Interview with Steve Goldman, CEO of Isilon

The company was originally founded by Sujal Patel and Paul Mikesell, two Real Networks guys. Their epiphany in being associated with Real Networks, was that storage systems used for rich media content that was streaming out over the Internet, was completely inadequate. Digital content is very different from traditional data that is stored, whether word documents, email, or databases. That led them to realize that there is a huge opportunity, because digital content is exploding.

Currently, Rich Media represent less than 5% of all stored data, but within the next two to three years it will balloon to over a third. We are talking one and a quarter million terabytes by 2006. For comparison, the entire text of the Library of Congress is ten terabytes. We are talking about a huge explosion. It is around you every day. When you think about DVDs, television, songs downloaded from Apple’s iTunes, digital cameras, etc.

Exactly, if you look at the specific markets, Media and Entertainment was the initial focus of our business. We identified them, because their pain is so acute with the management and storage of all this digital content. They are not in the business of running storage environments, they are in the business of creating movies and videos. One of our design goals, that makes us so unique is that our systems are very simple. For example, on my first day on the job, I was able to install our system in 20 minutes. Typically, with a traditional storage product, that same process would take two days.

Any company using digital content in their business is dealing with four things: 1) creating, 2) managing, 3) delivering, and 4) archiving that content. Isilon IQ is designed to work and live across that whole digital content lifecycle ? from production to archiving. Until Isilon, these companies did not have a storage solution that specifically met their needs or addressed the unique requirements of large, graphical files. Our customers are often trying to use an incumbent storage vendor?s product to store and manage the content it wasn?t designed for — the result is like trying to fit a square peg into a round hole. These traditional storage systems are big, complex, and hard to manage ? and not designed for large files and growing stores of digital content.

For example, one of our local clients, Corbis, has 70 million images of which three million have been digitalized. If you think about it pragmatically, they should not be looking for images in file cabinets, but rather have digital access to their entire image collection. A digital collection allows Corbis to access content, edit that content and deliver it far more quickly than before, resulting in new ways to earn revenues from their existing inventory. Corbis has ordered additional storage capacity from us four times and has more than 35 terabytes of capacity installed today. After installing Isilon IQ, Corbis could then repurpose their original EMC storage system for database and email applications, which is the type of data for which it was designed.

The way people thought about storage as an industry has changed very little in the last two decades. The basic design goals of the big incumbent vendors have been to increase the capacity of the storage system, but not address the speed of accessing that content. There has been a lot of pain associated with that. There hasn’t been a disruptive force like Isilon to come in and really address the issue of why it is so difficult for companies to manage their content efficiently. Storage should be a back office function and not one where legions of people are managing a company?s storage requirements. Going back to the Corbis example, they had one person on staff who fully understood the inner workings of their EMC environment. That isn’t a single point of failure from an IT perspective, but a single point of failure for their entire business. Now, with Isilon IQ, it takes half of one person’s time and anyone in their IT department can manage the entire cluster. There is this real-world, tangible need to make storage simple.

The three things we are going for and the reason companies are buying our solution: high availability, scalability, and simplicity. These systems are extremely important and must always be available. Also, a storage solution must be able to handle this explosive growth seamlessly. The two buzzwords we use is that it is a distributed cluster architecture, so that you don’t have all of your data in one place, and you cannot have a single point of failure. Simplicity is what all of the current systems are lacking. The incumbents don’t have incentives to change because they make a lot of money with consultation services.

In terms of company lifecycles, we are still in the beginning phase of getting the word out. In the 20 years that I have been in High Tech, I have never been associated with a products whose advantages are so readily apparent, and elicit such excitement when you do a demonstration. What I have seen in scores of demos is the literal jaw drop that it installs so easily, that you can grow it so seamlessly, and that you manage so readily. So the answer to your question, is that we need to show it to people. From a business perspective, we are still building out partners, distribution and brand so that we don’t have to tell the story one person at a time. The credit goes to our engineers and the founders of the company that had the vision to really solve this problem.

I have had experience with both models in my career, and we are essentially taking the message directly to the audience in the early days. Building up that critical mass of blue-chip customers that validate everything we are doing.

Exactly. Clearly, companies are judged by the marquee value of their customers and the interesting ways in which they use your products. The second thing is to build out a channel. Those take a variety of modes from people who resell your product as is to those people that integrate it into another system. For example, one of the obvious uses of our product is with the use of ultrasounds and x-rays, which are now digital in Hospitals. The companies that make those systems, such as GE and Siemens, would be a perfect fit for incorporating our technology into theirs. This is an excellent example of an OEM opportunity. Our core intellectual property is in our software. The other aspect is that we are currently focused on North America, and will growing our international capabilities this year.

We have the advantage of being the strongest player in this new product category; storage for digital content. This market is currently a $20 billion business, which allows us to clearly target a market need that is growing extremely rapidly. My main point is that digital content is unique and different from traditional data and that the established companies are not addressing.

I can’t answer that question in full detail, as I am relatively new on the job. I can answer the question by letting you know why I am so excited to come here. Beside the ability that I can make a large contribution to the company, I really thought that this is one of those rare instances where the time is so good, the opportunity is so huge, and the solution is so strong. The large storage companies retrenched, and focused on what they do well, which is building large complex systems for traditional storage. As we discussed earlier, digital storage only represented a few percent of the overall market, so they weren’t too excited by this opportunity. Meanwhile Sujal and Paul and world class group of engineers, were able to build a wonderful solution while the Giants were in effect sleeping. Also, the data is fundamentally so different that you will need different systems to handle the distinct needs of data and digital media; small versus large files, random access versus sequential access.

I have a lot of experience in long term plans. I worked at F5 Network for 6 years, and was the 15th employee. We all helped grow it from 15 people to the 600 person company it is today and take it public. One of the goals we have laid out is taking Isilon public by the end of 2005, which is a very ambitious goal. I laid out a 10 quarter plan when I got here, and there are a lot of things we need to do along the way with customer acquisition, growing revenue, product leadership, etc. I think that given how enormous our opportunity is and the lead that we have we are a very realistic candidate to achieve that.

That goes back to our initial strategy of targeting the Media and Entertainment vertical. The awareness of the issue is very high within that market, and there is less missionary work to be done on our part. As we move into more expansion markets through this year; Oil and Gas, Bio IT, security agencies within the Federal Government there is additional education that needs to be done. Oil and Gas, the amount of terabytes that is required to map a square mile is off the charts. Those are all examples of where our underlying technology is a perfect horizontal solution and where we believe there is reasonable levels of awareness.

Our biggest obstacle to overcome is “Who is Isilon, and why should I buy this mission-critical solution from you?” That goes back to simply proving ourselves over time. One of our advantages are the customers we have on board to date, including Paramount, Corbis, Experience Music Project, Technicolor, Digital FilmWorks and ResearchChannel. Another thing we have done is to partner with IBM Global Services so that we provide four-hour response time anywhere in the world to anyone who has an Isilon product. We based our system on industry standard hardware, so it has off-the-shelf components, with Intel processors, standard memory, etc. We have an advantage over others because our system was designed specifically for digital content, scales at a tremendous level, is extremely easy to use and mange, and is half the cost of other systems.

The use of the Isilon solution allows customers to continue to use their current system, but in a way in which it was meant for, for standard storage requirements. With our scalable approach we have seen customers maintain their older systems, and begin to do a lifecycle management approach. It is very easy with our architecture to manage that.

My pleasure.