Cultist Simulator: the retrospective

I talked at Develop this year about open production, a system of developing games in the full sight of your community and building your audience as you go. It reminded me that I hadn't shared the Cultist Simulator retrospective as widely as I should have, and that the successes (and mistakes!) might be useful to indies who haven't read it yet. So, here we are! Gama, spread this cautionary tale.

It’s worth mentioning that there are lots of things – both good and bad – which don’t get a look-in, because they’re not the headline points. We’re absolutely delighted by Maribeth Solomon’s work on the soundtrack, for example, and I’m very proud we didn’t move project goalposts even when we were tempted. But I don’t want to get stuck trying to tell you everything: below are the top-line woos and boos of Cultist overall.

☽★☾

☽★☾

1. THE KICKSTARTER

Alexis ran this back in September 2017, when he had a playable alpha and a clear vision for the final game. We asked for £30,007, were fully funded in 12 hours, and were 273% funded by the end with nearly 4,800 backers.

The numbers here are great for a small indie game, especially one planned for open production. But the campaign was also designed realistically, thanks to Alexis’s prior experience crowdfunding games like Sunless Sea. Cultist Simulator kept things simple, promising only that which a two-person team of our specific experience and skill could deliver. No complicated digital or physical rewards. No ill-thought-through stretch goals adding unknown complexity. No promises we hadn’t run through the Yeah But Can We Do This 10 Months Down The Line When We’re Really Tired And Are Sick Of The Whole Thing test.

We probably could have squeezed more cash out of a Kickstarter if we’d set more complicated stretch goals, but I’m very glad we didn’t.

2. OPEN PRODUCTION

This is about being open with your community, delivering consistently and incrementally, and incorporating feedback ALL THE TIME. If we hadn’t developed Cultist Simulator like this, it would have ended up incomprehensible and un-fun. We would also have produced a buggy, unplayable mess, because we’re two people with no dedicated QA resource and the fact that Cultist was being played all the way through development absolutely saved our butts.

Cultist Simulator’s open production highlights include:

3. DEADLINES

We met every single one of our project deadlines: sprints, major releases, publisher commitments, the lot. This may sound petty, but there’s a lot of hidden goodness here: we didn’t end up with a backlog of stuff we’d pushed back from earlier deadlines we wouldn’t otherwise have been able to hit; we didn’t have to cut large parts of the game to hit launch; we were able, early and often, to drill 31-ST-OF-MAY-20-18 into people’s heads.

It also meant we kept Cultist Simulator’s scope and budget small. All in, we spent £142k making the game, which was our original budget + one conscious and deliberate 10K increase in expenditure once we’d signed with a publisher. I’m really proud of us here.

4. LAUNCH

We launched on time, and after a half hour or so, somehow appeared as Steam’s #1 overall top seller (only for a couple hours, of course, but still!). We were featured on the front page of all storefronts (Steam, Humble, GOG and itch) and hovered somewhere in the top ten sellers and new and trending lists across all stores for the first week or so. We’d set pessimistic hopes of shifting 10k copies in year one, with our most optimistic projections at 30k copies in year one if everything went really well. At time of writing, two weeks after launch, we’ve sold over 40k copies across all storefronts, and have seen wishlists jump from ~14k to ~71k. We’re astounded.

Bearing in mind we definitely happened to hit a broad, spangly vein of good luck here, it’s going to be my job over the next few weeks to pinpoint what (or even if!) we did anything specific that had a tangible effect. It was something in the alignment of the following points, which broadly had two aims between them:

Aim 1: getting our game seen by consumers (NB: not fans, not devs, not social media – consumers):

Aim 2: getting people to buy in a cluster at launch

I gave myself a crash-course in marketing when I joined Weather Factory. The two most specifically useful things I found for launches, both of which shaped the above list, were Sam Coster’s ‘Marketing Judo’ GDC talk, and Lauren Clinnick and Matt Trobbiani’s ‘More than Luck’ Steam strategy video. Highly recommended, especially if marketing isn’t your strength!

5. THE GAME!

We set out to make a weird, experimental narrative game about yearning and apocalypse. Alexis wanted to combine mechanics and story like never before, specifically aiming to experiment with narrative structure and the effect of pithy text snippets as a story medium. We wanted to keep costs low so we didn’t have to sell a zillion copies to break even; we wanted to work with a publisher to amplify our voice and give us experience with publishers generally.

Cultist Simulator came out better (and more addictive!) than we’d ever hoped. From a creative and project viewpoint, it did exactly what we wanted it to, and then some!

☽★☾

Now, enough of the back-slapping. What did we bugger up, eh?

☽★☾

1. THE PRESSURE

There’s lots of work when you’re the only two people on a project full-time. Working with your other half in your shared flat is also, as you might imagine, quite intense. On top of that, we committed early to getting the game out by the 31st May, which meant we sat beneath a ticking clock for the whole of development.

This caused some tension between design and production, most of which resulted in a better outcome than either could have achieved on their own, but some of which caused some grumpiness between us. We wanted this constraint, because we wanted to make sure Cultiststayed small and experimental. But it did mean we had to section off a bunch of great creative ideas for post-launch consideration. The release build of Cultist doesn’t contain all the fun stuff we’d have ideally wanted, if we were operating in that magical wonderland of infinite money and time.

There was also a particular pressure on Alexis. He’d left the safety of an established studio with high-performing IPs, and his previous game had the whiff of once-a-career success. Now he was making a new IP with a team half the size of the one that made Sunless Sea. Could he do it again? Did he actually have talent, or was he just lucky? I watched my developer, and my fiancé, feel the pressure of proving his own worth. It paid off in the end, but it was tough on him as a human.

In the future, we’ll definitely hire additional tech resource pre- and post-launch so Alexis ‘only’ has to worry about design, writing and the game’s overall success. We may also invest in other areas. The moral here is: games are heavy, so spread the load where you can.

2. ONBOARDING

The most contentious decision of the whole project was whether to include a tutorial or not. Alexis was dead set against it: it was fundamentally at odds with Cultist’s design. Our publisher, myself and a bunch of beta testers worried that without a tutorial, people wouldn’t know what the hell was going on.

Alexis was right. Tutorials have been mentioned in almost every review of the game, but most of them say that they understand why there isn’t one, and approve of the decision.

But the testers, our publisher, and myself were right, too: there was a problem. We’d originally intended the last month of development to be spent adding additional Legacies into the game, but Alexis instead ended up having to go through the text to add clues. These breadcrumbs had to be obscure enough that they didn’t undermine the rest of the game’s design, but apparent enough to lead all players usefully from objective to objective. We also spent a bunch more money and time on improvements to the UI system, so these too would lead the player subtly but constantly through the game.

I think we succeeded within the constraints of a game set up to be deliberately mysterious. But we had to cut some things to do it, there’s likely a better solution, and we should have seen and addressed the issue earlier. Peeling back layers of mystery is a hallmark of an Alexis Kennedy game, so we’re liable to run into this problem again. We’ll see if we can do better next time, and definitely won’t try to fix it just one month before we release…

3. REVIEWS

Cultist Simulator was designed to be a niche game. It was not designed for mass-market appeal, but mass-market was exactly what we got when we went to Steam #1 and started getting front-page store billing.

The direct result of this was that our user reviews took a beating. (For the love of the Hours, if you like the game and fancy leaving a positive Steam review, please do!) We’re currently hovering at around 73% positive, which isn’t terrible, but certainly doesn’t reflect the generally very positive reception we’ve seen elsewhere. We hope this will trend upwards over time, as people find Cultist via genre filters and recommendations. But we’re not very happy about it right now.

I’m also in two minds about whether or not I should have given the game early to big publications like Edge in the hopes of securing top-tier coverage. On the one hand, printed press is prestige, and gets games like Cultist seen by audiences you can’t reach with budgets like ours. On the other, reviews really matter if you’re the kind of studio who cares about its Metacritic score, and one half of your company is the ass who got Eurogamer’s last ever 10/10. I decided to take the potential hit and go for it: we did get featured in some major publications, but the reception was notably colder than when we sent the game out closer to launch. We now range from ‘Recommended’ and scores of 90/100 to mediocre 7s from earlier reviews. It plops us at a 78 on Metacritic, which again, isn’t bad, but is lower than we’d hoped for.

4. FREELANCERS

Let me first just say: we worked with a number of freelancers on Cultist Simulator and they were all great. Our mistake was our freelance pipeline, not our beloved freelancers themselves.

Ten different freelancers worked on Cultist Simulator, and most of them doubled up in some way (we’ve had three UI/tech freelancers, for example, and four people created our in-game art assets). This surprisingly high number is primarily because coherent, consistent production didn’t happen until the Kickstarter in September 2017: prior to that, Cultist had been this weird tinkery side-project Alexis had been playing with in his downtime in between his own freelancing gigs. We started off with some freelancers who then weren’t free when we came back to them six months later; other freelancers had their own projects and personal commitments which meant they weren’t always on tap.

The end result is that we didn’t end up with the same level of consistency we’d have had if we’d commissioned fewer freelancers for the whole duration of the project. Next time, I think we’d be better off either committing sooner to ongoing, uninterrupted production, or waiting to work with freelancers until we had a constant development road ahead of us.

5. POST-LAUNCH

With the exception of post-launch production, I didn’t plan for this at all. It’s one of those things where as soon as you say it, it’s incredibly obvious that you should have planned for it, you total moron. I should have a bunch of social media lined up to keep our momentum; I should have sales and storefront plans to keep Cultist relevant once our launch-week offers expire; I should have paid ads, press articles and podcast interviews booked in to capitalise on launch visibility. But I don’t, because we both worked flat out to get Cultist ready for launch, and I’ve never been in charge of a game’s marketing before. Don’t Do What Donny Don’t Does, kids!

We had a basic but useful marketing plan for the duration of development: ‘this month is about getting press previews’; ‘we should shout about Perpetual Edition here’. That ran out in the first week of June, when apparently I stopped selling our game. We’ve had a good enough launch that this hasn’t damaged the studio’s prospects, but I bet we could have made more money and more noise if I’d had the foresight to see beyond release.

Alexis also tussled with post-launch, though his demon was feedback, not marketing. Bearing in mind Cultist has had a really great launch, it’s proven to be contentious, meaning people had a lot of Feels and these Feels were often sent to us. Feedback is invaluable and we’re grateful to everyone who sent something in (apart from the guy who said Alexis should be tried as a war criminal in the Hague – screw that guy). But when your game’s featured, you get a lot of feedback. And when all of it is essentially a critique of its design or creative direction, both of which are Alexis’s bag, it very quickly begins to chafe. And then it just doesn’t stop coming, even if what you’re hearing are reiterations of the same critique others have mentioned before.

My takeaway here honestly is to book in a goddamn holiday before our next launch, so Alexis doesn’t get that torrent of negativity when he’s already tired and worried and painfully aware of all the flaws in the game (have you ever seen a designer sit back at launch and go, yup, happy with that?). We also need to set up a more coherent feedback process, so it doesn’t manifest as a deluge of DMs to Alexis’s Twitter account, or an unwieldy number of support tickets, or a bitter taste in the backs of our mouths when we’ve just had a release more positive than we’d ever dared dream. We’ll also devote more time to closing in our next project, so there’s a higher level of polish and a lower chance of getting avoidable UI/UX tickets.

☽★☾

Thus endeth the lecture on Cultist‘s development! There’s loads of stuff to unpack, even in a project that only lasted ~11 months all in. If you think we’ve missed anything, or you have any questions, tweet me, Alexis or Weather Factory and ask!

Oh, and if you haven’t tried it yet, you can get your cunning hands on the game below! See? Sorting out post-launch marketing already.