Analyzing the results of your leanstartup experiment!
This is the best part...! You have picked a risky assumption, converted it to a sound testable hypothesis and run an experiment. You have run the experiment and reported on your results. How best to truly interpret these results. If we were going to try and turn this into a formula - this is what we would do - we would go directly to the prediction vs results.
DID THE PREDICTION COME TRUE? YES OR NO
IS IT ACCURATE?
HOW FAR OFF WAS THE PREDICTION
DO WE KNOW WHY
WHAT DO WE DO NEXT
- Did not meet criteria... false negative
- Did not meet criteria... true negative
- Met criteria... false positive
- Met criteria... true positive
- failed numerator
- failed denominator
- tested at the wrong time?
- right people didn't BELIEVE the solution was possible
- tested on the wrong people (this is a true negative for the segment you tested on though)
- couldn't find enough people (this is an indication though)
- subjects flat out didn't understand the experiment
- High denominator with custdev - extremely low numerator... explicit learnings are "NO PROBLEM, HAPPY WITH SOLUTION
- LOW numerator after 3 tries or 6 months
- Asking friends and family - not the customer
- Not testing the riskiest assumption
- Asking other segments who have a vested interest in your startup succeeding (service providers, etc)
- The red herring experiment... "Free beer if you sign up for our Petfood delivery service" Entreprenru interpretation - people want the Petfood delivery service... Everyone else "people want free beer"
- WONDERFUL... despite all of our feedback... we are actually rooting for this! Can you repeat it? If you got 10 can you get 100 in the same amount of time? Can you automate it? Can