That’s the advice I’d give any aspiring web entrepreneur. Take a market, find the niche within in, identify what their itch (the thing that’s annoyinng them) and provide a solution to the itch.

Take three successful web apps:

• Eventbrite helps you manage tickets for your event. The niche is people running small ish events (20 to 100 people) who can’t afford big conferencing systems but don’t want to have to di it manually. Eventbrite charge 50p per ticket (pretty expensive if you’re running a big 1000 people event but for under 100 people it makes it worth it)
• Unfuddle is focusing on computer software developers (a pretty small niche) who need what’s called a “subversion repository” - it’s a place where you can store your computer code so that others can work on it at the same time. Everyone needs a repository like this when programming but they are a hassle to set up. From around $4 a month unfuddle runs it all for you (for 4 projects) as well as project management tools thrown in.
• Freshbooks solves the problem of sending out a regular monthly invoice (by paper or by email) for a set amount. MicroAid’s newsletter tool: Newsletto uses it to great effect, especially when billing schools with www.school-newsletter.com who still need a paper invoice and can’t use credit cards.

In fact school newsletter is a good example of how web apps work best when targeted at specific niches. The MicroAid team in fact gets more interest from users who want the school newsletter than from those who are looking for a generic newsletter tool despite the underlying product being essentially the same.

Too many web entrepreneurs start too general and never find an itch in a niche (try pronouncing it the American way as ‘nitch’) and their web application dies because anyone could use it but there wasn’t a specific set of people who had to use it.