OnLive, the Cloud Gaming provider and of late DaaS start up has been shuttered, only to be reborn as a new company.
Rumours of problems at OnLive first surfaced last week with hints of financial problems within the company, due in large part to a combination of high operating costs – said to be around $5M per month, and low usage, leaked information suggests an average concurrent user count of less than 1,500, peaking at around 1,800; far lower that the figure of 2 million signups that OnLive has previously offered. With a mix of subscription users paying only $9.99 per month and an a la carte pricing structure that offered only a marginal discount over conventional ownership its evident that meeting this run rate was clearly going to be difficult.
The story came to a head on Friday when company founder and CEO Steve Perlman called an all hands meeting to announce that it was ceasing operations, firing all its staff and was filing an Assignment for the Benefit of Creditors (California ABC). This is a form of bankruptcy that is often employed by high tech VC funded companies as a way to maximize the value of assets that might otherwise be difficult to value. At the same time moves were underway to relaunch OnLive as a new company backed by one of the initial VC partners. So far at least the gaming and desktop services appear to be running without incident, and it is already clear that the new company is hiring some of the old OnLive employees. OnLive’s tight lipped approach to communicating remains unchanged with OnLive spokeswoman Jane Anderson declining to say how many employees would be rehired, although The Silicon Valley Mercury News has reported that Perman will remain on board as an employee of the new company.
A report on gaming site Joystiq.com suggests that Perlman gambled on there being much higher demand for the gaming service than materialized. According to Joystiq’s source, Perlman explained the company’s failure in Friday’s all hands meeting, offering “There’s no way to exactly estimate how many servers we’d need. So we literally bought thousands of them, and all the equipment and networks to go with it.” Previous reports have revealed that OnLive’s servers were custom built suggesting that OnLive could not buy off the shelf hardware or take advantage of conventional cloud services to accommodate changes in demand. This pile it high approach tied up capital in ways that hampered the businesses chance of success, leading it its eventual demise.
“If you’ve got 8,000 servers and 1,600 users, how could we ever get to cash flow positive, right?”
OnLive launched its DaaS product at the start of the year to a mix of uncritical praise and more measured skepticism with concern being expressed about OnLive’s failure to comply with Microsoft’s licensing rules for Windows 7. OnLive made a ‘stealth’ update to its service in April, dropping it’s Windows 7 based VDI service in favor of a Windows Server 8 R2 Remote Desktop Services offering. While this move eliminated the threat to the service that attempting to run a set based on a licensing model that was not compliant with Microsoft’s licensing policies, other questions about it’s service and data retention provisions remained.
At the time I raised concerns that OnLive would have difficulty maintaining the appropriate licensing compliance exposing it to significant additional costs and treat of audit from Microsoft. The cost of licensing per device makes the delivery of free and $5 per month DaaS services a risky proposition. At $10 per month per user for the SPLA for MS Office, OnLive was loosing money every time a customer signed up. Microsoft’s licensing rules do not permit license reassignment within 90 days in the event that a customer stops using the service, meaning that OnLive would have to maintain significant additional licensing capacity to accommodate customer churn. Even if OnLive were able to secure a substantial discount for an Office SPLA subscriber license, the $5 per month service is still a risky proposition.
With OnLive 2.0 rising from the ashes, regardless of any possible viability of OnLive as a gaming service provider, my position on OnLive Desktop remains unchanged. The flaws inherent in both the remote display technology and the terms of service of OnLive Desktop make it is impossible to recommend it as a DaaS platform for even the most casual use. My recommendation to Perlman would be to abandon DaaS and focus on gaming, or at least hire someone who knows how to turn OnLive Desktop into a viable product.