Steam in 2017
How Steam Direct, PUBG and the rise of China affected the biggest PC gaming marketplace in 2017.
The following article is based on the presentation I gave at GDC 2018. The data was gathered during February 2018 and by now is probably slightly outdated.
Obligatory Fine Print
Steam Spy only tracks owners and players, not sales. Steam Spy doesn’t track DLCs or MTXs.
It’s impossible to distinguish between games sold on Steam, sold elsewhere and given away for free. For the purposes of this article we’re measuring the market Steam controls, not the sales coming directly through Steam.
Steam Spy uses 98% confidence range and is very inaccurate for small games, especially the ones below 30K owners.
Geography on Steam Spy relies on self-reported data. Only engaged users fill out the country field in their profiles, therefore the geo-data is inevitably skewed.
Steam Spy tracks users playing games, not idling in the Steam client or botting trading cards.
Steam Spy is very inaccurate as a method of tracking refunds. I do have a small sample of data on refunds from a dozen of developers, but I don’t feel confident about it to include it here.
The site uses 3-day samplings. For this article I used a 10-day sampling, so it’s a bit more accurate than the site.
Steam Spy only uses public profiles to gather and estimate the data. 99.9% profiles on Steam are public.
Steam Spy is always lagging by at least 4 days, and is completely inaccurate for new titles.
The market
2017 was the best year for Valve so far. Every single metric you can imagine grew: the overall number of games sold, the audience, the total revenue and so on.
Add f2p titles and in-app purchases to this number and you’re looking at the total number well over $5B. Of course, not all of this revenue comes through Valve, a sizable portion of sales happens through 3rd-party sites and even in retail. The size of the market controlled by a single private company is still amazing.
By the end of 2017, Steam had 291M that have played at least one game at least once. 22% of them joined in 2017, so if you were worried that Steam only appeals to the aging core demographics — it’s not. We don’t know the age distribution for sure, Steam has been adding more and more new users every year and is far from slowing down.
57M people played anything on Steam in the last two weeks. Steam is smaller than some single games (like Minecraft or League of Legends) but it’s certainly huge nevertheless.
Steam’s retention is nothing but amazing. There are no other games that can show the same retention over the years. That’s, of course, because Steam is a hardware-agnostic store and should be compared to Amazon or eBay, not to conventional videogames or even gaming platforms.
The slight decline in the user retention on Steam in the past years could be attributed to the huge amount of new players joining.
New players are buying less games than the old ones, especially with the majority of them joining from developing countries like China or Philippines. Though 1 as a median number of owned games might seem low, it should be noted that if the players retain and continue using Steam, they’re likely to buy more.
While 15 as a median number of games for users that have joined in 2003 might look high, it translates to only one game per year. Not exactly extraordinary or even surprising.
Games
Newsflash: Steam has a lot of games. But you know this, right?
Steam sales are skewed to the top with just 100 games (0.5% of all) accounting for 50% of total revenue. It’s probably even more skewed if we add in-app purchases and free-to-play titles.
The top 20 list doesn’t look so surprising. PUBG, being the behemoth it is, took home cool $600M in sales (again, not counting in-app purchases). It’s also nice to see a bunch of indie titles on the list — Divinity: Original Sin 2, ARK, Rocket League and fans-favorite Cuphead.
If you just look at the top 20 games, the situation hasn’t changed much. A game still has to make $22M to break into the top 20. So, why all the gloom and doom about the indiepocalypse?
Because while top games are safe, it’s not so pretty for the rest of them.
When I wrote my first article about Steam three years ago, a median game had 32K owners.
The pricing situation is also becoming more and more unhealthy and reminds the pre-iOS 11 App Store. The race to zero is on.
Although, a lot of this could be attributed to the sheer amount of new games released. Counting medians and averages is starting to lose its meaning, so I checked top 500, top 1000 and top 2000 games on Steam instead. In all of the brackets there is a slight year-over-year decline, but it’s insignificant, really — from 3% to 7%. It could both be attributed to players buying PUBG or any of the new small titles.
While Steam has its best year so far, the number of new users joining the platform and buying games can’t keep up with the number of new titles being released. On the graph above you can see the total time Steam users spend playing indie titles. It grows with every Steam sale, but not nearly enough.
That’s because Steam Direct brought in a lot of new titles to the platform. It wasn’t nearly as disruptive as Steam Greenlight, though.
But let’s go back to the prices. I’ve already mentioned that the average prices of games on Steam are going down. This situation is affected severely by the regular Steam sales that taught users to wishlist and wait instead of buying immediately.
If you look at the full price of the games on Steam, you can see a normal distribution around $9.99 — a so-called sweet spot that a lot of indies love.
But it’s very different if you look into the total revenue, even with all the sales and discounts taken into account.
We still see the same normal distribution, but this time around $19.99 price point with a huge spike for full-priced games.
Surprise: cheap games do not make a lot of money. People are still mostly buying relatively expensive and quality titles.
Geography
In February this year the Steam Hardware survey reported that 64% of Steam users had Simplified Chinese as the default language on their machines. That lead a lot of people to believe that China was rapidly taking over Steam.
All that while Steam is not officially available in China. Valve is only allowed to operate in Hong Kong. Of course, players from mainland China ignore this and play Steam games anyway.
If we look into the people that actually play anything, the numbers are quite different. Yes, there was a spike in the number of people playing from China, but not in the number of games they’re buying.
There is a huge difference between 19.5% and 64%, right? It’s even bigger if you consider the number of games they own: 64% vs 5.4%? This is insane.
The actual situation is weird and somewhat hilarious. With the rise of PUBG in China, a lot of local cybercafes started installing Steam on their machines, so their patrons could play PUBG.
Most of those cybercafes only installed the Steam client, they never bought the game. Therefore the discrepancy between Steam Hardware Survey (counts machines) and Steam Spy (counts players).
But some of cybercafes did buy PUBG. An average cafe charges $3 per hour of play and having the most popular new game available obviously helped them attract clients. Of course, being able to play on someone else’s account with no repercussions for cheating, lead to a rise of cheaters in PUBG from China, most of them playing from cybercafes. Even if a machine gets banned, it’s just $15 in the game’s cost for a cybercafe owner. Just a business expense.
Let’s look at the US for comparison.
The share of the users from USA steadily declines, but that’s not because Americans are playing less. It’s just Steam has reached the point of saturation in most Western markets.
US users are buying way more than people from China. Again, expected, given the difference in GDP per capita.
So, while you shouldn’t discount China as a market for your game, please remember that it still is smaller than the US market on Steam.
But they sure play PUBG a lot.
Talking about the markets. China and Russia are two huge markets where people buy indie games. But they buy less on average and enjoy special regional pricing, making these markets less lucrative at the moment.
USA and the big countries of Western Europe remain the biggest markets for indie titles on Steam for now.
Summary
2017 was the best year for Valve so far: more games, more sales, more gamers.
Although there weren’t nearly enough new gamers to buy all the new games.
Steam sales are heavily skewed to the top, but it was always like this. We see more games joining the tail than ever and affecting averages.
Which makes Steam similar to most entertainment markets (think movies, music and books).
Games at higher price-points make more money, especially full-priced AAA titles.
Duh.
There are many users from China on Steam, but they don’t seem to be buying much outside of PUBG.
China is the biggest gaming market in the world, but it’s not yet reflected on Steam.
Steam Direct caused a significant increase in the number of new titles releases but it’s nothing compared to the Greenlight launch.
Steam Greenlight started it all and while Direct removed some barriers, even with Greenlight in place we’d see more games this year than before.
Do you want to know more?
If you’d like to learn more about Steam games and current state of its affairs, visit Steam Spy and follow me on Twitter.
You can also support me on Patreon and be the first to get access to all the cool insights and features of Steam Spy.
