Building Fat Startups: Delphix CEO Jedidiah Yueh (Part 3)

Sramana Mitra: How long did it take you to launch your product from that round?

Jedidiah Yueh: It took us a long time. This was my first time as a founding CEO. I’d never run a development program. I still don’t know how to code. We were four years late in shipping product. We shipped the product, but we had lots of function and quality issues when we first shipped. It took us about four years to harden the product until we could really protect large volumes of data from large enterprises.

Sramana Mitra: Your target was enterprise and not small businesses.

Jedidiah Yueh: Yes. Data de-duplication or disk-based backup is used by companies of all sizes. We were generally targeting medium to large businesses.

Sramana Mitra: What was the customer acquisition strategy?

Jedidiah Yueh: It was direct sales. We did try a channel approach, but it didn’t work. We had a hard time with the channel approach because of our software. We built a software version of data de-duplication. We tried to use existing backup channels, but we had a hard time selling software to the channel. Most of our sales were direct sales.

Sramana Mitra: What is the next major strategic move in the evolution of the company?

Jedidiah Yueh: We had an acquisition offer from EMC and we ended up selling the company for $165 million.

Sramana Mitra: How long from beginning to this acquisition?

Jedidiah Yueh: I think it was about seven years.

Sramana Mitra: There must be some other things that you did in between starting and the acquisition. Could you highlight some of those moves?

Jedidiah Yueh: Most of it was about selling the product. We were iterating on the go-to market strategy. We tried selling through the channel. We tried using the channel as a lead generation vehicle. We tried combining the software with hardware. In the end, most of our sales just came from direct sales.

Sramana Mitra: No twists and turns here.

Jedidiah Yueh: The twists and turns were all on the go-to market side.

Sramana Mitra: The thing that we try to do in this series is to provide insights into strategic issues like go-to market strategy.

Jedidiah Yueh: If I could go back in time, what would I tell myself? The biggest thing I would tell myself would be, “No Redundant Array of Independent Nodes (RAIN).” We built a data de-duplicating hash-based file system. That’s what allowed us to strip away redundant data across all the backups.

One of the other things we built was this thing called RAIN. It allowed us to cluster servers and storage so you could add servers and storage to the cluster at any time. Even if one failed, you could still read and write around it. It was a high-availability technology.

On the surface, companies want high availability, but that high availability technology was so hard for us to architect and deliver at a high quality and scale that it consumed many years of development effort. In hindsight, we saw a competitor enter into the market that they didn’t even bother building a high-availability mechanism until much later in their life cycle.

It turned out that the market didn’t require high availability. It never really required that level of high availability for backup or recovery. It was a feature set that customers all said yes to, but it turned out to be an unnecessary feature. The complexity of shipping that is what really protracted the timeline for getting a quality product out to the market.

This segment is part 3 in the series : Building Fat Startups: Delphix CEO Jedidiah Yueh
1 2 3 4 5 6 7