Kinda interesting how A24 has built a brand/audience relationship more akin to that of a youtuber or twitch streamer, even though they’re a pretty major film studio. Now they even have a quasi-patreon membership thing. Makes sense since their films largely target millennials.

Within games, Annapurna Interactive could probably pull off something similar.

If you’re an indie game dev, trying to build a brand and audience relationship the way they do is probably going to be very helpful i think.

So you’ve launched your game on GOG

It’s been a month(ish) since the release of The Signal State on GOG, so it might be interesting to look at the sales data and compare it to our Steam sales data.

The total number of copies sold through GOG is approximately 1% of the total number of copies sold through Steam, with the vast majority of copies sold during the launch week in which we had a 20% launch discount.

This is a pretty absurdly low number! Data on GOG sales vs Steam sales is even harder to come by, and I’ve seen estimates for the GOG vs Steam sales ratio of anything from 2% – 15%. Ours is clearly on the very low end of this.

There are probably a variety of factors for this figure beyond Steam’s dominance. Our GOG launch was not simultaneous but came months after the Steam release. We did not promote the GOG launch particularly heavily. The lack of Steam Workshop support means the Puzzle Workshop in the GOG version of the game is clunkier in the way it works compared to the Steam version.

That lack of feature parity with Steam’s Steamworks backend services might in fact be the main deciding factor as to whether to release on GOG or not (assuming your game is notable enough for GOG to invite you to release it there). If your game relies on Steam’s backend infrastructure for anything more complex than leaderboards or achievements, you might find yourself having to find a whole variety of workarounds for GOG’s more suite of backend services.

Does your game use Steamworks for mod support, microtransactions, voice chat? If no, then a making a GOG build is probably quite easy. The GOG API for user authentication, leaderboards, etc. is pretty easy to implement. But if you need anything more advanced, you’re gonna have to find alternative solutions for your GOG build.

And if you’re not going to make much money off of the GOG release, is the time and effort even financially worth it?

So your game is Steam’s Daily Deal: Conclusion

The Signal State’s daily deal promo will be over soon, and the sales from that have actually exceeded my expectations.

With the exception of release day, the daily deal has resulted in the highest single day sales ever, with nearly 14% of all copies sold having occurred during this promo, nearly exceeding the copies sold during the winter sale (and it might exceed it by the end of the promo).

That said, there are other additional factors involved of course other than the visibility from being part of the Daily Deal. This deal is for 25% off, our highest discount so far. This also follows the release of our Puzzle Workshop, which is visibly promoted in the artwork shown on Steam’s frontpage.

So your game is Steam’s Daily Deal

Its been a few hours into our Steam Daily Deal promo, here’s what the sales figures look like:

Getting onto the Daily Deal requires manual selection by the Steam team, which probably involves some invisible metrics your game must fulfill. Given its short promo time its probably not going to be as lucrative as say the winter sale, but its very nice nevertheless.

So your game is discounted during the Winter Sale

The wishlist-to-sales ratio can be wildly different from game to game, with anything from less than 0.1 sales per wishlist to more than 1 sale per wishlist.

If your game ends up with a lower ratio than the widely publicized statistical average, it could potentially be quite disheartening, but it need not necessarily be. After all, a wishlist is an indication of some interest, and these wishlists usually bear fruit during Steam sales.

Copies Sold before and during the Steam Winter Sale

For The Signal State, the average number of copies sold per day during the Winter Sale was 6 times that of the daily average in the month prior. 17% of the total number of copies sold were bought during the sale, accounting for about 14% of our total revenue.

Its also worth noting that people will still wishlist your game during these sales, adding in more potential buyers for future sales. Our daily wishlists remained roughly the same through the sale.

Other stats like our positive reviews percentage, percentage of buyers on Mac/Linux, percentage of refunds remained mostly unchanged.

So your game has been out for a month

Its been slightly more than a month since release, so it might be useful to look at how The Signal State has done.

Our reviews stand at 92% positive right now. This ratio stayed pretty stable after 2 weeks or so from release.

Our refund rate is at about 5.6%, a slight increase from the rate after 1 week, which makes sense as players who buy later are players who may be less passionate about this kind of game, but I’m still pretty happy with this rate. As mentioned before, the average refund rate is about 5-8%, which puts us on the lower end of the range.

Our wishlist to sales ratio is now at about 18%, and a conversion rate (users who first wishlisted then eventually purchased) of about 7%, a slight increase from ratios at the 1 week point.

The current number of copies sold is about 47% higher then the number of copies sold after 1 week. This is pretty much in line with expectations from existing data from other games.

If you consider the current profits and the publisher advance (minus external expenses like freelancers), this game has earned enough to sustain at least 1 full-time developer for at least a year, possibly longer once we see the revenues from future sales!

P.S we just released an update that adds some requested UI improvements please buy my game and givememoneythankyouverymuch

So you’ve sold your game


It’s been a week since launch, typically the most crucial period that decides whether a game is a ‘success’ or not, so it’s an interesting point at which to review how we’ve done.

Overall, the game has performed better than I, or even our publisher, had expected! We recouped our development budget in about 3 days. The game is now profitable 🙂

We have a wishlist to sales ratio of about 16%, and a conversion rate (i.e wishlisting users who proceed to purchase it) of about 5.5%. These numbers are pretty low, despite our good sales as the vast majority of wishlists came post-release from streamers/influencers and press coverage. In other words, the conversion rate is low as the people who would’ve bought t immediately have already done so, bypassing the wishlist process.

What this also means is that our pretty good absolute wishlist numbers could translate to good performance during sales periods when the game is disounted, as we have a large pool of potential buyers, likely waiting for those discounts.

We currently have a total of 104 user reviews, 100 of which are positive. That puts our sales to reviews ratio somewhere between 25-50x (I’m not giving a specific number as you could calculate our sales figures from that, and I’m not certain if we can give our actual numbers out).

6.5% of players are playing on Mac. About 1% are playing on Linux via Steam Proton. While we don’t have a official Linux build players have reported that the game runs perfectly well through Proton. That said, the low player counts on Linux suggest a proper Linux build is not worth the effort for the vast majority of games. While a proper Linux build might result in more copies sold, you could also argue the niche nature of our game means the number of Linux players in our playerbase is above average anyway.

4.9% of sold copies were refunded. General consensus is that the average refund rate is about 5-8%, so we’re a tiny bit below average here.

44% of revenue came from North America, 27% from Western Europe and 17% from Asia. Note that this differs from the % of copies sold due to regional pricing, e.g more revenue came from North America despite the % of copies sold in North America being lower.

Based on Achievement stats, 86% of players completed the tutorial, 7% of players completed the game and 3% of players 100%ed the game. The great thing about our per-level leaderboards is that they also function as user analytics, giving us a pretty granular data about how far individual players have gotten through the game. If your game design approach is more data-driven (mine isn’t), this data could really help figure out how smooth your difficulty curve is.

The conclusion that I pesonally draw from the game’s performance is that yes, making games that cater to small niches is absolutely viable. You don’t have to compete with the big premium indie titles if you don’t want to or can’t. Fans of niche genres can also be forgiving of some of the rougher edges of your game, or the above-average pricing, simply because they’re starved of content that fulfills that niche.

What doesn’t change is the value of marketing. Having a publisher helping you with this is incredibly valuable. Knowing who to market to is important as well, with a large amount of our outreach occurring due to word of mouth.

Marketing isn’t just a question of expertise, but time as well. During release you are going to be swamped with bug-fixes and tech support and feature requests. You won’t have the time to handle the numerous requests from streamers and Steam curators asking for review keys or updating your store page with press blurbs or setting up video broadcasts on Steam. Having a publisher absolutely helps.

Anyway, I hope this data and my past posts are somewhat illuminating.If you’re looking to go indie hit me up 🙂

So you wanna release your game on Mac

In one of my older posts I made a joke about the tiny amount of hipsters who still play games on mac, so now that we have *some* sales data from our release, it might be worth asking, how accurate is the joke?

Mind you, this is only the launch day’s worth of sales data so things might change. Furthermore, the nature of the game can skew the results either way. You could argue that the programmer/engineering nature of the game skews it towards PC players. Conversely you could argue that its synthesizer/electronic music nature skews it towards Mac players. Maybe it just balances out?

Of all the copies sold so far, 5% were bought by Mac users.

Now, some of the copies tagged as PC buyers could also own it on Mac, but either way 5% doesnt look like a lot. Does it make it worth the hassle to release on Mac? It obviously depends! If you’re working on Unity/Unreal a Mac build is probably pretty trivial to make without much changes to your codebase.

That said the financial impact is still worth considering, especially as a small-scale indie dev. You’ll need Macbook/iMac/cloud virtual machine to test and notarize your Mac Builds (newer Mac OS X versions will not run un-notarized apps), which is a sunk cost if you don’t already own one. Obviously this is a negligible cost if your expected revenue is in the millions of dollars, but if its in the thousands or tens of thousands of dollars this is a cost to consider.

Mac support still also means another OS to support. If you are using a custom engine or your game involves a lot of lower level code, then the change in OS may mean more complications during development and post-release. Maybe the money isn’t worth a headache. One instance of this we experienced was the GIF recorder in The Signal State. This feature involved pulling frames of the game from the frame buffer and encoding them into a GIF. This broke on newer Macs due to the new Apple Metal graphics API. Fortunately for us the fix was to simply force Mac builds to use OpenGL, but I can imagine this potentially being an issue for more graphically sophisticated games.

So you want people to wishlist your game

If you’re a smaller-scale indie dev, “the algorithm” can be very important to your game’s success.

The Signal State made it to the ‘Popular Upcoming’ list on Steam’s frontpage (don’t ask me how), and this is what our daily wishlist graph looks like now:

For comparison, the first spike was from the playable demo during Steam Next Fest.(this is a linear scale graph)

So you wanna export your game

Ok brosquito! you’ve finished localizing your game and now its time to release it. You gotta make the builds.

Simple right? You gotta make a pc build and a mac build.

WAIT. new mac os x versions wont run apps unless you notarize them, so after you make the mac build you gotta notarize them, which can only be done on a mac, because reasons, so you do that.

oh yeah dont forget to pay apple 100 bucks for a developer account for the privilege of making games for the 7 hipsters who play games on mac.

WAIT. you also released free demo versions of the game, so you gotta update those builds with the recent bug fixes.

Ok, simple enough, 4 builds, 2 of which have to be notarized on a mac.

WAIT. the chinese regional publisher is gonna get your game onto the wegame store, so you have to make builds for them. These builds must not have links to/integrations with discord/twitter/steam/etc. since those are banned in china.

ok, so you write some special code for the wegame builds.

WAIT. the chinese publisher also handles publishing of the game on steam in china, and they have a qq chat group for players. so chinese players on steam should be directed there instead.

ok so to recap:

players in other languages on steam should see steam/twitter/discord integrations.

players in chinese on steam should see steam/twitter/qq integrations.

players in chinese on wegame should not see any social media integrations.

ok you got that all done. time to make the builds.

you need a pc steam full game build.

you need a mac steam full game build, notarized.

you need a pc steam demo build.

you need a mac steam demo build, notarized.

you need a pc wegame full game build.

you need a mac wegame full game build, notarized.

you need a pc wegame demo build.

you need a mac wegame demo build, notarized.

So you need to make *8* builds.

Sigh, so you sit down and make the builds and twiddle your thumbs while unity does its thing.

ok its done and you hand the builds over and you’re done!

WAIT. The chinese publisher just spotted a typo in their localization. Time to fix it and make 8 builds all over again.