Right now, there is a lot of excitement in Virtual Reality, and CES had a section dedicated to it. But there is little excitement in this office, as we think it’s a damp squib.
Everyone mistakes two ideas. Idea one is that it would be very cool to go and live in another reality for short periods of time, in order to experience things that we are unlikely to experience otherwise in our lives. Idea Two is that you can do this with a silly headset, built around a Smartphone, which not only makes you look odd and behave oddly, but like most other things that a smart phone offers, it makes you bump into things. It is this duality of thoughts – VR is going to be cool, but so far, we cannot make it work – which has dogged the technology since its arrival in the 1980s.
Back then it cost as much as $750,000 to build an effective HMD, and they made people throw up they were used for more than 15 minutes, due to the two streams of stereoscopic viewing being out of phase, due to the processors being too slow to keep up. The misalignments made people feel like they wanted to vomit or gave them migraine headaches. The processing is way faster these days, so at least this should not be the major unsolvable problem this time around.
So just how “successful” or “big” is virtual reality is going to be this time around, and how soon? And what problems do we have to overcome to get there. During 2016 there were many attempted answers to this question and we just wanted a brief discussion about this, in a week when the tech has just been hyped, and at a year end, when forecasting on the subject did little to reassure us.
Forecasts have ranged from this becoming a $162 billion market by 2020 from the indomitable, but rarely right, IDC from $5.2 billion today; to being a $120 billion market in the same timeframe, according to Digi-Capital,
At the other end of the spectrum Deloitte Global said that VR had its first billion-dollar year in 2016, with about $700 million in hardware sales, and the remainder from content, estimating sales of about 2.5 million VR headsets and 10 million games sold. Another report from Markets and Markets estimates that the market will grow from $1.37 Billion in 2015 to $33.90 Billion by 2022, at a CAGR of 57.8%.
While yet another from SuperData said last year that the market would reach $40 billion by 2020. And then slashed its forecast halfway through the year by 20% or so on sales of wearables being down.
The most interesting forecast of 2016 for VR comes from Goldman Sachs, who could not make up its mind between $80 billion, $182 billion and $23 billion. There is not much point to a forecast which lets you take your pick on the outcome. Perhaps it’s to do with the nature of the forecast – Goldman Sachs just wants you to be aware that it’s an arriving technology, and to invest in the stocks, others really need to help Venture Capital groups ascertain market size.
The problems with these Goldman Sachs forecasts and any that are higher, is that by contrast Notebook sales are just $111 billion, and games consoles are around $14 billion. The notion that VR, unformed and ephemeral as it is, will be bigger than gaming hardware, is tough enough to swallow, but being bigger than one of the biggest established platforms, is totally ridiculous.
We looked at one of the VR videos from the CES show. Everyone in them is young, and that’s not a criticism, but we were young once and it happened to be about the time of the first VR wave. We were equally impressed with good VR, but it just took too much money to create it. The VR industry was sure that what it needed to become cheaper were standards and to bring down costs of both making virtual worlds and delivering the compute power to run them.
In high tech if you wait ten years there is usually enough power by then, and then back in 1998, just as VRML 2.0 the Virtual Reality Modelling Language was made ready, all interest in VR dissolved, until it was awakened a couple of years ago.
In the intervening years gaming companies had to solve the problem of developing once and deploying on Sony, Nintendo, Microsoft consoles and the PC, and it helped when each of them got out of the purpose-built chip markets, and came back to the X86 Intel processor base. With that came gaming engines which output to any environment and which can run Virtual Worlds in real time, such as the Unreal gaming engine and the CryEngine. We have OpenGL, we have cross platform playing engines. Certainly, right now all of the technical hurdles to create what was desired back in the 1990s, are now overcome. And we have the addition of 360 degree cameras, also now possible due to more on-chip, battery driven, CPU power.
But the VR industry, and forecasters in particular, have gone straight back to the same examples that it had in the 1980s and 90s – Use VR to sell real estate, or to design your kitchen, and you can see what your new home will look like. People envisage a kind of super Google Street View, either in a headset or even on a flat screen.
The other constant example is VR porn. Funnily enough at CES last week there was one VR Porn distributor cautiously on show, Naughty America, and a number of the popular mags did write ups with porn content targeted at the Oculus, HTC, Samsung, Carl Zeiss, Sony, and Google VR devices. The exhibitor suggests that VR can only become around a $1 billion industry by 2025 despite claiming that Google Trends shows searches for ‘VR porn’ up by 9,900% over the past year and a half. This type of modest ambition does not sit well with those other more ambitious VR aims. But it may be closer to the truth.
We think there are clear reasons for this – the viewing format is not yet quite right, and many of the specialist areas are just that, specialist and will require a large investment to realize them.
We accept that Video games, live event sharing, real-estate sales, even designing a new sitting room or kitchen walking around at the height of an 18 month old child to see what he/she can reach, all have their place. We are less sure of health, education, engineering and the military and here’s why? All those applications allow for buying one rather expensive, fully configured, real time world and using it for a high value purpose. On the other hand watching video, is already a pretty cheap exercise and in order to replace that with a VR version we have to shoot for ubiquity of playing device and in the same way that most people did not want to put on 3D glasses when they watched video, most people do not want to have to view video and in the process become blind to everything that is going on around them, and – bumping into things, unable to see other people. And the few who are comfortable with that only want to do it for short periods of the time.
Gamers already play in a separate room, with a TV in front of them to share the experience. They would be happy to sit on a room and play a VR game. Of course, a room of 4 gamers will need four headsets and that could set them back $1,000 collectively, but that side of things is in no doubt. It will happen. But to presume from there that video watching “en masse” in the millions of homes, will follow suit – that’s not a done deal.
Like we said video games, live event remote viewing, real-estate sales – all those can build a business around the technologies which are already in play. To conquer the other markets we need new inventions, and that’s what makes these hard to forecast, and is why the forecasters cannot agree on the numbers.
In the end the key to all of this is “How much would you pay, for how many minutes a day,” for a VR version of what you are doing.
The second issue is, if not an HMD (head mounted display), then how would want to view VR? Well how many ways are there?
VR on a flat TV screen might be navigated with a mouse or tracker ball, being used to turn the angle of view? Augmented VR allows you to use visual overlays, which you can see right through and interface with the real world and the virtual world at the same time. But even this can have problems?
One device on display at CES was inside a motorbike helmet, offering a heads-up screen in the field of view, but only as an overlay – something like a map of directions could be placed there? That’s fine as long as the driving authorities don’t ban it as a “distraction” likely to cause accidents. And anyway it competes with auditory instructions as well, which simply don’t need an expensive screen.
You may agree that while there is little in life more pathetic than a slightly dumpy, couch potato, holding a virtual gun, and wearing a VR headset and pretending to be a marine in the middle of a virtual gun battle – this probably remains the biggest turn on for the VR market.
So, despite the fact that anyone could buy VR headset from top of the range down to Google Cardboard in 2016, it still did not go mainstream in a big way. And given the fact Oculus and HTC – the two market leaders still will not give out sales numbers publicly, there is every chance that VR won’t even be mainstream come 2020.
Which means these forecasts may end up being right, but will have a market START date of 2020.
The other thing about 2016 forecasts was that each and every one of them assumed that throughout the next five years it was the hardware that would continue to take the bulk of the revenue. We think that the longer it takes for VR to become mainstream, the more likely that content and software will take the lion’s share of the pie, which is why after years of VCs investing in VR hardware platforms, all the smart (VC) money is now going into content or software.
VCs have so far put some $2.3 billion into VR start ups, and during the third quarter of 2016 alone added another $500 million.
There is so much VC funding in VR that it has started its own alliance the Virtual Reality Venture Capital Association, with 42 members. Normally VC players would hate each other and keep their distance, and this association can only reflect their uncertainty about this market.
So, we know there will be tens of millions of VR headsets because some smartphone vendors have started giving them away with phones, and we know that VR content is fundamentally more compelling than flat screen content, and that this in turn makes VR advertising the hottest thing right now. But if it never gets out of the low millions or low tens of millions globally, then none of that counts a jot. And to do that it has to be built into screens we already use – either phones, TVs, tablets or set tops. Right now, the only one of those which has enough pixels for decent quality is the TV and most broadband lines (never mind mobile services) simply do not have the bandwidth to support them at high enough resolution.
This month a movie called La La Land has captured the imagination of movie goers throughout the US and Europe. LaLaLand is a story, set in California, about people who are out of touch with reality. Perhaps it would also be a good description of where VCs are in conjunction with VR right now.