Wednesday, April 16, 2008

Reason #1 startups fail: software underestimates

A friend sent me a link about a programmer who was approached by friends to work on a startup, only to find that they thought his role was so easy, he'd do it for free.

Between budgets, man-hours, and even CPU requirements, underestimating how long software takes to write and test is easy. And it's not only a problem for startups, companies like Microsoft have the same problem--Vista wasn't just late, it was missing features. The good news for startups is that deadlines and requirements are often flexible, but missed deadlines kill morale.

Basically, the only people worse at estimating software schedules than programmers are managers, and the only people worse at dealing with missed deadlines than managers are programmers.

Thursday, April 10, 2008

Former Lycos CEO on the startup market

Why is it that CEOs who cashed out of failed companies early end up becomming VCs? No matter. Kara Swisher of All Things D interviewed former Lycos CEO Bob Davis, hoping for a few bits on wisdom.

"I think there is a big, big play to be made in the mobile arena, [but] ...people have not figured out where it's going to be. Look at the raw number of handsets...As it relates to content, no one has put it together, yet"
Platform matters. Remember WebTV? It was (and barely still is) a set top box paired with a keyboard for browsing websites on TVs. Despite the potentially large audience and the built-in market of newcomers to the internet, it never caught on. WebTV was a nifty idea, but Tivo is what brought TV and the web together, taking features from each component, and solving a problem that affected people.

So what's a relatively new VC do, now? Shamelessly promote a startup he funded, Quattro Wireless, that reformats the web for mobile phones and provides ads for web-enabled mobile phones. Their problem is that they're the WebTV of Web 2.0, and while trying to start an ad network--an revenue source--is admirable, if monetizing the web is hard, imagine trying to monetize a platform that isn't used for purchasing, has far less traffic, and far less space for ads. predates Adsense by several years. Maybe people should figure out what people are selling via phones before they start building advertising platforms.

Here's where Davis gets something right: when asked what's overhyped, he responded with the natural "social networking." More interesting was his take on social networking 1.0. He looks back on Lycos acquisitions Tripod and Angelfire as the Facebooks of their day. What's changing isn't the technology--neither the web, Facebook, or Google did anything to add to the internet--it's that the internet is changing from being generalized to being specific. Someone using his 1998 Geocities homepage to post personal information, a résumé, family photos, and a his research on Mark Twain would today use Facebook, LinkedIn, Flickr, and Wikipedia to do the same things. Generalized tools might be powerful, but only specialized tools find an audience.