This also has the flavor of Warren Buffet's investing in 'boring businesses'. I think it is better to make something low-profile with higher likelihood of success. Bingo card creator is a good example. It takes out a lot of variables such as luck, connections, and timing.

There are plenty of places that you can choose your competition wisely. I would rather compete against a ma-and-pa than a fleet of the brightest engineers battling for space in the new fad market.

I guess it depends somewhat on if you are looking for f-you type money or something else. If I were to compete in a hard market, I would definitely go with the let the market move to me approach.

One other thing to consider about the big market hypothesis is: how accessible is it to a capital efficient startup. To take an extreme, the power generation market in the US is second only to health care, but it is quite unlikely that you can crack that as an entrepreneur. Similarly, even if you can get into a big market, the amount of margins available to subscale players is often a deterrent in and of itself (you can't for instance become a subscale ISP these days even though Internet access is a huge market as well). Outside of chasing a fad, going after inaccessible markets is the second biggest mistake I see.

DuckDuckGo is indeed useful. I just spent a while looking around the site and the only obvious monetization was selling logo clothing and a water bottle. Just curious: what is their plan for making money?

This is an interesting comment because it may determine if DuckDuckGo (DDG) can remain true to its privacy goals given that the privacy feature was not part of its goal initially compared to say 'offering structured content'. E.g., if privacy-conscious users started using DDG thus enabling DDG to reach critical mass attracting buyouts, huge advertising dollars for it's structured content, will DDG cave-in and ditch privacy?

It only came later because I wasn't educated on it. Now it's personal. I updated the end of the post to make this more clear. Thx.

Why couldnt you have highly targeted based advertising against the search terms and not the user?

I can see why it wouldn't be as attractive as knowing everything about the user and marketing towards them but I don't see why it couldn't work, and it isn't betraying the users trust so long as the data isn't matched to any IP or pattern.

You could! It could help DDG make lots of money.

However, my thought was if that (privacy) was not the main goal and something comes along, e.g., $$$$/partners, that could catapult you to become a market leader/giant in your core goal area then given the principle of most startups to focus on a core competency and to do it really well, the secondary goal (privacy) could be compromised.

EDIT: I don't have a list but I'm sure there have been cases where companies tweak their policies once they gain traction.

It has ads now, via http://www.carbonads.net/

I know of affiliate links but there may be others, as well.

If you're chasing a fad, you're already too late.

I love the point he made about search engine results not making too much inherent sense. It took a little bit of work, but I really tried to imagine what I'd like from a search engine, and it wasn't a list of textual links.

Absolutely. If I'm searching for journal articles, I'd love to be able to explore a network of citations. I want something to capture that experience of going to a library shelf organized by subject and browsing what's nearby. But it needs to scale to all works ever published.