SM: When did you actually get Rearden started?
PG: I wrote the business plan at the end of 1999 and hired my first employee in March of 2000. One year later, to the day, Microsoft unveiled their version of this. In the summer of 2000 they announced .NET, which was the foundation web services plumbing that would allow applications, devices and content to operate interactively. They then announced Passport, which was supposed to be your e-wallet.
They were still missing context. There is personal information and plumbing, but how do you bring all of that information back into my workflow? On March 19, 2001 they came out with Hailstorm, which was a collection of contextual web service objects such as myCalendar, myAddressbook, myProfile, etc. In totality, their idea was that everything on the web would connect through .NET, Passport would inform your experience on the web, and everything would be brought back into your Hailstorm identity. I was a little nervous when they announced what they termed ‘the most strategic initiative in our history’.
SM: Had you looked at context at that point? My thesis is you cannot do personalization without context.
PG: We were marching down a very big stack. Our platform had to be completely agnostic on the backend. We had to be able to plug in any piece of content, device or calendar.
SM: Does that make you context-agnostic or context-sensitive?
PG: Today we are context-sensitive inside the Rearden environment. We are location aware in everything we do. In terms of context and identity outside of the Rearden network I would love to get others out there to open up. There is a lot of talk about OpenSocial, but I have not seen much other than press releases.
SM: From your point of view I would not go context-agnostic because it is a harder problem. You can build a business on a context-sensitive platform. Microsoft ran into problems there.
PG: It was simply not in Microsoft’s DNA to be open and agnostic. This fundamentally required being locked into a Microsoft-centric world. Our view is completely opposite. Instead of technology-centric computing where you need to map to everything, it had to be user-centric computing. It needed to be about things working together on your behalf.
SM: From a workflow point of view, is Rearden a personal assistant for a knowledge worker?
PG: The initial application I wanted to build was around travel and entertainment. That involves many complex attributes such as time, location, availability, and cancellations, among other things. To the extent you could solve for virtually anything. It is one of the most complex things to solve.
When I looked at travel from the consumer space, there were others such as Expedia coming out with a lot of dollars and big backing. Trying to compete as a technology, application, and brand with those established brands would have been very difficult to do. Prosumer was just the right way to go. My assessment was that if I built out the application on travel with an eye towards anything the employee needed to purchase or schedule, then companies would pay me.
I knew a number of former Cisco executives. I got a Cisco contractor badge and in July and August of 1999 I lived on Cisco’s campus. They gave me unfettered access to everyone. I saw them as a laboratory for where the world was going. They pushed a self-service environment in the workforce optimization initiative. They had a portal which had links to a travel app, a dining app, a conferencing app and many other apps. It was good, but not as good as an administrative assistant. When I went in there and did an exhaustive analysis, I came to conclude that with the intro of every new application the adoption rate of existing applications dropped. Too many user names and passwords, none of which were integrated, led to diminishing returns. All we had to do was build, in software, a pervasive agent that could be an assistant that knows who you are.
The idea was to invoke policy, based on user identity, at the point of purchase. Doing so could save a company enormous amounts of money and you get the obvious benefit of productivity. The average knowledge worker at the time was spending an hour a day dealing with these types of tasks. My time at Cisco was where I developed the specs for Rearden.