Finding the Right Startup Idea

I recently read a blog post about “unusual sources of startup ideas“. None of the sources seem unusual to me (I see many attempts to re-create built-in apps or copy recently acquired or shutdown startups) and I don’t think that copying others is a good path to success.

I feel like I see a lot of clones (another lo/mo/so sharing app please) because it’s easy to build a copy cat. You’ll leave the hard work (building customers) for later. As if that works!. To me it appears copycat is more usual and that proper customer development is more unusual.

In the book Zero to One Peter makes a very good case for building a new business in a monopoly (or oligopoly) space. That seems impossible if you copy a built-in or recently closed service.

What Worked for Me

I’ve tried maybe a dozen times to start something new; the two that worked are technology and workflow innovations in an existing problem space. Here are the steps I followed to find the product that customers wanted, using the most recent one as an example.

  1. Define the General Market Space you want to work in: Weed
  2. Define the Customers: Growers, Processors
  3. Contact these prospects and listen to stories for six months: IDEA SHOWS ITSELF! (or fails)
  4. Build MVP that solves core problems, demonstrate, collect feedback: IDEA REFINES ITSELF! (or fails)
  5. Validate by asking for money: VALIDATION (or failure)

#1 – The Marijuana business space was just opening up in Washington State, that was immediately a strong signal of opportunity. A new space, with few competitors.

#2 – Of all the problems I assumed existed I narrowed to ones I thought I could fix (technology, process automation) and who would have the more urgent need. Pick a game you could win if you try.

#3 – I went to meetups and listened, I went to prospects site to listen and watch. I wrote pages of notes from each visit. This is commonly called Customer Development, it works. Do It! It took about six months calendar time but man-hours invested was less than 200.

#4 – The first MVP was just HTML/CSS mock-ups of the application, which I showed around and took feedback on. From this I received a massive amount of additional insights into my customers desires. It’s critical at this point to shutup; listen and watch intently. Received strong signal that we were on the right path but without that the entire project would have been over. Repeat this step ensuring that your concepts match customer desires and incrementally improving the MVP.

#5 – Every stage ask about money. How much does current solution cost? How much is your time worth? How much time does current solution cost? Is feature X worth $120/mo to you? Is feature Y worth $240/mo? Will you pay me right now for software I’ll deliver in eight weeks? If you don’t talk to your customers about money, who will? Receive strong signal (read: money) and proceed else bail out.


It’s very difficult to innovate a copycat that is 10x better in a crowded space. Find a new, narrow problem, identify the group with money that is willing to pay for a solution, build relationships and become the big fish in a pond of your own making.