I’ve got a great idea for a startup, now what?

November 9, 2011 · 14 Comments

On Oct 20th, I was invited to give a talk at NYU Stern as part of the Entrepreneurship Speaker Series. Below is a video of the talk, key takeaways and my slides. Enjoy.


Key Takeaways

  • Investment and tech talent is earned not given.  What can you do to further your idea on your own?  How can you validate your idea and demonstrate traction without spending a huge amount of time and effort building product?
  • Pick an idea in an area where you have strong domain experience, know the problems and have the connections to sell.  Look for markets that are big, have growing user bases and have rapidly changing technologies/distribution channels.  Love the problem space.
  • An idea is an assumption.  Turn that idea into a series of questions and figure out what’s the easiest way to find out an answer.  (examples: Skillshare, TheLadders, Zappos, Groupon, Seamless, Dropbox, SinglePlatform).
  • Fake it till you make it, delay complexity (Ghetto testing examples: Aardvark, Zynga)
  • Defining a minimum viable product – What are your biggest questions/assumptions about the product?  Identify those and devise and execute tests.  Will people use your product (and come back)?  Will they tell their friends?  Will they pay for it?
  • Launch Early
  • Don’t do a big launch.  Find a few people who feel the problem you are trying to solve most acutely.  Get them using your product in the way you want them to.  Then expand your audience.  Communicate with them openly and honestly.
  • Personal development: Everyone who is aspiring to be a tech entrepreneur should be: developing a strong online presence, comfortable with html/css/js/hosting/dns, understand online marketing (apply for Google Grants and manage SEM campaign for an NGO), understand web analytics (read Web Analytics 2.0), read The Lean Startup, attend OHours, Meetups, Skillshares.


  • http://www.facebook.com/ChaitanyaPramod Chaitanya Pramod

    It was an awesome talk! loved it. Especially the evaluating the idea and launch before polishing parts.

    • http://amitklein.com amitklein

      thank you!

  • Pingback: Chaitanya Pramod blogoscoped

  • Startupjerkfest

    mi notez:

    Identifying Good Markets to Enter:

    You know the industry really well, you understand your customer, therefore you can build the correct solution to the real problem, and be able to sell it to them

    Size of Market: growing, not shrinking

    Change in Infrastructure or distribution, i.e. something has disrupted the normal way it used to work

    Passion – a drive & motivation to keep going when you absolutely want to quit

    Goal is to take any idea you have, and turn it into an assumption that you can test, to see if your assumption was true

    research the problem (before you propose a solution). understand the REAL problem, not just what you THINK the problem is. ask why it’s a bad idea, you’ll get better answers that way

    validate your ideas:
    design the experiment.
    set goals
    what do you want to measure
    how will you measure it
    run the experiment.
    analyze the results
    did the results confirm your hypothesis?
    do you have new questions for the next hypothesis?

    if you can show that people are interested and demand your product, then you can build a minimum version

    you want to DELAY complexity, don’t build any technology yet, until you can’t do it using manual techniques anymore. i.e. make your idea work using manually techniques. i.e. using manually written emails to do what your application would do if it was built. do this to test demand before you build it.

    you have to think up 10 ideas of how you will get customers, and do this ahead of time before you start. because the first 3 ideas you have won’t work, and you’ll be tempted to give up if you don’t have 7 more ideas to try.

    by collecting email addresses and building interest for your product/service before it is build, you will be able to survey and then promote your product launch to people who are interested in it, which is better than building something and then start cold calling strangers, trying to get them interested in it.

    Minimum Viable Product

    will people use it & come back
    will they tell others about it
    will they pay for it

    identify your biggest risk of why your idea won’t work.
    then develop a way you can test to see if that risk is real or imagined
    i.e. if the risk is people will not pay money to see their name on a webpage, the you make a test to see if they will fill in a form with a credit card number to buy the service.

    this will not damage your reputation if you do the testing correctly – find early customers who have the pain you are trying to solve, and they have a lot of pain of it (i.e. this is an important solution to them), and work closely with them, be open, honest, make them feel that they can help build the solution by trying your idea and giving feedback. they will realize it is a testing environment, and they won’t freak out. also – have courage, you have to take a risk to get a reward.

    celebrate after you accomplish a test, you’ll need to take the edge off so you can stay alert and fresh, so you don’t burn out

    personal development items:

    start blogging about your experiences, mostly the bad. tweet them and push them to startup communities. get people to read them. doing the blogging will strengthen your ability to communicate which helps in customer development and sales. having other people read them will get you exposure to people who want to help you. that is why you should write about things you are having trouble with. when people see trouble, they want to help.

    • http://httpcolonslashslashwww.www.startupjerkfest.com StartUpJerkFest

      what is the link to this comment? view source doesn’t show it. i want to link to it for reference

      • http://amitklein.com amitklein


        thanks again for the notes!

        • http://httpcolonslashslashwww.www.startupjerkfest.com StartUpJerkFest

          thanks, now i can bookmark this and use it as a basis for some presentations i’m planning for Buffalo NY

  • Startupjerkfest

    i could tell in the video how you were getting frustrated with the lack of response and feedback from the audience, which looked like a very small audience.
    next time you should have a drinking contest before the lecture, to get people to loosen up, lol.

    • http://amitklein.com amitklein

      there were actually close to 80 people there which is why it was a bit frustrating… with that many people I expected folks to talk, I’ve given this talk one or two other times and the crowd has been better, I agree mandatory jager shots wouldve set the mood

      • Derek

        We are much more lively down here at Penn

        • http://amitklein.com amitklein

          Invite me and get a room full of students and I’ll come

  • Mike G

    This was a great talk, some good logical common sense ideas, and some great insights into some other successful start-ups.

    Amit – what are your thoughts on the 37signals approach to building a profitable business from day one, building true scale, not taking VC money, etc? Seems like many of your ideas intersect with theirs, though maybe not to the same extent?

    • http://amitklein.com amitklein

      Thanks Mike. I would generally advocate building in a revenue model initially thought for some businesses that doesn’t make a lot of sense. If there is a strong reason (and you have the runway) to achieve scale before monetizing, that can be a powerful approach (instagram, pinterest, twitter, etc…).

      I would always suggest going as far as you can on your own before considering taking on VC money (though in many cases strategic angels – insiders in the industry – can really help you if they have skin in the game).

      I like the 37signals guys a lot and enjoyed reading Rework. We are heavily influenced by Steve Blank, Eric Reis, etc…

  • Stephen L

    Extremely insightful! Thanks for posting the video/takeaways/slides.

    (and great intro/outro music too..)