Sep 14, 2010
“Three weeks before graduation when everyone went off and partied in the Caribbean, I came back out to San Francisco and my co-founder and I sat down and did a lot of work. I went back to MIT for graduation, graduated on a Saturday and two days later on a Monday morning we started fulltime in our new office together. We hit the ground running,” recounts MIT alum Frederic Kerrest on launching Okta while still at MIT. Okta recently attracted a $10 million backing from such venerable investors as Ron Conway and the VC firm Andreesen Horowitz. In Part I of this Student to Founder Series interview we hear from Frederic about Okta and how he incubated it while in school. In Part II next week we will hear Frederic’s story on how Okta pulled off backing from VC and angel investor heavyweights.
MITER: Tell us about Okta in your own words and what makes you excited about building this company?
FREDERIC KERREST: Absolutely. Okta is a cloud-based application management platform to help companies better manage, secure, and integrate the applications that they use – both on-demand cloud applications and on-premise applications behind the firewall. In the 1980s, 90s and early 2000s local area networks grew up as a function behind the firewall. IT managers would manage their company’s users, servers and systems all behind the firewall. When an employee was let go, the company would just take away their RSA token and their employee badge to the building, and the employee was effectively locked out of all corporate systems. Today’s corporate IT is a very different world, with on-demand systems expanding throughout organizations and the advent of what we call the Cloud Area Network – a hybrid IT infrastructure that requires new tools. It’s a paradigm shift, and we’re still just in the 1st inning – which is what makes it so exciting!
My co-founder Todd McKinnon and I met at Salesforce.com. Todd was the head of engineering there from 2003 through 2009. I started at Salesforce.com in 2002 as an early member of the sales and business development organizations and built a number of businesses including their OEM and Wireless groups. While at Salesforce.com, and thereafter, in our interactions with customers Todd and I both realized that as companies adopted one, two, three, or more cloud applications, they essentially were developing a Cloud Area Network. They didn’t know it, and it didn’t behave like a Cloud Area Network – it didn’t have single sign-on, central provisioning, central user management, etc. But once a company adopted even a single on-demand enterprise application, the IT department started running into problems. How do I manage the users? How do I manage the systems? How do I know who is using what services? How can I monitor this activity and report to my auditors? How can I make sure I am in compliance? A host of new problems arise..
Software-as-a-Service (SaaS) is a very recent phenomenon over the last decade, and everyone here on the management team and in the engineering and sales groups has grown up with SaaS. I’ve been in the enterprise software world for about 15 years: I started writing enterprise software at Sun Microsystems (now Oracle) and Applied Materials, then co-founded and ran an enterprise software consulting firm in Latin America deploying on-premise software. So the first half of my industry experience was really focused on installed software behind the firewall. But after spending five years at Salesforce.com, it became clear to me that enterprise software is gradually, but undeniably, shifting to an on-demand SaaS model, for myriad business and technology reasons. The SaaS industry is still small, only a ten billion dollar market, which is four percent of the $250 billion annual spend on enterprise software. But it’s growing. It’s growing 40-50% a year and ten years from now it will be 20-30% of all enterprise computing spend. The SaaS industry needs to mature – ISVs (Independent Software Vendors like Salesforce.com, WebEx, NetSuite, SlideRocket) need to offer more transparency into their applications and companies need to feel more comfortable using them because they are outside the corporate firewall and you don’t run the servers and you don’t run the systems. Today, Chief Information Officers (CIOs) are being told by their CEOs, “Business units are adopting these on-demand applications, the sales guys are using Salesforce.com, the accounting guys are using NetSuite, HR is using SuccessFactors. Hey, Mr. CIO, support these things!” And then the CIOs shake their heads and say, “What does that mean, support these applications? I don’t run the servers, I don’t run the systems, and I have no control over users.” We help those CIOs regain control.
At Okta, we think there’s a huge opportunity here to help the industry mature, by providing more transparency to customers about what’s happening with their users and the systems, ultimately giving companies a level of comfort that they don’t have today and which is impeding the broader adoption of cloud computing.
MITER: You began building your startup while at school at MIT. Tell us about the steps you took and the advice you’d give to students looking to launch startups while in school. In essence, what are the best practices for starting a company out of school?
FREDERIC KERREST: One of the reasons that I went to MIT was to learn more about the venture business. I’d done a little bit of work in the venture business before MIT but nothing meaningful, so the Summer after my first year in business school I went to work at a great venture capital firm called Hummer Winblad Venture Partners. Mark Gorenberg is the partner there who sponsored me; he’s an MIT alum and is currently on the Board of the MIT Corporation. I got to work closely with Mark, John Hummer, Mitchell Kertzman, Ann Winblad, Todd Forrest, Prashant Shah, and Lars Leckie who is now a close friend and one of the great young enterprise software investors today. I had a fantastic experience, learned a lot of things, among which was that I didn’t really want to work in the venture industry at this point in my career. It’s exciting to be in venture capital, it’s a great business, but you really need to have a lot of experience building and growing companies to be a good investor, especially a good early-stage investor. I enjoyed my time at Hummer Winblad but I realized that I wanted to go back to operating with the intention of building an important software company.
My background experience is all in enterprise software, so as I started thinking about that through the beginning of my second year, I started reaching out to a lot of my networks here in California. I started talking to people and I realized that what I really needed was a technical co-founder. I was reaching back to my network, spoke with some executive friends of mine at Salesforce.com, and they mentioned that the head of engineering, Todd McKinnon, was leaving. Todd and I originally met because we were both involved in the first acquisition that Salesforce.com made. I co-founded the Wireless Business Unit at Salesforce.com and participated in the first M&A of a small software company called Sendia, and I met Todd when we discussed how best to integrate the technology that we bought into the Salesforce.com infrastructure. He wasn’t that excited about the prospect, so while we met as rivals somewhat, we developed a professional respect for one another. But I hadn’t really spoken with him in a while.
So I think it was in January of my second year that I committed myself to starting a company. In January I came back to San Francisco during IAP, started talking to Todd, meeting with a lot of my network, and getting a feel for what I was going to do. In March Todd and I spent some time working together here in San Francisco. Three months before graduation we agreed that we were going to go into business together, and I started doing a lot of work in terms of building some financial models for investment, starting to do some due diligence in the market, and starting to get our heads around what we were going to build and how we were going to build it.
Three weeks before graduation after Finals I came back out to San Francisco and Todd and I sat down and did a lot of work. I went back to MIT for graduation, graduated on a Saturday and two days later on a Monday morning we started fulltime in our new office together. We hit the ground running.
So my recommendation really is that you need to find a partner, you need to find a team because you can’t win by yourself, and no one will invest in a solo entrepreneur, it just doesn’t happen. And you need people who have complementary skill sets to you, who you can work well with, and ideally, have the same domain experience that you do, because that’s what will give you a sustainable competitive advantage in the market.
MITER: At MIT you were working on a number of different ideas. What did you do to validate your different ideas? Why did you pick Okta in the end and discard other ideas?
FREDERIC KERREST: You’re right, during my second year I looked at a couple different things. I talked to a number of friends who gave me an idea, which a few VCs that I later talked to actually liked: the easiest way to describe it is PayPal meets Mint.com for 5 to 13 year olds. When small children get allowances they don’t know how to manage them, they don’t know how they should be managing their money, and their parents don’t know what the kids are doing – the kids are buying songs on iTunes, they are buying games online, magazines at the store, ringtones for their cell phones, and there’s no way to track it. So I thought about building a service through which parents could more easily manage and control allowances, and also teach their kids how to manage money and through which the kids themselves could more sensibly manage money.
I built a prototype and spoke to a bunch of folks. Ultimately I decided that I didn’t care enough about the business that I wanted spend the next five to seven years building it. I didn’t have the core domain expertise because I don’t know that much about consumer Internet. And I didn’t really care that much about the problem because I didn’t have any kids myself. So it was an interesting academic problem to solve, there was clearly a need, I got some good response and feedback from the market, and the venture community might have been interested in investing in it, but I decided that it wasn’t what I wanted to spend the next several years doing.
What I’ve done largely in my career is enterprise software – I’ve built it, I’ve sold it, I’ve built a company around it, a high-tech enterprise consulting firm that I helped found and run in Latin America from 2000 to 2002 (with MIT Sloan MBA ’05 alum Benton Moyer, incidentally). So that’s really what I know and what I really like. But the previous idea with the children’s PayPal service was actually a good experience because I realized, “Hang on, I know what I like to do.” I used that experience as an independent research project where I got to learn a lot about that business, but discarded it because it wasn’t what I loved to do, and if you’re going to be an entrepreneur, and if you are going to spend fifteen hours a day doing something, seven days a week, then you really have to love it.
You have to be able to sell everyone on it. You have to sell yourself on it, you have to sell your employees, because good employees are well employed – you have to convince them to quit their cushy jobs at Google or Microsoft and come work for you when you’re still completely unproven. You have to sell your investors on the fact that you have a big market and the right team and the right product and that they should give you money, which is hard to do. You have to sell your customers on using your product even if you are a new, young startup and so forth. You have to sell your vendors on extending you lines of credit because you really don’t have any credit since you haven’t been around for a long time. You really have to have a lot of faith and belief in what you’re doing. That’s the most important thing in order to decide what business you should be going into – in addition, obviously, to making sure that there’s a giant market, you have the right team, and there’s actually a need.