Often I will collect data from hundreds of game devs and I can see a very clear cause and effect relationship between an action a game developer takes, the Steam algorithm (*) responding, and then traffic and wishlists increasing. 

But then later Valve will tell us through their great youtube lectures and documentation that the algorithm doesn’t work that way and whatever we are looking at was just statistical noise or us over-reading the tea-leaves. 

Sometimes though, even if the Steam algorithm isn’t responsible, there is a human behavioral algorithm that is impacting a game’s visibility. 

In today’s blog I want to look at some of the most common marketing techniques developers use and clarify whether it is the Steam Algorithm that is responsible, or the human behavior driving the effect they might be seeing.

So here are 8 marketing techniques and a look at who is responsible for them doing anything:

#1 You are penalized if you launch your Steam page without a trailer

Explanation: Can you launch your coming-soon page without a trailer? Does anything bad happen?

Steam algorithm perspective: FALSE

  • There is no flag in the Steam algorithm that says “if a game doesn’t have a trailer, reduce traffic by X%.”
  • When you submit your Steam page for approval, a trailer is not a requirement. 
  • The Steam store will use other assets to cover for the fact that you don’t have a trailer. For instance on hover, your screenshots will appear instead of the trailer.

Human algorithm perspective: TRUE

  • Yes you need to launch with a trailer.
  • When I held a qualitative study of how people browsed Steam, the trailer was not necessary; and many shoppers skipped watching it entirely. HOWEVER, my theory is that even if people don’t watch your trailer, if you don’t have one it just makes your game look like it is missing something. There is something odd about not having it. It’s like wearing a nice suit but not having a belt or suspenders. It’s not like you are naked, it just looks odd.
  • There is so much “my first game” stuff on Steam that you need to stand out above all those amateur developers by having a trailer. A trailer shows that you are serious and you took the extra effort! You want to show potential players that you know what you are doing and this isn’t your first game (even if it is).
  • Also, launching a steam page without a trailer probably means you didn’t make an “announce trailer” and you are missing out on a potential marketing beat. See the next myth.

#2 Launching a Steam page gives you a visibility boost

Steam algorithm perspective: FALSE

  • There is no Algorithm boost when you launch your steam page. 
  • There is no section in Steam called “New Steam Coming Soon Pages”
  • There is no extra traffic they give you for two-weeks.
  • This myth might come from this blog post where I compiled the number of wishlists earned during the first 2 weeks a Steam page is live. I wrote this post because so many people asked “I just launched my Steam page and got <x> wishlists, is that good?” I just picked 2 weeks arbitrarily because I figured it was enough time that it would level out any noise. There is nothing special about 2 weeks.  

People algorithm perspective: TRUE

  • Humans like new things. So yes, you should announce your game with a big TADA! And that includes contacting the press, influencers, and making a trailer. When your Steam page goes live you can do a bunch of stuff and get free coverage by other human beings and if you don’t partake in that, you are just missing out on free visibility. SO DO IT!
  • When you launch your coming soon page, you can put out a tweet that say “I am excited to announce my new game.” Because people are nice, they often will retweet it because they are excited for you. When I studied when and where people greatly increased their social media following, it was when they announced their new game. 
  • Before you announce your game and share your trailer, you can reach out to journalists and ask them if they would like exclusive access to it. This is HUGE. See this blog post where I explain how to get covered by IGN.
  • Sometimes festival organizers are more likely to accept you into a festival if you are announcing your game during the festival. 
  • Japanese Influencers and Press outlets on Twitter really like to tweet about new games. If you do a big announce, they will cover you and you will get more visibility. I wrote a blog post about getting covered by Japanese press on announce. 
  • If name a video “Announce Trailer” you are more likely to get coverage from it than if you name it “July Dev Log #58” because the human algorithm says we want to be the first person to uncover a hidden gem. 

#3 I didn’t get many wishlists when I launched my “coming soon” page, Steam hates me now don’t they? I have a black mark on my game and I should just delete my Steam page and make a new one.

Explanation: There was a GDC talk a while ago (link) where the very successful publisher said they work extra hard to promote their game at the coming soon page announcement because Steam will give you a “black mark” if you underperform at the start.

Steam algorithm perspective: FALSE

  • The Steam algorithm does not “bury” games if they do not achieve a certain number of wishlists within a certain number of days. 
  • The Steam algorithm DOES, however, boost games if they suddenly get a ton of wishlists. But the algorithm doesn’t care when those wishlists come in. 
  • It is demonstrably false if you look at games that broke out when they launched their demo. Here is the follower chart for Peglin (which I talk about in this blog post). They had a terrible first 2 weeks, in fact, they had a terrible first year! But after their demo was posted and Youtubers found it, Steam gave them incredible traffic. No black mark here:

Human algorithm perspective: Somewhat True

  • So why is it “kind of true” that if your game doesn’t go viral within the first 2 weeks, it is bad? IF you are making a game that can go viral (for example a horror game, or a game that has beauuuutiful visuals, or a game with a creative hook like “You must dance to earn bullets”) your announcement is the first test to see whether that idea has viral potential. If nobody cares about your game after seeing your game, and you don’t go viral even though you made a really good trailer that does a good job showing off what is unique about your game, it could indicate that the collective human algorithm said “eh, that game isn’t that interesting.” If you did a good marketing push in those 2 weeks and the human algorithm didn’t respond well, you might need to rethink your game. 
  • BUT DON’T WORRY, if you are making a deep, systematic, crafty-buildy-simulationy-strategy game, those games rarely go viral via visuals, or on social media. They usually go viral after the developers publish a demo and streamers play it (like Peglin)
  • Also, if you announcing your game without a trailer or big PR push, it’s like not jumping on the next train leaving the station. You just missed out on an opportunity to get somewhere faster. Why are you not boarding the train? Why are you not launching with a trailer? Humans are waiting to hear about your game and they want to see your trailer! JUST DO IT!

#4 I got caught in the spam filter when contacting streamers

Explanation: One of the best and most consistent ways to get wishlists is to send an email to content creators asking them to cover your game. However, many developers say something like “I emailed 1000 youtubers and I got caught in the spam filter.”

Email algorithm perspective: FALSE

  • If you are emailing from your personal, normal, email address (like a gmail, or one from your company domain) you most likely won’t trigger any spam filters. Gmail will actually cut you off before you do anyway. See this document that says if you send 500 emails to 500 recipients in a single day you will get a warning.
  • Similarly, if you are grabbing a bunch of streamer email addresses, throwing them into an email service provider (like mailchimp, or constant contact, or kit) without their consent, and then sending out a bulk email, that could trigger a spam filter.That is also technically illegal for CAN Spam act and GDPR because the streamer didn’t agree to be on your list. You can’t add their email address to those services. They need to opt into it.

Human algorithm perspective: FALSE

#5 You get an algo boost when you get 10 Reviews

Explanation: In an old blog post I documented how developers can see a very clear traffic jump the moment their game gets 10 reviews. Is this the Store algorithm? 

Steam algorithm perspective: FALSE

  • I have asked representatives at Valve is there is anything bout the 10 reviews and they confirmed that there is no “if 10 reviews, give visibility.” The basic Valve vibe is all about not having hard thresholds because it causes developers to “game” the system. Notice how Valve doesn’t allow you to ask for reviews in game? They worry that 10 review thresholds would just incentivize that. So I believe them that there is no limit there. Stop freaking out.
  • The actual algorithm is that before you get 10 reviews of your game your game is listed without a thumb icon. Once you get 10, your review score is computed and the game is assigned an icon that is a thumbs up, a line, or a thumbs down. You can see it in action here:
  • Also, before a game gets 10 reviews, on the store page, the game doesn’t get a bit of text that says the score, you just see a low number. 
  • Other than the thumb icon and the rating text, Valve has confirmed that there is no specific visibility algorithm change after the game gets 10 reviews. 
  • Valve has also confirmed that access to the Discovery Queue does not depend on getting 10 reviews. 
  • Valve also says that if there is an algorithm boost, it really isn’t that much. Even though that graph I show above is dramatic (it jumps from 100 to 878), in the grand scheme of things, it doesn’t matter that much. The Steam algorithm is built for million-unit-selling games and those games get 10 reviews in a nano-second. If you are having a hard time getting to 10 reviews, you have more problems than just the tiny boost it gives you. Basically your first 100-1000 copies sold is up to you and once you prove to Steam that you have something that sells, then they take over

Human algorithm perspective: TRUE

  • There is no denying that the traffic greatly increases once the 10 review mark hits. Even recently people have been sharing examples of it. It seems like human beings just like seeing that little blue thumb and Positive / Negative text. 

#6 You get a boost when you reach $50,000, $150,000, $500,000

Explanation: This myth is kind of new to me but several people have asked about it and I think I saw it mentioned on some reddit post. 

Steam algorithm perspective: FALSE (kind of)

  • There is no hard number that triggers some sort of visibility boost. Again, Valve doesn’t like these clear thresholds. I actually think this myth comes from developers who have experience releasing games on the Google Play store. I was talking to a mobile developer once and they said that Google gives you bonus boosts in visibility after reaching certain earnings milestones. 
  • HOWEVER, when you get about $15,000 ish Valve unlocks trading cards and achievements that can be featured on player profiles. Before you reach that dollar amount this message is shown on the game’s Steam page. However, Profile Features do not give you a visitability boost, it’s just something players like
  • HOWEVER, when you reach around $150,000ish gross revenue, SOMETIMES Valve will unlock the Daily Deal tool so you schedule one or if you use the support request tool they MIGHT approve you for one. The Daily Deal does give visibility and money. 
  • Similarly, when your game earns millions, you might be eligible for a Weekend Deal which gives you even more visibility. 
  • But by itself there is no “boost” that magically occurs. 

Human algorithm perspective: FALSE

  • Humans have no idea how much you make.

#7 I have “mixed” reviews my sales are impacted

Steam algorithm perspective: FALSE

  • Valve clearly states in their documentation that “Mixed” rating does not hurt sales. Here is the text “As long as your game’s reviews are Mixed or above (40%+), review score is not a factor in algorithmic visibility. If your game dips below 40% into Mostly Negative, your game will be less likely to be featured.” There is no reason for them to lie. If mixed hurt sales they would just say “Positive or greater has no impact on sales.”

Human algorithm perspective: TRUE

  • I have had developers who have a game that oscillates between “mixed” and “positive” say that they can see their sales dip when it hits “mixed.” This seems to be entirely due to shoppers worried about buying a product that isn’t up to their quality standards. 

#8 My launch failed, but I might be able to turn it around

Also known as saying “Ah but look at Among Us!”

Steam Algorithm Perspective: FALSE (kind of)

  • Officially, the Steam algorithms only care about money, not time. It just listens for huge spikes in dollars earned, and if you earn those dollars within a short period of time, Steam will start featuring you in widgets like the Discovery Queue, More Like this, and you might get invited to a Daily Deal or Weekend Deal.
  • I say “kind of” because Steam has a ton of widgets that only promote your game at launch. For instance Popular Upcoming, New & Trending, and the Discovery Queue all kick in usually right when you launch. If you don’t hit the right numbers at launch, you miss out on those widgets FOREVER. So that is a big piece of free visibility taken off the table.

Human Algorithm Perspective: MOSTLY TRUE 

  • Unfortunately, if your game didn’t sell well within the first month, things probably won’t turn around later. The reason is because humans just didn’t like your game. 
  • If you marketed your game adequately before your launch, it means you gave the game to Streamers, you put out a demo that people could play, you entered into festivals, and launched it with at least 7000 wishlists. By doing that the audience was exposed to whether the game was Fun or Not. At launch, enough people knew whether your game was fun or not and so there isn’t much more you can do to turn your prospects around. If you sent game keys to 300 streamers and they all tried it out but thought it wasn’t fun, there is nothing more you can do. You showed it to the influential people and they said “no thank you.” 
  • The best analogy here is before launch games are like Schrödinger’s cat. When you launch the game you reach into the box and hold that cat high above your head so the Steam audience can “observe” whether it is dead or alive thus sealing its fate. But instead of “Dead” or “Alive” what you are discovering about the cat is whether it is Fun or Not Fun.
  • Sometimes there are games that appear to “come out of nowhere” or they were a “late bloomer” or “were discovered hidden gems.” This is very very rare. But typically, games that this happened to were not marketed adequately at launch. A perfect example of this is Terminus: Zombie Survivors which I wrote about in this blog post. The game launched with 100 wishlists and only in the Korean Language. Even though it launched, the lack of marketing meant that the developer essentially kept the cat in the box and it wasn’t “observed”. The developer patched and patched and patched the game over years and added English and Streamers slowly discovered it. So it looked like the game turned around a bad launch, but really, the game wasn’t launched traditionally.
  • Again, it is very very very rare for a game to survive a weak launch because most of the time, if a game is marketed adequately, the reason it didn’t sell was because people didn’t like the game, not because not enough people heard about it. 
  • If you release and you find out your game isn’t fun it is really hard to change the game to make it fun later. The concrete is hardened. It is much easier to start over with a whole new game with a whole new game design rather than trying to transmogrify it into a different game. 

Appendix:

There is no ONE Steam Algorithm because Steam really is a collection of widgets and each widget has its own rules and behaviors. For the rest of this blog when I say “The Steam Algorithm” I really mean the meta algorithm that results from the constellation of widgets that make up the Steam store.