Half-Life: Alyx led to a huge leap in VR usage on Steam when it launched back in March. Four months later, there’s nearly as many VR headsets on Steam as there were right after the game’s launch.

Each month Valve collects info from Steam users to determine some baseline statistics about what kind of hardware and software is used by the platform’s population, and to see how things are changing over time, including the use of VR headsets.

While we had expected that Half-Life: Alyx would lead to a surge in VR usage on Steam, we also figured that only a portion of that surge would remain in the long run. Surprisingly, Steam appears to have retained nearly all of the growth in VR usage brought on by Alyx four months after the launch of the game, according to Valve’s monthly Steam Survey.

In fact, if you look purely at the percentage of Steam users with VR headsets, this month’s figure is slightly higher than the peak brought in by Alyx.

Monthly-connected VR Headsets on Steam

The data shared in the survey represents the number of headsets connected to Steam over a given month, so we call the resulting figure ‘monthly-connected headsets’ for clarity; it’s the closest official figure there is to ‘monthly active VR users’ on Steam, with the caveat that it only tells us how many VR headsets were connected not how many were used.

The latest figures for the month of July show that 1.93% of Steam users had a VR headset connected to their PC over the course of the month, which is actually the highest the percentage of VR users on Steam has ever been.

To put this into perspective, Steam users with connected VR headsets are about twice as common as those using Linux, and a little more than half as common as those using MacOS.

While Valve’s data is a useful way see which headsets are most popular on Steam, the trend of monthly-connected headsets has always been obfuscated because the data is given exclusively as percentages relative to Steam’s population—which itself is an unstated and constantly fluctuating figure.

To demystify the data Road to VR maintains a model, based on the historical survey data along with official data points directly from Valve and Steam, which aims to correct for Steam’s changing population to estimate the actual count—not the percent—of headsets being used on Steam.

From the model we estimate there was 2.6 million monthly-connected VR headsets on Steam in July. Although the percentage of VR users in July (1.93%) was ever so slightly higher than the April Alyx peak (1.92%), from a raw number standpoint we estimate that April still held the edge with 2.7 million headsets that month. Still, four months later, nearly all of the Alyx spike has remained.

As we can see on the graph, June seemed to have a sudden drop but recovered quickly. It isn’t clear if this is just a blip in the data or if June really did see around 400,000 fewer headsets than July. Valve did make some tweaks to the Survey (which now breaks down more headsets and VR streaming drivers), though they didn’t specifically say that a dip would be expected.

Share of VR Headsets on Steam

As for the share of individual headsets on Steam, the biggest winner in July was Valve Index (+1.55%) which now represents 14.45% of all headsets on Steam. That’s surprising considering the significant global backorder that the headset has faced during the Coronavirus pandemic. This may be a sign that Index has begun shipping steadily again, but that Valve is still churning through its backlog of orders.

Oculus Quest also had a big month (+1.26%) for an overall 10.13% share of headsets. If it keeps growing at this rate it may well eclipse Valve’s Index.

On the losing side, first-gen headsets are giving way to newer headsets. The original Rift CV1 (−1.03%) at 13.63% now holds less than Index for the first time. The original Vive (−1.06%) is declining too but still holds an impressive 23.03% of the share of headsets on Steam.

Windows VR headsets (−0.99%) have continued their steady decline, now at 7.36% from their June 2019 peak of 11.13%. Later this year HP will launch Reverb G2, the first new Windows VR headset, since Samsung’s Odyssey+ in 2018. We’ll be interested to see if the headset alone can breath new life into the bloc.

Exclusive Hands-on: HP's Reverb G2 is the King of Clarity

Looking at the headset vendors overall, the ‘big two’ still hold strong; Oculus headsets (+0.69%) now hold 45.55% and HTC headsets (−0.89%) 28.82% of all headsets on Steam. Valve (+1.55%) is still a somewhat distant third at 14.45%, but closing the gap as HTC’s overall share of headsets continues to decline from its 39.28% share this time last year. Windows VR headsets (−0.99%) are giving way too, now at just 7.36%.

As ever, it’s worth noting that the Steam Survey only gives us a glimpse of the overall VR market, as it only counts headsets connected to Steam. That means it doesn’t count some portion of Rift & Quest users which may not use Steam at all, nor other major headsets like PlayStation VR on PS4 and Quest standalone.

Newsletter graphic

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. More information.

Ben is the world's most senior professional analyst solely dedicated to the XR industry, having founded Road to VR in 2011—a year before the Oculus Kickstarter sparked a resurgence that led to the modern XR landscape. He has authored more than 3,000 articles chronicling the evolution of the XR industry over more than a decade. With that unique perspective, Ben has been consistently recognized as one of the most influential voices in XR, giving keynotes and joining panel and podcast discussions at key industry events. He is a self-described "journalist and analyst, not evangelist."
  • TechPassion

    Senseless assumption that VR usage will drop after a few months. Those VR headsets are bought and in someones hands. If somebody pays 1000 USD for Index or 500 for other headset, they will use it after they play the game.

    • Flo

      If good VR games keep coming out sure.

    • patfish

      I and if not every game will be downgraded to smartphone graphics for the Quest like Onward! :-/

    • VR5

      It’s not senseless at all, considering it did drop in the two following months. Some people probably just took their otherwise unused VR systems out of storage (or lend out older retired systems to friends) to play Alyx, then put them away again.

      But the number finally being on the rise again seems to indicate that VR systems keep selling and enough new users keep using them. No surprise, since that is also evident from the Index not leaving Steam’s revenue charts, and other headsets selling out.

      • TechPassion

        How many % lent them? 0.02%? Would you just rent your index and base stations to somebody? The increase was in hundreds of thousands.
        Yes… unused Index or WMR or Rift S from storage, right….

        • VR5

          The fact is, it did drop. You can clearly see that in the graphs. We can just guess the causes.

          Reconnecting an othertimes unused headset is one of the first things to come to mind and also has been mentioned as a possible explanation both here on Road To VR, and Kotaku (which have an anti VR-bias so of course they want that to be the main explanation for the spike in general).

          I upgraded from a CV1 Rift to Quest, so I could have lent my Rift to someone. Many people should have upgraded from a first gen headset like Rift or Vive to Rift S, Index or Windows MR. If you don’t sell your old headset, it is available to be borrowed by friends who just want to play Alyx.

          Either way, whatever the cause was, it did happen. So it is not senseless to acknowledge that fact.

    • Ad

      It’s possible, it was common in the past and I know people who put their headsets aside because there wasn’t software to use consistently. But that is changing so we’re in a much better place.

  • Kyokushin

    Next bump will be on HP Reverb G2 launch.

    • cirby

      Yep. The two-stage combo of the G2 with a VR-capable Microsoft Flight Simulator will make a LOT of people jump into VR.

      I was surprised at how many people I know who have fairly complex multi-screen dedicated flight sim setups at home, and several of them are already into VR.

      (I already pre-ordered my G2, so now I have to figure out if I want to drop another hundred bucks on Flight Simulator…)

    • Andrew Jakobs

      Don’t think it will be an extra bump, as most people who will go for the HP Reverb G2 are replacing their current headset.

      • Ad

        HP doesn’t have the marketing, and tech news doesn’t seem to be giving them the kind of free press they give facebook. It’s a great entry headset.

        • Andrew Jakobs

          I agree with you on that. It really seems to be a great headset, but it could have been just a tad better (like a tightenknob at the back for instance). But the headset is still not actually released and we still haven’t had a real final test of the headset (including the controllers), but by the end of september when most people hopefully have their pre-order, we’ll see how it actually is perceived and we’ll see what kind of problems it will have (just like last year with the release of the original reverb which wasn’t a good release).

          • Ad

            Oh for sure. I’m hoping it all goes great and valve helps them reduce any compatibility issues, make tracking even better, etc. Sebastian didn’t show the camera feed, but hopefully it’s at least passable passthrough, as this seems like the ideal headset to do work in and a desktop view overlaid over passthrough is the best way to do it solo.

    • I don’t think so. Reverb G2 is a prosumer headset. It may eat other Oculus and HTC shares, but not increase the number of VR users

      • silvaring

        I think Microsoft gave Dell or HP early access to develop the original WMR headset, so maybe its the same, and other vendors will release late this year or early next year. Thats what I’m thinking anyway, and it will be priced more competitively for the other newer headsets from Oculus and whoever else decides to bring something out. Exciting times.

      • Ona Mynuos

        I don’t think functioning VR devices will be left in the cupboard. They’ll be sold on the used market at the value they have left.

  • wheeler

    Are you certain that the Steam Hardware numbers are for connected headsets and not used? Uploadvr is reporting them as actual usage. I ask because one means something very different from the other, especially if VR has retention/usage struggles.

    This is good to hear. When I saw last month’s numbers I thought my suspicions were correct about us heading downward post Alyx. Perhaps this is because supply levels have picked up a bit.

    As uploadvr reported, the “gen 1.5” split between Index (14.5%) vs Rift S (~21.5%) is really impressive for a headset that is over twice as expensive, and now the Index has surpassed the original Rift. It pretty much solidifies that even within the niche of VR there is a market for high end VR. The Vive, while decreasing, is still hanging on strong.

    • Schadows

      Previously, the company explained, VR headsets were counted by the Survey only if they were plugged in via USB at the time the data was collected. The new method will allow SteamVR to report to the Survey any headsets that have been used by the system in the last month, regardless of whether they are plugged in at the time of data collection.

      • wheeler

        Yeah, that’s what I mean. Does “used by the system” mean the user is actually using the headset to play games? Or does it mean that this background monitoring system “used” the headset in its data collection? (regardless of whether or not the user actually played with it). The difference could be massive–I have several friends that leave their old Rifts connected but they haven’t used them in over 6 months

        • Schadows

          For me, the only moment Steam detect which VR heasets is plugged in is when SteamVR is launched (which doesn’t necessarily means a game has been launched).
          I have no proof, but I would be stupid to log its “usage” at any other point than this.

    • benz145

      I’m nearly certain that the numbers do not account for “use,” they only account for headsets that were “connected/recognized” over the course of the month. That’s why we call it monthly-connected headsets instead of monthly-active users.

      RE: @schadows:disqus below, I think “used by the system” in this case simply means recognized by the system.

      • Baldrickk

        Can I suggest the usage of more colours in the graph?
        Black, three greys, and a dark navy blue make for a tough to read graph.

  • Ad

    Worth considering that many people use steam for extremely basic gaming. So when I was a console gamer I had maybe three or four games on steam, and I actually played them on a laptop. Console gamers and laptops, which are actually two thirds of steam, make the 2% number misleading. Not to mention that steam is used in tons of countries that aren’t relevant when gauging VR’s success even if just because tariffs alone make obtaining headsets as impossible as getting consoles.

    This is great though and it shows that better software across the board has made VR more useful and consistently enjoyable, and hopefully that continues because it’s how VR will actually grow. And PCVR is quite healthy.

    • benz145

      This is why we estimate the actual number of headsets; it’s more useful than a percent of an unknown population.

  • Ajedi32

    I’m skeptical that the spike in the April survey was primarily due to Half-Life Alyx. My suspicion is that the bigger factor there was the Steam Survey’s change to their counting methodology. Supposedly that change happened in March, but the spike in the March survey was way smaller than expected; it wasn’t even enough to offset the previous inexplicable drop in the February survey’s stats. I strongly suspect that the much larger spike in April was when the changes to counting methodology _really_ took full effect.

    With that in mind, I disagree with the headline of this article. Obviously “growth” which was caused primarily by a change in counting methodology isn’t going to suddenly disappear just because a popular game is no longer in the limelight.

    The drop in the June survey is also a bit weird. I think it might be best to wait a few months for the stats to stop fluctuating wildly (here’s hoping) before drawing too many conclusions from this data.

    • Rogue Transfer

      Your premise requires that a significant percentage of users use more than one headset each month.

      It seems likely most users wouldn’t keep switching between multiple headsets, as most upgrade and stop using their older headset after a short time or sell it on.

      Any rise from the counting methodology change would likely be not much more than the amount of transitioning users to a newer headset, or the few that have two setups they use each month. Both likely much smaller percentages than we saw in the huge rise at HL:Alyx launch.

      The drop in the May & June are likely due to the warmer months where less people will feel like using VR.

      • Ajedi32

        I don’t understand your point. What does people switching headsets have to do with the change in counting methodology?

        The major effect of the counting methodology change is that a large number of headsets which previously weren’t detected by Steam VR at all now are. Headsets which get unplugged when not in use are now much more likely to be counted than they previously were, as are wireless headsets and virtual headset drivers like Virtual Desktop and Riftcat. I’d fully expect those changes to have _huge_ impacts on the overall numbers.

    • benz145

      Important to note: when Valve publishes the latest data, the data we’re looking at is for the prior month. So the data published on June 1st covers what happened in May.

      The change in survey methodology was complete as of April 1st (https://www.roadtovr.com/valve-tweaks-steam-survey-accurately-count-vr-headsets-pimax/), and it was used for the March data. You can see the holiday spike and then the post-holiday lull, and then the spike in March was the change in methodology.


      Valve says that most of the Survey data collection happens in the first few days of the month. Alyx launched on March 23rd; Valve specifically advised us that March’s data was unlikely to account for much (if any) of the Alyx activity—and remember, March’s data used the new collection methodology.

      We didn’t get a good look at Alyx activity until April’s data was published in May, at which point we saw the big spike, and this was well after they changed the collection methodology for March’s data.


      Granted, let’s go with your premise that the April spike had little to do with Half-Life: Alyx. In that case where is the Alyx bump in the data? If we don’t see it, wouldn’t we have to conclude that Alyx didn’t make a measurable change in the data?


      • Ajedi32

        My current hypothesis is that the bump from Alyx is partially present in the January survey (people getting new headsets for Christmas in anticipation of later playing Alyx) and the April 2020 survey (masked by the much larger bump due to the change in counting methodology).

        Alyx is a great game, but I very much doubt that many people bought new headsets and then just left them in the box until March. Certainly not enough to increase the total number of VR users by ~50% overnight anyway. The fact that the numbers haven’t reverted to their former levels yet is, I think, further evidence that the counting methodology change is the primary factor at play here.

        That said, it’s hard to be sure of anything when the stats are fluctuating like this. The last few months have seen swings larger than anything we’ve ever seen prior to the change in counting methodology, so we don’t have a whole lot to compare them to right now. I’m going to wait and see if things stabilize, or if this is the new normal before I put too much stock in this data.

        • benz145

          Don’t forget that the bump isn’t all new users. A large portion will consist of people who bought a headset over the last four years but haven’t continued to use it, but Alyx gave them a reason to pull it back out and then check out all the content that’s come out since they stopped using it.

      • Ad

        I think all these people using Indexes, Vives, and Rift S’s really shows how the Quest has revolutionized VR, is the largest market, and anything in VR that can’t be done on a 835 shouldn’t be done at all. If Valve doesn’t give up and port Alyx to the Quest (with the PC version cut down for crossplay) then they’re crazy. s/

        • Baldrickk


          The Game Boy colour was a fantasicly popular device too. We should never have made games that couldn’t run on it. Anything that can’t be done on a GBC shouldn’t be done at all!

          • Ad

            And no one said you should get a game boy instead of a console, no one said it was the best gaming device to buy, and no one said the market should be “game boy first.”

        • VR5

          /s stands for strawman, right? The favorite tool of the fanboy. Alyx doesn’t even have multiplayer, so how could there be cross play?

          It’s up to devs/pubs where to put their games. Valve obviously will put them on Steam.

          Not every game needs to be on every platform.

          • Ad

            It stands for sarcasm, and yes obviously Alyx wouldn’t be on Quest. Epic games has steamVR games though which is nice. The point was that PC VR is so healthy and is growing fast but thoughtleaders have moved on to hyper pastures.

          • VR5

            Growth has accelerated since Alyx but I’m not sure I would call it fast yet. Obviously now that headsets are selling out production is limiting growth so let’s see how much they can do with ramped up production.

            According to the Nikkei report, Oculus has ordered 2 million of a new headset, the rumored Quest revision. Unless other vendors also increase their production, PC will again fall behind in momentum. I guess if HP is confident in the Reverb, that could help PC keep pace.

            Of course, it is entirely possible that one or more of these will fail.

            Another thing to consider though is how the increased install base will help software. I haven’t seen any positive reports from devs about increased sales on PC since March (but I might just have missed them).

          • Ad

            There are more games on PC. Again, the quest has so little competition that sales numbers are pretty worthless in many cases (it’s actually not about how many quests there are but how much software people are buying), whereas on PC there is constantly more competition and most of the top ranking games aren’t really the kind that talk about it in these circles. Like Pavlov, Blade and Sorcery, etc

          • VR5

            You’re definitely correct in pointing out that there are more (much more) games on PC. But that is exactly the problem. With competition being fierce it is hard for any given dev to make money. Even with an install base approaching 3 million (since 2016).

            On the other hand, even with a great first year it is unlikely that Quest already outsold the combined PC headsets. A million might be a reasonable estimate. But we get these success stories of first week sales being many multiples of first week sales on PC (which of course also were on smaller install bases than they are now).

            Basically, with less competition and customers willing to spend, Quest has supposedly been much kinder to devs. We get those reports. We don’t get them for PC.

            If devs keep porting to and sometimes putting priority for future updates on Quest, that should be indicative that there is more money to be made there. We’ll have to see how this develops longterm but as of now, Quest seems to be the best VR platform to sell your games yet. If you can get into the curated store.

          • Ad

            This… makes no sense. Besides the fact that you cannot laud the quest for having no games on it, more devs piling onto Quest will eliminate these benefits. Onward likely rushed to launch since Pavlov is coming to Quest in a cut down port for free. I’m just saying it’s incredibly dumb to say Quest is the best place to sell your games… because there aren’t games on it. It will be interesting to see what happens with Horizons, which will definitely compete with small games.

          • VR5

            There are games on Quest. It launched with over 50 games and now has well over a hundred. All of them quality titles thanks to the curation. That means it is easy to find something to buy and customers will buy more titles because they aren’t getting burned by low quality efforts.

            Having fewer games doesn’t equal none. You need to grow out of this binary logic of all or nothing, good or bad with nothing in between.

            And facts are facts. Many devs reported they sold more on Quest. I haven’t read any news where a dev claimed they sold more on PC than on other platforms, PSVR or Quest.

          • Ad

            It’s literally not a fact when you mislead that much. You have just side stepped and side stepped. First there were just so many quests, then that wasn’t true so it’s that people buy so much software, but that’s not true so now it’s that there is a small amount of software and therefore it’s great for all devs that a tiny number of them can port their old games onto the quest as new since people aren’t interested in them on PC anymore.

            “because they aren’t getting burned by low quality efforts.”

            This is because Oculus on PC is a mess of terrible quality and zero curation compared to steam. And you’re not including any stories of Devs who were not allowed on the store after they did all the work of making their game.

          • VR5

            The fact is that many devs report high sales on Quest, compared to other platforms. The other fact is that some devs give up on VR because they’re not generating enough sales. That happened mostly on PC.

            Nothing we argue here, analyzing, interpreting, or whatever changes anything about those.

            Steam has almost zero curation, which is not necessarily a bad thing. I think it’s odd that you would try to claim otherwise. Oculus stores except Quest are mostly the same as Steam, hardly any curation. That wasn’t always the case, both Rift and Gear VR stores started out heavily curated.

            For Quest the curation works, because they have a lot of quality titles.

            And you really should stop accusing me of twisting facts when anyone can easily see that describes your MO much better than mine. And when you keep saying that something that goes over your head makes no sense, it only reflects badly on yourself.

          • Ad

            Steam has built in curation because community response is a tight sorting mechanism compared to Oculus.

            What you’re saying about high sales on Quest is completely valueless because it is entirely backwards facing. I’m not sure why you can’t see that. You’ve admitted it’s an issue of reduced saturation so it’ll evaporate pretty quickly. Interesting how none of the people blowing smoke up the Quest never show the full hand on that.

          • VR5

            Community response is not the same as curation. PC audience tends to be harsher too and it waits for sales a lot. All that doesn’t seem to help with getting report worthy success stories.

            As long as Quest keeps selling at the pace as it apparently does, saturation should be maintainable at favorable levels. Basically, if the audience grows with releases, they should be able to maintain the high sales.

            And no one in VR is showing full hands really. How much did Alyx sell? We don’t know. We get bits and pieces of information, mostly when there is some milestone. Shows how fragile this market still is.

          • Ad

            Your faith in the quest is still, as always, a mix of hype and assumptions.

          • VR5

            It’s not faith, it’s observation. And they aren’t assumptions, they are dev comments, number of community reviews and average ratings. All facts that compare better to lack of positive dev comments on PC, lower number and average of ratings on PC. Quest audience is empirically buying more. It’s ridiculous that you pretend that data doesn’t exist.

            If you were just pointing out that we don’t get a lot of concrete comparative data, just the developers’ word for it, you’d have a point. But saying there is no data, just assumptions, is plain wrong.

            And at last year’s Oculus Connect it was reported that Oculus doubled its revenue of 3 years of store history within a few months after Quest release. Of course since Steam has better prices a lot of Rift game purchases aren’t actually made on the Oculus store. But given the exclusives Rift has, it wasn’t a small feat for the Quest to completely defeat PC (and mobile) Oculus revenue at its launch.

          • Ad

            Even journalists who are obsessed with the quest aren’t making the claims you are. This is nonsense.

          • VR5

            Yeah, sorry I misremembered. Two weeks after launch Quest had generated 5 million dollar revenue. At OC6, it made up 20% of all Oculus store revenue (20 million out of 100 million). Still, that’s three months compared to three years. On a much smaller install base than mobile and Rift had by that time.

            First paragraph I wrote has no fault though.

  • Andrew Jakobs

    Well, that’s why I think it’s strange for some older VR-games to not put out a small update to get their game to work with newer headsets. Some games are still excellent but are being hold back because they still use the awfull Resolution multiplyer instead of just using the native resolution, and support of controllers (I’m still amazed at how bad Journey has implemented the Vive controllers, it seems like they just haven’t tested it at all). People are looking for content, and if they see a lot of thumbsdown with ‘controller not supported’, ‘crashes when I start with controller A’..
    And ‘remastering’ and older VR game to also have more locomotion options would ofcourse also be a good thing. In a lot of cases it shouldn’t be that much extra work, and yet it would bring in a lot of extra money.

    • indi01

      I guess many of those devs have moved on to other projects, unfortunately it’s common to stop supporting older software, not just in gaming.

      • Andrew Jakobs

        Yeah, I know, but for a small developer it would sometimes mean a couple of days work, off the new project, but instant extra income due to the update which would fix most of the problems with the newer headsets.. Ahwell, it’s a shame some of the older VR-games are actually really good, but crippled by the problems with newer headsets (Robinson: the journey, and Alice-VR are prime examples of that, at least IMHO as an adventureplayer).

  • JACrazy

    “Later this year HP will launch Reverb G2, the first new Windows VR headset, since Samsung’s Odyssey+ in 2018”

    Hope the author realizes that there was a Reverb that came before to make this a G2, and that headset came mid 2019.

    • DeenVR

      Yep author is braindead. G1 Rev2 is best headset besides 8KX right now

  • Clownworld14

    I ordered the Valve Index months ago, still waiting to receive it… I hope the HP reverb doesn’t come out when I just get my index ffs

  • David Cole

    Based on all the info MRTV on Youtube has provided. If you don’t own a Headset and you want the best visuals you can possible get the G2 is the one to get and they finally modified the Controller where they have the look and feel of the Touch Controllers.. At $599 you are basically getting a 4K Headset..

  • Gerald Terveen

    I wonder what is hidden behind the “other” category growing that much. Is it Pimax delivering headsets?

  • mateuszapol

    Warhammer: age of sigmar is comming VR…

  • Tore Bjolseth

    It will be interesting to see if Flight Simulator VR will be able to make a similar dent as Alyx (if it was indeed Alyx). Based on the hype for the sim itself, it might just be a killer app for VR. Seeing your own city in VR, in a world that feels alive with dynamic and live weather, cars moving etc will be unprecedented. Google Earth VR was also awesome, but didn’t really want to make you stay.