Part of a series of responses to questions posed in the "have any questions for me" post.

Q: Lee

OK, here's a good question. Say you're a programmer with a particular idea - what are the most effective ways that you know of to research and evaluate the market before you jump in? Both qualitatively and quantitatively?

A:

Hmm, I'm trying to think back to everything I did getting ready for HelpSpot. First off I think it's different for each market. Some have lots of data all over the place whereas others are a bit harder to find. In many ways I think that can be your first clue. If it's very easy to find lots of good data on your prospective market then it may not be a great market for a small software shop. If that many people are spending time aggregating data about your market there's a reason and it probably means lots of strong players.

As far as evaluating I like to find high dollar value markets that have lots of competition, but which are highly fragmented. Lots of competitors allow you to use their marketing dollars to create general awareness of the need for your product. That's really expensive general advertising type stuff. Once they've created demand you can focus your limited budget and time on pulling in those leads they created.

Another thing I like to look for is who exactly the competition is. Personally I prefer to go up against relatively large companies and open source projects. What I don't want to see in a space is lots of small ISV's like me. The reason is that I rely pretty heavily on grassroots type marketing. The big guys have no idea how to do that and open source projects don't really market at all, though they do get press which feeds back into creating general knowledge about the market and creates more customers for me to capture. If the market has lots of smaller ISV's there's likely to be a lot more noise in the grassroots streams you'll want to use and it will make it harder for you to get noticed. For example, trying to create a hosted project management app (like Basecamp), a bug tracking application (like FogBugz), or a blog search engine (like Technorati) is going to be much more difficult because many of the grassroots channels are clogged up with talk of these other products already.

As a programmer one place I think you should never go to evaluate your product ideas is a programmers forum. Many programmers seem to think those are good places to ask about their product ideas, I completely disagree. Not for anything you're truly serious about anyway. The problem is that programmers like to shoot from the hip on those type of questions. So you've got this really well thought out product idea and then you post about it and 20 people come in and say it's junk in the first 5 minutes. I've seen many people get dissuaded by this, but the truth is those 20 people have no vision of your product, haven't thought about the market at all and generally are trying to stir up conversation as much as anything else.

Otherwise I don't have have much formal advice. I didn't do loads of computational analysis. It was much more driven by feel. Searching around, reading everything, and as I found areas that felt right that felt approachable I followed those paths. You definitely can't rush it. It takes months of research to really know the market and have a strong feel for where you fit in. I do suggest though that you go after an existing market that already has many millions or even small billions of dollars in sales. That gives you a nice big pond to work in. Trying to invent a new market or edge out a piece of a tiny one really lowers your chance of success IMHO.