Marc Andressen blogs that HP has agreed to acquire his datacenter automation company, Opsware, for $1.6 billion in cash. This ia an exit that makes a lot of sense for HP and Opsware.
Opsware is a company that after some twists and turns settled on providing managed services for large datacenters. In his blog post Marc outlines how the company took its original goal of providing automation to datacenters and worked up to 350 customers including many of the larger datacenters in the country. Well done.
I wonder if this sale will shuffle up the datacenter automation space?
A little before the original incarnation of Opsware (called LoudCloud) was founded in 1999, I was working in a startup (ArcanaNetworks) with somewhat of a similar mission although--at least in terms developing technologies to automate datacenters. Arcana has done quite well for itself over the years although it's interesting to see how the two companies have taken different paths. For an engineer like myself it's a lesson in how different business evolve.
From the outset, Arcana was very technology focused and developed a core technology that has kept it going to this date. It centers around an "XML automation engine," called OpsXML. In broad terms, the idea was to take high level descriptions of tasks and transform them into appropriate device commands using whatever protocol was needed. Personally I enjoyed the part of working on the original incarnation of the engine--before it was really understood what was needed. Unni Sreekumar, the technical leader and co-founder of the company, was intent on building more than just a 1x product. He wanted something he could use for years as the company grew out. Although I haven't worked at the Arcana for awhile, it's interesting to see how this has worked out as planned. The company has been able to leverage its core automation engine time and again ranging from small deployments to extensive ones. It's also led them into other product categories and building a partner network to provide services.
As I read the news today about Opsware, it's interesting to see how Opsware dogged and weaved through the business cycles to get to where it's at. It's done well. The company took on a lot of investment early which forced upon it some high business demands. At the time, I wasn't too sure how this would play out in the market. Could Opsware survive long enough to make it through as it refocused itself after the Dot Com cycle? It sure looks like it has and then some.
It did take over $350 million of equity and debt financing, plus going public to keep Marc Andressen's dream afloat. That's a lot of cash although with a sale price of $1.6 billion it looks like the water level is safe for everyone. I contrast this with Arcana that's stayed private. It's also a smaller company (Opsware has 550 employees according to Marc's blog post) and to this day I'd say Arcana is still more of a technology-focused company, but that's probably just my bias. Arcana also has managed to survive outside of heavy VC funding. Early on there was lots of discussions about going the VC route, but eventually the company settled on Angel investment from a select group that knew the industry very, very well. Personally, I think that helped a lot during the down cycles. They provided key advice, understood the potential of the market, and could see the bigger picture.



Incremental Blogger