Evaluating Startup Ideas

Noah Barr
Noah Barr
Jun 19, 2015 · 3 min read

10 steps

Last February, I decided to take a swing at starting my own business. Living in San Francisco, of course, it was going to be a technology startup.

For the last couple years, I had been keeping a running list of business ideas in a Google Doc. I turned there to get started.

Business ideas can come from anywhere. Mine usually come from gaps I’ve noticed while trying to do my job or live my life.

Here are the 10 steps I took to evaluating these ideas:

  1. Take the first/next idea off the list. Block every other idea out for the time being.
  2. Initial diligence (3 days max): Do some legwork around target customers, quick market sizing, competitive landscape, what the product really is and how you are going to make money. Make a list of questions / concerns that need to be resolved. Commit this to a single document (<3 pages).
  3. Marinate. Sleep on it, take a long hike, or take my grandfather’s approach (aka, “Consult your board”… by going into the bathroom, take a leisurely #2, and re-emerge with a decision made).
  4. Involved diligence (<2 weeks): Find, schedule, and complete 3 target customer interviews. Based on this, revisit and revise your work from Step 2.
  5. Wireframing (1 day). Get shoulder-to-shoulder with a local product/UX person. Iterate together with pencil and paper. Walk away from this day with something that captures the essence of your product. Negotiate a daily rate, not an hourly rate. Be a mensch and pay for lunch and coffees.
  6. Slap some lipstick on the pig (1 day contracting, 2 days processing). Enlist the incredible and shockingly inexpensive skills of far-flung contractors. Upwork: Convert wireframes into high fidelity mocks and convert a simple, handwritten customer pitch deck into something decently presentable, and (b) Fiverr: Get a cheap logo made.
  7. Pricing (1 day, while Step 6 is happening). The last thing you want is to successfully pitch a prospective customer in the next step, and get caught with your pants down when they ask about pricing. Don't let yourself spend forever on this. Whatever you decide, it won't be right, so just pick something.
  8. Look semi legit (2 hours, while Step 6 is happening). Buy a placeholder domain, redirect the website to Launchrock, and get Google Apps setup. This builds a shocking (and perhaps unearned) amount of credibility with prospective customers, especially those less technical.
  9. Sanity check (1 day). Lean on 3 friends who you trust to beat up on the idea, avoiding confirmation bias and general delusion. Ideally, 1 person in venture capital, 1 startup operator in an adjacent space, and 1 smart engineer.
  10. Re-pitch (1 week). Pitch product offering to Step 4 folks or find some fresh meat. Don't move off this step until you get 2 or 3 yes’s. By “yes”, I mean something in writing from the customer saying “I am committing to paying $X for what you just showed me”. An email to that effect is fine too! Anything else is bubkis.

18 seemingly awesome ideas, 0 survivors.

Step 4 and 10 were the predictable killers. Step 5 was the sleeper assassin. With all my ideas exhausted, it was time to take a full-time job.

Next up: How idea 19 came to me while at 42Floors, and life after Step 10!


Looking to be a technical cofounder? Know someone who might be? Idea 19 is called MoveQB (website) and progressed way beyond Step 10! Reach me at first.last@gmail.com

Noah Barr

Written by

Noah Barr

Healthy disdain for socks. Dog Lover. @TreasureData, previously @42Floors @GetBackOps @Helpshift @Crittercism