Ask people who have money and problems what their problems are. Solve their problems. Charge money for it.<p>Or, failing that, you could probably create a Web 2.0 Twitter-integrated mashup of social network aggregators.
Some general advice from my personal experience. It's just what I learned, I don't know if the following can be generally useful, but my 2 cents.<p>1) search in your needs that no one is addressing well enough currently.<p>2) search in some business needs that is already addressed but in an old way (and turn this into a web service where the company don't need to care about backups and broken window systems)<p>3) Niches are good but there are very different kind of niches. Some are good, some bad. For example my new product is about women and their menstrual flow, this is a niche "product" but not a niche "traget", and in the Italian version is working very well (we are launching in english, it's a matter of days).<p>Instead a new very powerful feed reader is a <i>wrong</i> niche. It's about very few skilled guys. It's very hard to make money and grow fast with this kind of stuff IMHO.<p>4) Provide a decent simple graphics from the start. We saw services that started to get users only after the graphical design stage was done. Not all the people are able to think as a simple clean site as good, like not all the people are able to buy fruit by taste and not by appearance. You need both...<p>5) Try to find a business model early. It's not needed to write a business plan or this kind of nonsensical numbers, but just, think "ok, eventually if/when I'll have a lot of users I can do some money selling this PRO feature, or attaching an ecommerce to this, or with this kind of service google indexing and advertising is going to work great".<p>6) Don't ruin your site just because of SEO. First user experience, than everything can help with Google not conflicting with user experience.
Or, you know, find a business guy as a partner. People from investment banking, consulting, venture capital, and private equity backgrounds have years of experience dealing with $50M - $10B companies and know markets (hey, look, the first thing VCs look at when making investment decisions) and their soft spots intimately. They also tend to have cash that can pay for an apartment and food. They're hybrid co-founders/angels.
The takeaway from the article is that there are many many ways to turn a simple problem to an idea and every small idea can be a viable business if you work on it hard enough.
The easiest way to find an idea is to target a big company, research it, figure out what are its potential next moves, then do it before they do it.<p>The fundamental advantage of a startup is that you can move faster than most big companies, and you need to use that to your advantage where it matters.<p>Here is a tip: if you can, try ideas that stay away from "advertising" as the main source of revenue, companies like that rarely become profitable quickly.
LISTEN!<p>People love to bitch about what bothers them. If you know people that work at a big company, listen to them bitch about their problems. Get them honest...or drunk...or both.<p>Then, figure out how to make those problems go away with software. I don't know the first thing about what goes on in a doctor's office but I'll bet if got the nurses talking I could find a product that would help them to do a better job.<p>If you can take that software and prove an ROI for the people that would use it, you've got a product.
The guy quoted in the article, who "wants to start u startup" has put it all wrong and is doing it for all the wrong reasons.<p>You don't get to "start a startup" -- startup is essentially a means and not an end; it is actually one possible form how a business could develop.<p>Generally, startups are formed from two motivations: a) you want to start a business (i.e. to work on your own) and b) you want to capitalize on a technology. In the former case there is a ton of possible ways to start a business, and some are listed in the linked article: you can become a consultant, you can set up a simple Web site etc. The same is true for the latter: instead of founding a company and looking for funding, you can sell the technology to another company, or simply license it etc.<p>I think that the quoted guy was attracted to the notion of a startup by the recent hype going on, and that he would like all the attention and "lifestyle" (or at least how he sees it), and is not really interested in building a valid business or technology.
Does it matter?<p>if you're smart one would hope you'd release early, get feedback and keep iterating till your original idea has evolved to something awesome.
Lame list of ideas. Here's what I think the basic skeleton is for all the most revered startups:<p>1) Find a product which a large company is selling and give that away for free, have lower operating margins, make revenue some other way.<p>2) Find a product around which a large company has created a man-hour intensive management process, commoditize the product by automating the moving-people-parts out of the process, standardizing the product.<p>3) Give away the razor, charge for the blades: Find a product that gives away something customers typically pay for, upsell a percentage of them on some higher margin product.