Beginning Marketing Experiments

Posted on Thu 23 July 2020 in experiments

Marketing has been something that I have long avoided. Always focusing on product development instead. I went against the advice around the internet - I did not validate the idea of Conveyor.dev before I built it. I was my first customer, Conveyor initially was built to solve a problem that I ran into every time I wanted to deploy a Python application.

Currently, Conveyor only has two paying customers - in order to start moving forward, I am going to have to start showing Conveyor to more people. It is a clear fact that I know nothing about marketing or getting customers.

Do I hire this out or try it myself?

In order to help myself learn I have decided to start a series of marketing experiments. The plan is to use an old approach... The Scientific Method.

The idea is simple, ask a question, develop a hypothesis, run an experiment, validate or invalidate the hypothesis, and then try to extract insights from the data analysis to use in future experiments.

Here is the general criteria of the experiments to start with:

  • Small experiments lasting 3-5 days
  • $10 max ad spend per day on each ad set
  • Experiment with interests/audience a lot initially to start finding the type of people this product speaks to
  • Keep each ad set to a single specific interest