Joe Kraus: Confessions of a Startup Addict
This is part of my set of notes from the Startup School 2006 sessions at Stanford.
Joe Krause is addicted to startups, and during his session he offered seven lessons he’s learned from his time at Excite and JotSpot. In addition, he summarized the major differences he sees between the environment for entrepreneurs in 1996 and 2006.
Lesson #1: Persistence pays
Netscape (early on in 1995) put two buttons from its browser (the web search and web directory buttons) up for bid. There were three bidders: Infoseek, MCI, and Excite. Excite had $1M, but bid $3M – the thinking was that if they won, they would figure out how to raise that much money – unfortunately they lost the bid. Instead of packing it in, they continued to pester Netscape, acting as if the negotiations weren’t over. MCI failed to meet their goals 21 days later and Excite won the bid. In the end, the difference between Excite being a $6.7B company, and nothing ended up being 21 days of persistence.
Lesson #2: It’s all about hiring
Joe’s core hiring philosophy: no false positives. Bad hires really screw up your company. A players hire A players; B players hire C players; and C players hire Losers. Although it’s hard to resist the urge to hire for the short term – but don’t give in. Hire slowly and carefully – it will make the difference between success and failure
Lesson #3: You make what you measure
Only measure the goals you want to achieve. For example, all subscription-based businesses (like JotSpot) measure ARPA, Average Rev per Account. The idea is to figure out how much money you’re getting from the customers of your business. Jotspot had two problems: it wasn’t measuring it initially, and it wasn’t going up. To turn this around, they started reporting on it every week. The first time they did this, the ARPA was $17.36; the second week, it was $31.72. By putting the measurement in front of of everyone and giving people the right feedback, you’ll generate the right results.
Lesson #4: Better to be a trend-spotter than a trendsetter
It’s better to be an early trend-spotter, than to create something entirely new and educate the market. No startup has enough money to move the market. For example, Jotspot is positioned as wiki – but why? Because it would take too much work to position as collaborative website. By leveraging the existing focus on wikis, they can use that coverage to gain awareness, and then branch out from there later. Being early is the same as being wrong – so many companies are great, but ahead of their time.
Lesson #5: Opportunities create opportunities
When you’re looking at deals, you can never predict where a given deal will take you. It’s hard to make rational decision. As Joe’s grandmother used to say, when someone offers you a cookie, take the cookie! You never pass up a cookie that’s put in front of you.
Consider the chain of events that led to Excite getting funding:
- Joe’s college girlfriend gave him a book called Accidental Empires by Bob Cringley.
- In his book, Bob encouraged people to call him up (“I’m a cheap date”) – and ended up introducing Joe and his team to InfoWorld.
- Infoworld was looking to put archives online and needed a searchable index; they offered Joe and his team $100K to help this out. Plus, if they did a good job, InfoWorld promised to introduce them to IDG.
- They did a good job and got introduced to IDG
- At an IDG board meeting, met Steve Coit, a VC from Boston for Charles River Ventures. To do a deal, Steve needed a west coast partner, and thus introduced Joe and his team to Geoff Yang.
- Geoff didn’t know what to do with them, and so introduced them to Vinod Khosia
- In a meeting with Vinod, Vinod asked how the Excite technology would scale – unfortunately, Joe and his team couldn’t afford hard drive large enough to test large database indices. Vinod bought them one, ordering a $10K drive on the spot.
- Through that relationship raised $3M without a business plan.
The moral of the story: it’s better to take the opportunities and see where they lead.
Lesson #6: Put your business model into beta when you put your product into beta
Jotspot screwed this up, by putting their product into beta without testing business model. The problem with this is that it resulted in skewed feedback. How you make your money is equally relevant in how people perceive your product.
Lesson #7: Celebrate your successes
Startups are a lot like the Olympic trials – a series of progressive milestones. As you cross the finish line in each stage, you think to yourself, “Wow, I made it!” And half a day later, you think to yourself, “Oh my God. I made it.” You’ve finished in first, but all that does is qualify you for the next heat, where you’re competing against similar companies who made it through the last round. No milestone is the finish line. As a result it makes it very important to celebrate every success you have.
How is 2006 different from 1996?
- It’s easier than ever to start a company: Hardware is cheaper, Open Source software is free to use, there’s plentiful offshore labor, and search advertising allows you to cheaply reach audience
- It’s no easier to start a business: It’s still difficult to get loyal customers
- Funding is more available than ever
- There’s a new funding model: As it’s a lot cheaper to start a company, it takes a lot less money. When you raise less money, it means that there are more successful exits than ever, especially with people who want to buy your company. Counterintuitively, taking more money may mean less opportunities for success. Considered two scenarios: Let’s say that you need/take $250K, selling 1/3 of the company to raise that amount and resulting in a post-money valuation of $750K; to generate a 5X return means that the company must be sold for $3.75M – lots of companies that can afford to buy you for that! Compare that to a case where you need/take $5M, sell 1/3 of the company to raise that amount and resulting in a post-money valuation of $15M; to generate a 5X return means that the company must be sold for $75M – not many companies can afford to buy you for that.
nice article. There are couple of typos, Vinod Khosla that’s the right name.
At one place, it has been written as Vonid instead of Vinod.
Whoops! Thanks for the heads up – fixed!