Vinod Khosla Helps Startup Entrepreneurs Think Bigger

I had been planning to talk in today’s column about the shortcomings of the Google Chromebook, which are numerous. But I’m in too good of a mood this week to play the critic, partly because of a terrific Silicon Valley event that I attended Tuesday night called Get Bigger! With Vinod Khosla. So instead, I’m going to tell you about the event itself, which featured the legendary Sun Microsystems co-founder and venture investor on stage with the CEOs of six early-stage startups.

Microsoft hosted the event at its Mountain View campus, and Joyce Park of 106 Miles, an increasingly active and important group of engineer-entrepreneurs based in and around Silicon Valley, was the organizer. Park—who is the co-founder with Adam Rifkin of stealth startup PandaWhale, but is perhaps better known for being the first person ever fired for blogging—came up with an unusual and refreshing format for the session. In six 15-minute bursts, Khosla alternately quizzed and advised the CEOs about their business strategies. In the process, he provided a wonderful glimpse of how his own mind works when it comes to evaluating investment opportunities and mentoring startup founders.

Vinod Khosla on stage with David Johnson of ArmedZilla

Below, I’ve summarized some of the nicest gems of startup wisdom bestowed by Khosla in the course of the interviews; it’s stuff that any entrepreneur working on an Internet or social media startup should listen to. Alongside my (fairly faithful) paraphrases of Khosla’s words, I provide some context that shows how each point came up. But first, a rundown of the six companies:

Stipple, Rey Flemings, CEO. Stipple turns Web images into mini-stores by allowing publishers to annotate the images with links to product information and e-commerce sites. (As it happens I’ve written two articles about Stipple—one after their Series A funding round and other after the startup rolled out a new set of services for publishers.)

ArmedZilla, David Johnson, CEO. ArmedZilla is a social network for members and former members of the U.S. armed forces. Veterans can create profiles documenting their military service, and this data can be used to help connect them with benefits, services, and products.

Oomnitza, Arthur Lozinski, CEO. Oomnitza is building a suite of cloud-based, mobile-compatible, highly customizable enterprise resourcing planning apps, including apps for tracking physical assets, expenses, and timesheets.

Relevvant, Craig Davis, CEO. Relevvant is building a system that helps advertisers tailor SMS-based advertising and promotional campaigns for specific psychographic, demographic, or geographic groups.

CodeLesson, Jeffrey MacManus, CEO. CodeLesson organizes 4- to 6-week online training courses for people who want to learn Javascript, Ruby, MySQL, Node.js, and other computer languages.

Zerply, Christofer Karltorp, CEO. Zerply is a professional networking and jobs site for “creative class” individuals; it’s like LinkedIn with portfolios.

Now on to Khosla’s observations (in italics) and my own commentary/context.

There are three possible reasons to start a company: so that you can be famous, so that you never have to balance your checkbook again, or just to make friends. They are all good reasons, but if you are clear about your reason, you will build a much more successful company.

This came up in Khosla’s conversation with Rey Flemings of Stipple. Flemings confessed that that he and his co-founders aren’t really in it for the money—“if we were we’d probably be doing something else,” he said—and that everyone on the team would be doing something related to images on the Web, even if they weren’t getting paid for it.

Passion for what you are doing is the single biggest determinant of success or failure. Get passionate, believe in something, so that when the problems come, you can survive the roller coaster ride.

–Prompted by Flemings’ confession.

The best companies are created when you create a new experience. Facebook didn’t come along and do better by being Google; they created a new experience.

Khosla was asking Flemings whether Stipple’s success depends on retraining Web users to mouse over images, so that they can discover the linked product information and offers. Flemings said people are learning from Facebook that images can be tagged and explored; Khosla advised him to find ways to identify Stipple with the experience of shopping through images.

Startups are about creating the most options at every step of the process, since you never know where the path is going to lead you. I liken the early stages of a startup to a roundabout. There are six roads leading in different directions. Too many startups, egged on by their VCs, pick a road too early. Before you decide to spend four or five years pursuing this road, it’s valuable to go around the circle a few more times.

This point came up in a couple of different contexts—once with Stipple, when Khosla was advising Flemings not to focus too soon on just one way of monetizing Web images, and later when he was quizzing MacManus about CodeLesson’s business model, which currently revolves around fees for programming courses (a potential problem when competitors like Codecademy are offering free instruction).

Figure out what your unfair advantage is. Look for unfair forms of credibility. Go hire the one person your customer would take seriously. Lock up one critical feature that every enterprise needs.

Khosla made this point several times—it’s apparently one of his favorites. The first time, he was urging Flemings to figure out features Stipple could add that Getty Images or other players in the publishing and advertising business would have a hard time duplicating. Later, he gave similar advice to Lozinski of Oomnitza, whose enterprise applications are similar to software made by SAP, Oracle, Microsoft,, and others. If Oomnitza came up with a feature that every business wanted—the way Amazon got an early patent on one-click purchases—it could be in position to grow much faster, Khosla argued.

Be the integrating force.

Another point related to Oomnitza. “SAP can do SAP really well, and Oracle can do Oracle, but Oracle will never do SAP,” Khosla said. If Oomnitza offers a way to help companies work across their existing enterprise software systems, “then you have a proposition that none of your competitors can beat.”

One of the most under-invested ideas in the Internet space is the idea of emotion. If you think something is beautiful, you want it. Read Fascinate by Sally Hogshead and engineer at least one of her seven emotional triggers into your product.

After Johnson’s initial description of ArmedZilla, Khosla seemed underwhelmed. He felt Johnson was emphasizing the site’s utility to veterans, and missing its potential appeal as a place for them to forge social and emotional connections. “I would love to see a product that had emotional connections as a principal target and utility as a side benefit,” he said. “That gives you a lot more stickiness and a lot more affinity, in the real sense of affinity. Lots of ‘affinity’ products fail because startups have failed to engineer emotion into the product.”

Your liabilities are just as important as your assets. You build a business plan to take advantage of your strengths, but you also need to engineer the business plan to minimize the impact of your liabilities.

That was Khosla’s preamble to a question for Johnson about ArmedZilla’s liabilities. Johnson replied that one challenge for the startup would be figuring out how to “open the right doors” and “play well with others” in a field—veterans’ affairs—with many existing organizations, interest groups, and service providers.

I don’t believe in business plans. I have never seen one that was accurate. If you are going to do anything, do a flexible plan that says, “Here is what is important, here is what I’m trying to test, and here is the sequence of testing I’m going to do.”

Johnson seemed set on building ArmedZilla as a standalone social network as the best way to serve his target community, but Khosla urged him to question that assumption and test a different approach—offering an ArmedZilla app within Facebook. “I would want to test both,” he said. “One of your liabilities is reaching all of these people, getting them signed up and started, and you should look for any way you can to reduce barriers.”

Appoint one person in your company who will be at every senior staff meeting and will only represent the users, the end consumers.

Khosla made this point to Davis of Relevvant, who spoke about the power of social data to help brands like MTV or Victoria’s Secret send ads and promotions only to the most receptive audiences. Khosla took the consumer’s point of view, asking “What’s in it for me?” No matter how targeted the mobile ads, Khosla said, they might still come across as spam-like. Having an executive appointed to the role of “consumer advocate,” Khosla said, would help the company to remember the end consumer-“the guy who makes the final decision.”

It’s fashionable to use the word “social,” but you want to be very clear what the value proposition to your consumer is of this word “social.”

Speaking with MacManus of CodeLesson, Khosla said he thought the company should look into ways to expand beyond its existing instructor-student interactions and allow students to help each other. “That makes the experience better for the person learning, offloads some of the burden on your professionals, adds scalability, and has the student who helped feeling really god about what he contributed—he is now a member of this community.”

A very specific value proposition is valuable.

This came up in Khosla’s interview with Zerply’s Karltorp, who said the company was struggling with how to expand beyond its initial market of Web designers to the larger community of creative-class professionals. Khosla said it was a good strategy to start out with a product for a specific group—“To get intense involvement, look for narrow niches, where you can get purity of branding and word-of-mouth propagation,” he said. But to get really big, Zerply might need to formulate a plainer value proposition. “Many years ago, we started a business school in Hyderabad, India, and we made the value proposition very simple—‘If you come here and spend one year, your salary will increase 300 percent.’ The school turned profitable very quickly,” Khosla said.

You don’t get paid for solving non-problems.

This came during the audience Q&A session, in response to a question about the future of genomics and personalized medicine. “This is a very promising area, because we are starting to see these vast amounts of data,” Khosla said. “For some people, that is a problem. For me, it is an opportunity. I would encourage you to explore that area.”

In the “sergeant style” of management, you decide where you are charging and you get all your troops moving toward that goal. But most of the time, your objectives aren’t clear and you don’t understand your market. You have to be flexible and ready to change as you learn things. I call that the “shepherd model” of managing. Some of the sheep are going left, some are going right, some are going backward. But generally, the shepherd is corralling the sheep forward. When the road isn’t clear, that kind of model finds the greenest pastures.

—In response to a question about whether Khosla is in favor of startups “pivoting” to new business models when their early plans go awry.

Finally, I’ll leave you with an extended Khosla quote in response to one audience member’s question about whether there’s a “valuation bubble” in the startup world.

I read a blog post that said that in the first half of this year, there were 26,000 new startups created. That is an incredible number. Whether there is a bubble or not, there are some fundamental changes going on. The post-PC world has created a new set of opportunities that will result in some big, interesting companies. Twitter, Facebook, AirBnb, and Groupon have already happened, but there will be 20 more like that. I believe that Marc Andreessen’s Wall Street Journal article was correct, and that the opportunities are increasing. However, whenever that happens, everybody starts jumping on, and you start seeing silly ideas. You are seeing some of that now, and valuations are clearly getting ahead of themselves.

But if you are stupid enough to invest in a bad thing at a high price, that is your problem. The relevant question is about mentorship. With all this frenetic activity around angel funding, with some angels doing 20 investments a month, they can’t possibly mentor and guide their companies. And if you miss that, it can take a high-potential, high-probability idea and make it into a high-potential, low-probability idea.

At the same time, VCs aren’t able to raise money, while the number of angel-funded startups is still expanding. There will not be enough money to fund all the startups. So picking your partners early is important. More than anything else, get the help you need. You only know 10 percent of what you need to know to build a successful startup. Over time, you will learn the rest of the questions, and having a partner who has screwed up enough times will help you find those questions and address them earlier.

Wade Roush is a freelance science and technology journalist and the producer and host of the podcast Soonish. Follow @soonishpodcast

Trending on Xconomy

By posting a comment, you agree to our terms and conditions.

Comments are closed.