Back in the early days of my career, it seems you could just pick an idea, do it, and then sell it.  Times are different now.  How do you navigate the entrepreneurial waters in this era?

Getting back to my first blog entry in this series, I certainly agree with my sales and marketing friends that disconnecting from the world for a long period of time while attempting to write a perfect product is dangerous.  On the other hand, I've had a lot of challenges trying to utilize the personalized, one-on-one interview technique that they suggest.  Some challenges I run into include:

1.       It's a process that doesn't fit well with my core personality.

2.       Identifying good people to interview is hard.  For me, interviewing software developers doesn't seem to be good enough.  I need people with expertise in ALM to get valuable feedback about the ideas I have.

3.       Despite the assurances of my sales and marketing friends, no common pattern of need has emerged as I've done these interviews over the past few years.

4.       It's an emotional rollercoaster.  If I talk to someone that gets what I'm talking about, I'm excited and ready to start cranking out code.  If they don't get it, I get discouraged and wonder if it's really a very good idea or not.  I suspect that I might have thrown in the towel on some good ideas because the handful of people I talked to about it didn't get it.

5.       You have to conduct a large number of interviews to get any statistically valid data.

6.       Most people don't give you actionable feedback.  Most people just say something very generic like "That sounds interesting."

So what do you do?  I have a process in mind that I think, based on observations of other successful startup companies, will work.  To describe the process, I'll start with an analogy.

Suppose you're an adventure tour operator specializing in shark petting.  You want your tourists to see as many sharks as possible.  So how do you go about bringing the sharks and the tourists together?

In one approach, you start with a detailed map of the ocean floor.  You conduct detailed data mining on every shark that's ever been caught by a fisherman.  You then dive down into the ocean with a remote controlled submarine to make contact with each individual shark.  When you find a shark, you tag them with a tracking device.  Then, using the world's most advanced radar system, you map the movement of each shark, feeding all of that data into another data mining program.  From here, you discover where the highest concentrations of sharks live.  You take your tourists to that location, dispatch each tourist in their very own submarine, and chase the shark down to get a glimpse of it.

Pretty complicated approach, right?  It's hard to do because the ocean is so big, and the sharks are too few.  Let's look at a simpler approach that's more doable.

Armed with some general knowledge of shark migratory patterns, you drop anchor in a location where you think you have a reasonable chance of attracting some sharks.  You then drop chum in the water to get the sharks to come to you.

And that's how I think you have to do things.  The location represents your product idea.  The sharks represent your potential customers.  The chum represents activities that you do that will get potential customers to come to your web site and take a look.

The key is to write code and chum at the same time.  If you only write code, then when the product is released you'll be dropping it into a world that is unprepared for it.  One of the main reasons for chumming isn't just that it provides research data, but that it builds buzz and momentum so that the market is primed and ready for the product.

How do you chum?  Try these things ...

·         Get the product web site up and running with software that supports groups and forums, so you can communicate with visitors.

·         Start an early access program.  Some information should be available to everyone, but  detailed product information and downloads should require registration.

·         Contact everyone on your Facebook friends list and your LinkedIn network to let them know about the product web site and early access program.  Invite people to join the early access program.

·         Write white papers and make them available on your own web site and other applicable web sites, if possible.

·         Start your search engine optimization efforts prior to product launch.

·         Start your own blog.  Don't just blog directly about the product, but also blog about tangentially-related topics.

·         Visit other people's blogs and post comments that link back to your web site.

·         Get other people to blog about the product.

·         For people registered with the early access program, publish early mockups and screencasts about the product.  I'm using Balsamiq Mockups for this (www.balsamiq.com).  Create a forum on the web site for people to provide feedback.

·         Start an open source project that provides functionality that is tangentially-related to the product you're working on.  If such open source projects already exist, become a contributor to those projects.

·         Join mailing lists or Internet groups that serve users with similar needs or desires to your product.  If none exist, create your own.

·         Use agile software development practices to get incremental beta releases in the hands of beta testers early.  Try to get something useful out fast.

All of these things are great, after you've settled on a product idea you want to pursue.  But that still leads back to the original question I posed, how do you decide upon the idea itself?  My advice is stop looking for a perfect idea.  It doesn't exist.  Pick a problem area that you've personally experienced.  If it's a problem for you, it's a problem for other people also.  Pick something you're passionate about.  Pick something unique.  Create.  Innovate.  If the product space you're going into already exists, then try to figure out at least two or three things that will make your product unique from all others.

Then, just do it.