Monday, August 17, 2009

Paul Graham on Enterprise Software

A link in today's TechCrunch posting about Y Combinator's "Request for Startups" idea brought me to an earlier "Startups ideas we'd like to fund" list, published by Paul Graham about a year ago. It's a terrific list of ideas, and I applaud Paul for sharing them (execution is everything!). If you're toying with the idea of founding an Internet startup and you're not sure what kind of business you're going to start, I highly recommend going through the list. I'm sure you'll find lots of inspiration.

One thing that I'd like to quote from the original article is the paragraph on Enterprise Software:

Enterprise software companies sell bad software for huge amounts of money. They get away with it for a variety of reasons that link together to form a sort of protective wall. But the software world is changing. I suspect that if you study different parts of the enterprise software business (not just what the software does, but more importantly, how it's sold) you'll find parts that could be picked off by startups.

One way to start is to make things for smaller companies, because they can't afford the overpriced stuff made for big ones. They're also easier to sell to.



ole said...

"One way to start is to make things for smaller companies, because they can't afford the overpriced stuff made for big ones. They're also easier to sell to."

Tough call. The assumption is that the small companies are still "big enough" to afford any kind of software. Easier to sell to is also relative - you need to find them, contact them and still get the product sold.

Ari @ World on a Hanger said...

It's a tough call. They may be easier to sell to. But they need to be much easier to sell to for it to make sense to target them over larger companies, because of their smaller budgets.

The best solution is to come up with a sales/marketing strategy where the marginal cost of customer acquisition is somewhere near zero. Selling to small businesses has much in common with selling to consumers.

Christoph Janz said...

You're right, Ari, they need to be much smaller indeed. If the price is in the $10-100 range it needs to be a self-service sales process with little to no human interaction. There are many great examples (37signals probably being the most famous) of this model.

Of course it's an equally valid strategy to target large enterprises, sell them an expensive solution and incur big customer acquisition costs. The most important thing is that you're clear on which of the two paths you go, because the place in the middle is tough. :)

Jim said...

I like selling to the small business for many reasons. I think they deserve better products then they are frequently targeted with and I think you have to be on your game to be successful with them. Your product has to work, be intuitive to use, have the right price, the right level of support and then you need to develop a recurring revenue stream.
I don't want to say that anyone can build an Entreprise software business, but developing a commercial off the shelf or over the air software product means you need to get it right millions of times, not just hundreds.