Steam’s discoverability strides & Steam Labs

[Hi, I’m Simon Carless, and you’re reading Game Discoverability Now!, a regular look at how people find – and buy – your video games. Or don’t. You may know me from helping to run GDC & the Independent Games Festival, and advising indie publisher No More Robots, or from my other newsletter Video Game Deep Cuts.]

With all the sourness about Steam and game discoverability that’s sometimes going on, I did want to note that I – personally – feel like the company is making an effort recently.

This post by Valve’s Alden Kroll – published in early October – about Steamworks improvements lists a bunch of the changes, Steam Labs being one of the most notable.

The state of Steam Labs

For those not paying attention, Steam Labs is a whole bunch of experimental methods that allow Steam users to find new games. Some are coded internally at Valve, and others created by notable indie scene folks including Ichiro Lambe from Dejobaan Games & Lars Doucet from Level Up Labs.

You can read all the Steam Labs blog updates for more detail on what Valve & various contractors have been up to. (Although there hasn’t been an update since the end of September, and I think a bunch of stuff happened – time for another blog, Valve?)

And I’m intrigued to see that multiple Steam Labs recommendations DID make it to the front page of Steam. For starters, there’s a ‘Community Recommendations’ area (“Discover games the Steam community is currently enjoying based on recent reviews.”)

This isn’t personalized based on the games you purchased, but has some advanced controls that help to tune it.

In addition to that section, if you log on you get ‘Learning Machine’ and ‘Deep Dive’ choices based on what you’ve already bought, too. The front page of Steam has capsule version of the full tools, that you can click through and use.

See below for how my Steam purchases affected parts of the Steam front page – so this is actually fairly advanced personalization (although I’ll note that the displayed games all seem to be fairly popular ones):

Pretty cool, huh? (And it seems fairly prominent to me, too. I was fairly vocal with Steam folks on Twitter around the launch time that the experiments would only be meaningful if they extended past the ‘small-scale prototypes’ stage. And it looks like they have. And not just because I asked, haha.)

[IMPORTANT NOTE: if you want to go and look at all of this now, the Steam front page is reconfigured for the Autumn Sale, so you won’t see these sections. You can still see the Steam Labs experiments on their dedicated pages, of course.]

Discoverability sidebar: the Steam Discovery Queue

On a side note, please check out Erik Johnson’s recent-ish Gamasutra blog on the Steam Discovery Queue if you want a deep-dive into yet ANOTHER discoverability-related part of Steam that you might not be completely aware of.

Basically, the Discovery Queue is a set of recommendations – originally launched in 2014 – and described at the time as follows: “Steam recommends twelve games to you, be they popular titles or titles it thinks you might be interested in, and you simply go through and either add them to your Wishlist, say you aren’t interested, or skip the suggestion and move on.”

In the end, Erik’s 2019-era numbercrunching suggests that the Discovery Queue, which is much more long-running than Steam Labs, does rely on ‘popularity and recency’ quite a lot. But that’s probably why Steam Labs exists in part, right?

Steam Discoverability changes: do they fix discoverability on Steam?

There’s one major issue re: ‘discovery’ for Valve. If it looks like you’re not making an effort as a platform, people will presume that it’s your lack of effort in helping discoverability that is making their game sell ‘poorly’.

For example, Apple has always had prominent editorial features for interesting ‘premium’ games. This means that – in my view – they have historically got less complaints over being an ‘uncaring, money-hungry’ platform. Why? Because people believe that they are trying to make a difference for high quality titles, not just top-grossing ones.

But premium games don’t necessarily sell that well on iOS (or Android) – even features aren’t a fortune-changing boost on their own. And most platforms (save maybe Itch) are in the absolute abstract equally uncaring, because those harsh, harsh capitalist market forces win out.

So this is as much a dev community relations issue as anything. Although Valve continues to believe that it’s algorithms – rather than dedicated editorially-focused picks – that should determine what people see on Steam. Which naturally makes Valve seem less lovable as a company, even if some of those algorithms are handcrafted. We’ll see if that changes in the future!

How people actually buy games on Steam.

And why isn’t your game selling well on Steam? It could just be supply/demand changes in the market – so many games nowadays! Or possibly, it could just be that your game may not be attractive enough to the ‘average’ Steam user, or be very attractive to a smaller niche subset of the market.

And honestly, as we learn from Chris Zukowski’s latest blog on how regular Steam users shop during sales (check it out, it’s a must-read/view!), it may not be ‘Steam Labs’-style discoverability mechanisms that make people choose and buy games.

(Nonetheless, I would love to see more stats from Valve on ‘this discoverability feature made this game sell X more copies’ of course, some of this is spitballing based on my own impressions.)

The more that I see surveys like Chris’ latest (and his excellent previous one), I realized that much of discovering games is not based on using the platform to discover new ‘unexpected’ ones.

It’s based on checking out already popular games, asking friends, looking elsewhere online (not on Steam itself!), and – most importantly – making decisions based on games & game genres you’ve liked before.

The Steam user decided this themselves, all on their own, in their brain. There’s a limited amount the platform can do about it – besides showcasing ‘top games’ and ‘similar games’ efficiently, which is WAY less sophisticated than Steam Labs, discoverability wise.

Conclusion: Do/don’t blame the platform!

In the end, I know it’s easy to ‘blame the platform’ for sucking at discoverability when your game doesn’t sell as you’d like. But a lot of the time it isn’t the platform, it’s you not developing enough pre-release interest in your game (as measured in wishlists, followers, etc).

Sure, platforms can help past a certain point – and there’s a definite ‘rich get richer’ bias to Steam sales, whether platform-created or not. But you need some initial momentum, right?

However, I’m not saying platforms should do nothing. Discovery systems not being the primary sales driver shouldn’t stop Steam from adding them. And I think Valve could make the front page of Steam even more personalized. That likely would help the smaller guys.

Even now, you can argue that Steam is currently tuned for maximum customer satisfaction. A lot of these top-selling, often-updated games genuinely do delight customers & get highly rated and are played for a long time. So why mess with that? That’s probably been the historical position for Valve. Even if it means the top games just keep selling. Dialing that ‘prominent games get more prominence’ vibe back a bit would help.

Still, Steam is still better than online game stores that have no real-time charts prominently displayed or that over-display ‘favored partners’, let’s not forget. And this level of front page discoverability notoriety is further than I thought Steam were going to go when they first introduced Steam Labs. So… good job, Valve? (Again, not Valve Defense Force-ing it here, just calling it how I see it.)

Take care,
Simon.

PS – I don’t think Valve should take a 30% cut of your revenue if you publish on Steam. But that’s another thing entirely, I guess!)

PPS – Three REALLY good game discoverability/business-y articles you should read after finishing the one. Firstly, PC Gamer on whether your game is underpriced. (Spoiler: it probably is.) Secondly, Victoria Tran on ‘the $0 Marketing Game Guide’ – great stuff as always from Victoria & Kitfox. And thirdly, Ryan Sumo actually showing his full, annotated developer/publisher contract for a recently signed game, wow. Neat stuff.