make it profitable: high-cost entrance DSPs, influencers marketing, combo UA+ASO misleads-based campaigns, and so on — but any of that is expensive and long-shot, meanwhile we were already on the verge of bankruptcy, so we honestly told to studio that it's better for them to try with another publisher.

And the other case was a profitable game that still might happen as a huge hit — yet developers terminated our agreement. There was a moment to reconsider the contract and I've failed to explain our value. That time we've already run out of resources and I suggested a third-sided deal with another publisher, thus refusal was expected (yet painful).

But those examples are not some kind of misfortune — it was clearly my mishandling. In the first case, we could have tried: for that, I should've raised more money and replenished UA team with new specs, but I weighted chances versus efforts and decided it wasn't worth it. I might have been right, though I can't say “I did everything”.

And in the following case, I just missed the moment. Two months before ‘divorce’ there was a time when both our and dev's teams were excited about promising test results, also our internal problems were yet on the surface and I should have brought up agreement prolongation then.

Let's resemble the chronology:

And now let's aggregate by periods:

The point is: two years passed from the start to the end, but hardly half of the time was effective. Assume there wasn't a breakup with investors; OR no Putin's war; OR I've raised enough for a whole year w/o cuts — any of those assumption might have been a game changer able to provide us more than 6 promising games.

Sure, also I should have better handled the crisis: stayed in operations, raised team morale, and made hard decisions at the best moments — but I'll give myself the right to disagree. No, I couldn't done better. Might be an option for someone else, but I tried my best, it just wasn't enough.

I suppose it's the most exciting question for many readers:

Or speaking cold logic: how was it possible either to increase N(umber) of causes (tests), or each (P)robability of success?

One of my essential functions everywhere I work is a bullshit meter: I'm usually the first who told “Meh, it's bullshit”, “it's not gonna work”, "why we spend time on this" and so on. Although we had a strong team - regrettably, no one could replace me in that. That's why when I returned as a CEO after almost half a year of resignation — I almost broke my hand in facepalms, reviewing what trash we were working on all this time (got rid of 3/4 by next two weeks — actually, not all projects were that bad, but our scope shortened cause of resources so we were forced to cut even some of what we liked).

Maybe, before that I wasn't pushy enough either. I myself was sometimes in a trap called “it keeps getting better” though I should have known it would never come to "good enough”.

Also, in the previous product team I built, I decided to let people learn from their failures, so I didn't insist on immediately abandoning 'bad' projects, instead, I waited for the producers to get that by themselves (so next time I won't be needed). It went well that time but failed in 10K probably and I'm not considering this as a good practice anymore: currently I think it's better to create such a process that highly motivates the product team for harsh selection of its products — and we established one in the second half of 2022.

Another underperform of ours was in the lack of documentation, references deconstruction, balance reverse engineering, level design and so on. We've done all that, but late: often we wait till some studio delivers us a promising-looking yet somewhere screwed prototype and only at that point we did dive into balance, deconstruct, etc. That was based on the assumption (and prior experience) that a studio unable to produce a nice playable proto based on reference — is ultimately unable to make a performing game, so it's okay if they fail.

More practice showed I was wrong: we received enough well-performing prototypes that failed specifically regarding numbers, balance, and other sides of the design. Plus when we did have docs in prior — the average quality of results has been higher. Probably we should make it more. Though it dramatically increases the price of each new idea we share with studios.

So yes, with the benefit of hindsight, we could get more successful launches just by doing better from the start. But I can't say for sure that it could really change the game like if there were no problems with money and as a result no periods of ineffective survival — but there were things under my control so it would be unfair not to mention and reflect on them.

We did make mistakes. I personally did mishandle some and lost opportunities when we needed them the most. But unlike most businesses that ended up just like we did — it wasn't predefined and expected failure:

Considering everything above it should be obvious to ask why don't I try to raise more money and start over again? Fair question.

The answer is simple: I'm sick and tired of startups and building something new from 0 to 1. I did that several times in 10K only: 1) when the company was founded, 2) when it was rebuilt after the second investment, 3) after staff cuts and focused on marketing & liveops only. Before that, I did the same in AppQuantum, headed the company in 2018 while it had two junior employees (they were great though), and led to a thriving business. And before that, it was ARPU.GURU, a small startup founded as a side project by me and Elena Bugakova then came to be one of the first gamedev consulting agencies with hundreds of experts involved.

So despite that one failed, I have my right to fuck it all and target something already working to grow it from 1 to 100 and beyond. My contacts are below :)