GameDiscoverCo's Founder Simon Carless talked about the company's prediction models, what they are based on, and how available data can help indie developers to sell their games.
My name's Simon Carless, and I'm a 25-year veteran of the video game industry. I've had a very wide experience of different sectors of the game biz, since I started as a Game Designer and Team Lead on the developer side of things, working on games for the first PlayStation and Xbox at companies like Kuju Entertainment in the UK and Accolade/Infogrames in the U.S. From there, I ended up working on Gamasutra/GameDeveloper.com as the EIC and Publisher and then overseeing Game Developers Conference and the Independent Games Festival from a biz perspective for many years.
Having been very close to the independent games scene with the IGF, and as the co-founder of the Independent Games Summit at GDC in 2007, I've always been very interested in the indie games scene, and in recent years I was the initial investor and am still an advisor to UK publisher No More Robots (Descenders, Yes, Your Grace, Hypnospace Outlaw). So partly due to that, I got involved more behind the scenes in further understanding the state of the games biz and started working on a newsletter about game discovery a couple of years ago, leading me to founding my current company, GameDiscoverCo.
So, GameDiscoverCo is described as "a new agency based around one simple issue: how do players find, buy and enjoy your premium PC and console game?" It's designed for people making regular 'pay once' games on Steam and consoles – not so much free to play or mobile – because I found that there wasn't a great deal of great analysis or research in this area. Whereas mobile and free-to-play often have lots of in-house jobs around monetization and market analysis, there isn't really an obvious information source for all the other platforms.
GameDiscoverCo is a boutique agency – i.e., it's intended to be mainly structured around myself and other contributors, which include a lead developer working on our GameDiscoverCo Plus back end (which grabs information on pre-release information for Steam games!) and a data analyst who helps me put together charts for other platforms that are less easy to grab information from. The main things that we do are:
The GameDiscoverCo newsletter: this gets sent out twice a week to free subscribers, and we sometimes reprint/syndicate highlights via sites like GameDeveloper.com and GamesIndustry.biz. It's focused on things like:
- printing real-life sales, wishlists, and trend data from actual Steam games
- analyzing trends and ways to make your game more successful and get more Steam wishlists
- the general state of various game platforms
Consulting in the space: we don't proactively advertise, but we have a number of clients – often mid-sized publishers – that we help with game discovery in general: for example No More Robots, Dotemu, Akupara Games, and about 5-10 others we help regularly on things like evaluating possible games to sign, discovery strategy in general, timing on releases, pricing, and so on. (We don't do marketing and PR directly, but we talk about strategy and approach.)
The basic goal here is – game discovery is a really complicated concept, and you have to start it VERY early in development. It's not 'just' marketing and PR, it's a holistic view of why people might be initially attracted to your game, and how you get them to become fans before it comes out. With an increasingly crowded market, it's really important to think about things like 'game hook' and community in a much more integrated way. So that's what we try to teach!
Helping Indie Developers
It's very hard for independent developers to break out nowadays! You need to have a game that has a real visual 'snap' to it, you need to be working in a genre or subgenre that isn't overcrowded, and you need to be realistic about putting time in before the game is released in discovery, community, and marketing. You can't just sit there and expect people to jump on board and get excited about your title. (Although it does happen sometimes!)
So you need to think of things like managing a Discord to talk to your fans, having regular alpha or beta builds or demos in private or public to show possible buyers, ways to get your game out there on social media that makes a big difference to interest like TikTok, Reddit, Twitter, and beyond, and just in general build up things that mean your game has a lot of interested parties at launch.
If people find out about the game and tell their friends, that's great – but if you only have a few people like that at launch, it's incredibly difficult to build up momentum. So we try to expose various ways that devs can be proactive – or work with publishers who are proactive on their behalf – to really prime their game for success.
Sources, Data, and Models
There's a lot of public data available via Steam on PC, fortunately! So we track all unreleased games in our Steam back end and we use data like a number of Steam followers, Steam unreleased game wishlist ranking, posts on Steam forums, 'track record' of the publisher, and some other data we manually collect like Discord server population, Twitter followers to make an overall ranking for all 5000+ unreleased games.
We also manually track Switch, Xbox, and Epic Games Store charts for new games, and have some plans to expand into post-launch Steam data analysis too.
We also do a lot of public surveys to ask other devs for their experiences on Steam, and then print the results for everyone! For example, we asked about first week sales compared to launch Steam wishlists and have the results here – all this type of public information is very helpful!
Separately of data, we compile a lot of information on best practices from all over the web. So if there's an interesting talk, graph, or other sources of information, we keep track of it and put it in the newsletter. We think niche newsletters are a really helpful way of keeping track of a great deal of information in this super complex field, which is why we enjoy writing and sending them out so much!
So this spreadsheet was part of two spreadsheets we released at the same time – the other one was from the indie publisher Fellow Traveller. You can check out both of them here. As for how accurate the prediction model is – frankly, it's 'indicative' more than accurate. As we explain:
"There are just two things you can change without editing formulas: the number of Steam wishlists at launch, and the average global price in USD of your game. From there, the spreadsheet makes a bunch of extrapolations to work out: Week 1 Steam sales (as well as Day 1 Steam sales as a bonus); Week 1 Gross & Net Steam revenue (presuming you’re not in a country that has 30% tax withholding from the U.S.!); Year 1 console net revenue, Year 1 total net revenue, and then 3-year Steam revenue, 3-year console revenue, and 3-year total net revenue.
So the example here: a game with 10,000 Steam wishlists on launch that converted its first week at 0.2 sales per wishlist (so 2,000 sales in its first week) would end up making a net of $222,290 across PC and console over 3 years. (Presuming you discount it, and all that good stuff.) The biggest gotcha here by far is ‘Week 1 sales compared to launch wishlists’, as we constantly mention - our data says it can be as low as 0.03 and as high as 1 sale per wishlist. So tweak the formula for a variety of scenarios. Also: this particular spreadsheet is a bit more conservative on Week 1: Year 1 revenue ratio (which we put as 3x) than our actual surveys. And console revenue vs. PC is just a total handwave. But if you want a super easy look at possibilities – here it is!"
So a lot of the information is quite accurate – for example, 'gross to net' Steam revenue is something we've written about before – and varies based on refund rate, but not a lot. And then Week 1 to Year 1 Steam revenue also varies quite a bit – from 2x to 6x, let's say – but it's still somewhat useful.
Where this really breaks down potentially is in the 'Steam wishlists to Week 1 sales', since our survey data suggests this can vary by as much as 10x to 20x! So with 10,000 wishlists at launch, you could sell 300 copies in Week 1 or 6,000 copies. This is why the Fellow Traveller version of the spreadsheet includes all kinds of possible scenarios. How you do is dependent on the quality of your Steam wishlists and the natural 'convertibility' of the genre you're in – and also if the game gets picked up by streamers, etc. There are a lot of variables. But the point is, you can do SOME planning based on this type of info!
What Should Indie Developers Focus On
Breaking it down to just a few things new indies should do, I would say the following:
- Think very carefully about what game genre/subgenre you want to be in. There are some areas that are more sales-friendly than others, and there's provable data on median revenue for Steam tags to show this.
- Design the game for discoverability! Does it look great – or interesting – if you just see a GIF for 5-10 seconds?
- Is it a game that 'looks' deep or interesting? You really need that upfront hook to interest people with so much competition out there.
- Announce early! Have information and a video trailer out for your game for 12 months or more before you want to release it.
- Build a community from the very beginning, and nurture it with alpha demos, regular dev updates, and lots more.
- Be smart about being in showcases (announcement events, Steam Next Fests, other Steam pre-release features!) This can make a major difference.
- Put time into actually marketing the game (via social media) and reaching out to streamers too. This can really pay dividends.
I know this sounds like a lot more work than just announcing a game and then releasing it, but it's all really important!