When OnLive first launched to great fanfare in 2007 it was was met with a mixture of awe and cynicism. To some it was the herald of a new technological dawn: an era which would see us return to terminal computing, powered by distant humming mainframes - a model which would lower the price barriers for PC gaming and allow publishers to revitalise the market.
For others it was another interesting anomaly - an idea which had escaped from the drawing board well before its time. Without the infrastructure and install base to support it, they predicted that OnLive would remain an ambitious footnote in history of gaming's evolution.
Article continues below
The truth transpired to be somewhere between the extremes. OnLive's technology did work, plenty of publishers signed up for the service and, by giving away plenty of the micro-consoles which allowed games to be played on TV instead of a PC, the install base was grown significantly. Nonetheless, the UK's infrastructure did prove to be inadequate, with jerky and unresponsive performance rendering many fast-paced games unplayable, and subscriptions quickly fell away - the public wasn't willing to pay full price for the experience and publishers saw no reason to cut their margins and encourage channel cannibalism.
Today, the business is pulling back the covers on a fresh new approach, with new business models, management and technology. OnLive is back for round two.
"in the last year we've really looked at what makes sense as a business model and what things were broken"
"When we initially launched in 2007, OnLive was positioned as a platform," general manager Bruce Grove explains. "As a platform it was dependent on getting content, and users thinking it was great - as well as a good internet connection. It's now five years since it was demonstrated at GDC, four years since launch. In that time a lot of things have happened, but in the last year we've really looked at what makes sense as a business model and what things were broken.
"First off, as a user, you looked at it and realised you didn't own the game. Ownership is important. If the network is down, the company goes away, anything happens, everything I've paid for disappears. You were asked to pay the same as you would for a console or PC game. So there's channel conflict. People also don't want to buy the same game twice.
"For the publisher, it was channel conflict. too. There was work involved in getting a game to OnLive, previously. If a user bought it there and then didn't buy it elsewhere, publishers were costing themselves money for the same sale. So there was a whole chain from beginning to end that wasn't working."
It's a frank enough explanation and Bruce Grove is the right man to deliver it. He's been with the business since day one, so has more than enough knowledge of the tough times it's been through. A third of the team's original members are still at the company. Working alongside them now, however, is a new team of executives with a worthy pedigree.
Mark Jung is the business' executive chairman and acting CEO. He co-founded IGN before selling to Fox, then was CEO Vudu, a streaming box company which was then bought by Wallmart. Jung's previous partner at IGN, Rick Sanchez - who went on to work at Metaboli and Playdom - is also on board. On the business development side is Carrie Holder, an ex-EA employee with deep industry ties, whilst Dom Gordon, previously of Gracenote and a multiple patent holder in streaming tech, is the new chief engineer.
Enabling all that new hiring is chief investor Gary Lauder, who bought all of OnLive's assets, IP and patents in 2012, when the company underwent a very public insolvency which saw the departure of founder Steve Perlman. A long-time Silicon Valley VC, Lauder steers the company as chairman.
But a fresh boardroom isn't an answer in itself. OnLive is reinventing itself in its customer propositions, too. Enter CloudLift.
Article continues below
"OnLive is streaming technology," says Grove. "None of that has changed. Everything we've been working on has been about how we deliver games over the network. We've continued to focus on bringing a graphically great experience and to reduce the latency, as well as the features that you can't get anywhere else. We're just now starting to see technology come in that copies the stuff we've been doing for the last few years.
"We wanted to stop being a platform that was going head to head with everything and look at tech and what we could support, so we came to CloudLift. Platforms are starting to deliver cloud saves, which are a really significant thing. As a gamer, no matter what happens to your machine, you got your saves somewhere - but before, if you lost them, it was a disaster. Syncing your data to the cloud is very important, but people like Sony are now doing this for us.
"We're just now starting to see technology come in that copies the stuff we've been doing for the last few years"
"What we're doing is giving you a way to play your cloud-saved games across more platforms. Say you have Steam. You have a library of Steam games that support the Steam cloud. You've already bought the game, so the publisher is happy. You're happy. You can play it on your local machine, but now, when you go travelling you can play it on your Mac, or your laptop or your Android device. The same save, anywhere in the world. Then we you hit save on OnLive, we sync it back up so when you get home that game is ready for you.
"We have gone from being a platform that forces you to make a choice to being a platform that complements your activity and gives you more flexibility and mobility. You could be in the home, downstairs playing on the TV. You could be playing on one of the new Android set-top boxes. You could be travelling. We're seeing solutions talking about home streaming, but that can't go outside. As soon as you get outside, you can't stream from your home device."
Launching that Steam game also puts you in Steam's ecosystem, Grove tells me. That means you'll have access to your friend list and all of the social options that entails. It also means you don't even have to download the game first - just link OnLive to your Steam account and it will give you access to any compatible games you've paid for, meaning you can play remotely until you get home.
Whilst there are obvious advantages to being able to pick up a Steam game where you left off from you laptop or tablet, Grove also points out that this could be a revelation for Mac gamers. Now, they can buy a PC game from Steam which hasn't been ported to Mac and play it via OnLive.
Article continues below
"At the minute it's PC, Mac, Android. In the US we have LG and Vizio connected TVs and some set-top boxes and we plan to bring this to more devices that make sense."
iOS devices would seem like the obvious inference, here, but Grove says that there are a few barriers to be overcome there.
"We're working on that. The idea is to extend to more of the platforms. As much as anything that's about working out the right business model - the rules say you can't actually be a store within a store, so we have to comply with that. We've got to work out the right model - we're on the Google Play store, so we're hoping to do something similar to that. Obviously iOS is a very popular platform so we have no reason to not support that.
"For us now, hardware is all of these devices," he explains, although I'm assured that the original micro-console will still work if customers want to play through a TV. "We're a service, so really from our perspective, the monetisation model is that there's a subscription. Just in the same way there is for Netflix or Lovefilm, any streaming services. This is subscription-based gaming. We don't care where you bought the game from - we will examine your library and import anything that supports the cloud sync, based on the deal we have with publishers."
Currently that list of publishers includes Warner Bros. Deep Silver, Born Ready and Codemasters, but Grove says many more deals are in the offing. Origin and UPlay are also mentioned as future targets for collaboration. A couple of major gaming networks with heavy-cloud focuses are missing though, so I ask Grove about XBL and PSN.
"We're a service, so really from our perspective, the monetisation model is that there's a subscription. Just in the same way there is for Netflix or Lovefilm, any streaming services"
"Well...bear in mind that they're their own platforms so that's probably not going to be the next one we move to," he says, diplomatically. "We're still really a PC-based platform, it's really about extending the reach of PC gaming platforms like Steam, Origin and UPlay.
Article continues below
"That gives us more opportunities to partner with digital retailers. They're selling keys, we're selling a service that gives people more reasons to buy those keys and keep playing. It's a model where we're not in contention with everyone else, but augmenting their service."
At £9.99 a month, the subscription is non-negligable, especially given that it comes on top of the £6.99 a month price for the PlayPack collection of games which OnLive has always offered. It's not clear what sort of numbers OnLive needs to subscribe to make this part of the business a viable one, but there's revenue from the publisher side, too.
"We have the usual types of publisher models and in the background we have other ways of monetising this with publishers," explains Grove. "So this continues to be a partnership model. But for them it's a more attractive model because it doesn't cannibalise their sales, it's supplements their sales."
In addition to the consumer business, OnLive is moving into B2B work with a service called OnLive Go. Go is a "solution to many problems," a service which will be tailored individually to the needs of corporate customers. Two very different use cases are already on the books - Gaijin's War Thunder and Linden Labs' Second Life.
"They're obviously very different. In the case of Linden Labs they have a very stable audience who love what they have, but the requirements for running Second Life have become quite high and high-end PCs don't have much mobility. What they wanted was to take that audience and improve their access.
"So we built an app which is a direct launch of Second Life - it puts you in the Second Life world from a tablet or a low-end PC. We've had users testing it and so far they love it. The Gaijin model shows the flexibility of the tech. Gaijin has War Thunder, which is a free-to-play game which is great once you get it going, but it takes a big download and takes a while to get going. Our app launches directly from Gaijin's download page and lets you get into the game immediately, to familiarise yourself whilst it downloads and installs.
"The expectation is not that the customer will play the game via OnLive for ten hours, but it gets them through the first bit.
Article continues below
"What they're trying to do is to increase customer acquisition, they want the customer to be able to get a feel for what the game is straight away, without the need to wait. To be able to try it out. Then you can worry about the download later, become more engaged. It gives people that initial emotional attachment to it, reduces the breakage of being too busy and thinking 'I don't have time for that now.'"
"We have the usual types of publisher models and in the background we have other ways of monetising this with publishers"
But the crux of OnLive's problems was always the inconsistency and unreliability of internet connections. Has technology caught up with OnLive's concepts?
"We've done a whole new UI, we've continued to move forward the systems we build in the data centres, they look to all intents and purposes like mid-high gaming rigs. That allows us to really power up what the games look like, we're able to really slide things to the right. We're also all about driving for 720p at 60fps. Graphically in our world 720p looks good, but the 60fps is all about latency.
"We've added additional data centres in North America, particularly around Chicago and Seattle. We're doing the same thing across Europe. We're also continuing to optimise for fast connections. When we first started doing this a fast connection was about 30-40 mbs. We came piling in at 6mbs - we were an HD stream when Netflix was still an SD stream. Now all of this stuff is HD and I'm on a 120mbs connection.
"We can't reach everyone, people in rural areas just don't have access to this sort of infrastructure, but our heatmaps match very closely with the growth of broadband deployment. We see our users growing with broadband. So if people have a faster connection we're dropping the 6mbs cap and letting them go up to 10 or 12, which creates a much smoother experience, graphically."
OnLive CloudLift is live now.