itch.io

It depends on which growth engine i decide to use (Sticky, Viral or Paid), basically they are 3 different ways to improve user acquisition, they all work in different ways and to know if they work, they need the developer checks 3 different groups of KPI... for example at the moment i'm assuming my game requires the Sticky growth engine, which relies upon those KPI: retention, churn rate and compounding rate.

After 8 months is evident that my game is not selling decently and not even engaging with the demo (tbh i don't remember how many wishlists it has at the moment, but it's not relevant: wishlist alone is just a vanity metric, if not supported from actual numbers).

What it's less evident is the specific reason why it's not selling/engaging, from the outside for example it might be because i'm skrewing with marketing... and that's true as well, but on a closer scope Steam said some time ago something more accurate, with the KPIs collected from my tiny early numbers: 70% of those few players dropped after 10/less minutes, besides the retention after 1 hour of gameplay was below the average... in other words my retention KPI sucks.

Is it because my game sucks? Maybe, but the answer isn't necessarily yes or no: it can also mean that in the first 10 minutes of the game something silly happens and that makes players stop playing... and that was true, in the very first releases i put a scripted story event which would enforce the end of the run (permadeath), in order to show a specific story and to close the tutorial run: it wasn't a good idea, since the game is quite tricky and minimal at the same time, 10 minutes of gameplay are not enough to introduce the player enough of those unique features, so the players probably wouldn't be convinced from just that first run.

After noticing it i removed that custom event interruption and it will happen only when the player will actually lose. Did it work? I need more datas to know the answer but for example a (probably quite small) streamer played it for maybe around 1 hour (i'm assuming that because he told he tried 5 runs, heck i hate the video on twitch wasn't highlighted... i missed a huge change to study an actual unknow player's playthrough), so maybe it's a good sign, we'll see. Anyway he still stopped within 1 hour probably, and mentioned that he didn't perceive enough progression while iterating more runs.

A friend of mine (i didn't asked him to play) actually enjoyed a lot the very first release and played it for maybe more than 20 hours, i actually had to stop him since at that stage it was still missing lots of features that i wanted he could experience. He told me a quite signicant thing: for the first hour he admitted he played the game because it was my game, but then he started to grasp the core gimmicks and understood what to do and where to go... and that's the point: i need to work on the first hour of gameplay, and that, said in another way means the same thing: i need to work on retetion (more precisely first hour retention's range), even my publisher told me that.

And that's the importance of KPIs: default analytics of itch can't tell me that degree of detail, on the other hand Steam told me that information, but it's not a one-time event: you need to keep monitoring your relevant KPIs and understand if they are improving after new changes.

I could ask to actual people who played the game on a fair event, in 2 days about 32 people played it, but the permadeath gimmick cannot be tested in such events, since almost 100% of times a permadeath means the player will stop playing.

I might even ask on forums about sharing opinions about the game: as a matter of fact some tens of people downloaded the game but no one wrote to share opinions, let alone sharing about the fact that after 10 minutes they would stop playing... and that's quite obvious, why would they have to lemme know that? I won't blame anyone for that, i wouldn't to that as well.

Fortunately, analytic datas can speak for players: even if tens of people played the game without telling me that, their analytics told me that issue about retention... not like it's the only one, but there's no doubt that KPI are a powerful tool that devs should learn to use.

Assuming i could spent a significant amount of effort, time and money on Ads, if did that some month ago, that retention would never allowed the demo to convert the player into the full version (purchase), so i would have wasted tons of resources without understanding that problem.

By the way aren't in-game ads supposed to be used for Mobile games? The game is Premium so i'm afraid it cannot work, one day we'll make even the mobile version but that's another stage of developement. No Dlc for now, i should wait at least until the base game starts to sell decently. On a side note shouldn't Ads surely imply that you have to deal with GDPR, PIPL and CCPI?

About the A/B testing, even me i thought the first time i learned about it that it was supposed to be used only for mobile and monetization, but that's wrong: it can (and should) be used for your MVP as well (in my case the released game which keeps getting updated), because it actually works for that as well! What you test with the A/B testing is how KPI and funnels behaves differently among the two versions of the test.

While i mentioned statistics of Steam and i know they are supposed to be used for the achievements, actually i was talking about another use of them: while it's true that you can create specific achievements to create a sort of Funnel, that will work only for the full game. The problem is that, especially in the early stages, what you want to track is the demo, which is the closest thing similar to your MVP, and it's bound to have much better gross numbers than full version. Steam Statistics on their own are simply custom variables, until you match them with achievements they won't even talk among each other. The actual interesting property is that you can setup such statistics in order behave as aggregate global datas which can be tracked from Steam as you can for their analytics... and that's how you can setup a sort of funnel for the demo as well (it doesn't even requires extra assets like GFX for achievements), i'm gonna try to make them work in the next days.

Funnel is probably even more important than your KPI: they are the steps which lead to user acquisition (purchase), if you manage to learn where the player is gonna decide wheter to purchase or not, then you can setup a list of micro-steps of gameplay which leads to that point... something like 1 - The player watches the intro; 2 - The player finishes the first dungeon; 3 - The player finds the second artifact; 4 - the player triggers the first boss event; 5 - the player reaches the end of the demo. Does the players accomplishes all those steps? Does the funnel have an actual match with purchases? With analytics you can convert those steps in % values and improving them after lots of A/B testing, and/or changing the funnel on the way.

Of course, A/B testing to be effective should be done with new users, but that's another problem to handle and is rather about the page analytics (that luckily even itch can offer)

Wait they are not indexed? (To be honest i don't even know if i should be surprised or not, i don't know causes and neither implications of indexing) That probably explains why they cannot be found on the search bar. Yeah it's The Rogue of Nexus game what i'm talking about but i didn't want to use this thread to talk about it (it was about analytics in general, but i sticked with that to explain better about KPI and funnel).

I should learn more about that web version you're talking about, maybe it could be a good idea for the demo, i don't know... so you can collect different datas than default ones and in that case without GDPR, etc. getting involved?