2
2 Comments

Talk with 150+ potential clients before building MVP

One of the first suggestion for indie hackers is to talk to your clients. Understand their pain points and then guage whether your idea will work. This is before you can build a MVP.

I decided to do it in parallel - (build + customer survey)

For customer survey, I had a mixed kind of response. Reached around 30% of my target of 150+ likely customers. And the response has not been positive! :)

The low rate is because of my laborious manner of reaching out to people. i have used the following channels:

  1. Personal connections from my previous job
  2. Linkedin connections, I reach out. These are relevant to the product that I plan to build
  3. Meetups and networking groups.

Are there any other channels that the community utilizes?

on December 10, 2021
  1. 2

    I don't know anything about your idea but there are quite a few assumptions here.

    • likely customers:

    How do you know? Since they're LinkedIn connections and the like, they're customers to whom you aspire to sell your products. Not likely.

    • the low rate is because of the laborious process

    Really? The low rate could be from a million things. 150 data points isn't enough to validate that assumption.

    I think it's worth rethinking if your customer interviews are getting you what you want. I have one customer I'm building an app for. I have 3 other potential customers who have issues with the same process. This is enough validation for my space. Build something very simple that will solve this issue, see if I can get a little traction. Go from there.

    Your post history says you're in ML so it's likely you need a little more validation because of the time complexity of your project. My intuition however, is to cut scope to something smaller and build the more complex parts in the future.

    1. 1

      Thanks for your perspective.

      Yes, on LinkedIn they are friends, colleagues (peope who I have worked with ) etc. They may not face the same problems.

      With ML, there is a big technical hurdle to cross - in tems of time + money. Till date I have got a negative review, regarding the idea.....so planning to drop it.