Launch Effectiveness : Lab = must have a testable hypothesis

In modern Product development we use every discovery technique at our disposal. One of my personal favorites is similar to a live-data prototype, but in production… I call this Labs.

Putting an offering into Labs means you are wanting to deliberately learn a thing. Therefore a hypothesis is required. Just putting it out there to see what happens is ineffective at best, lazy most likely, and a vehicle for confirmation bias at worse.

DON’T DO IT.

Do this instead. BEFORE putting it in the Lab, set Launch Effectiveness. But, with a decision tree.

If after [some testing period]:

• If X, then we invest in the needs related to graduating this offering from Labs to General Availability
• If Y, then we abandon and turn it off
• If Z, then success is inconclusive and we will lean on the qualitative feedback that we received and our product sense to guide our next steps. Likely iterations and re-testing.

Note: Testing Period = could be time or quantity based (this number of folks try it), but must have a limit. I prefer quantity based… # of ## folks who try show signs of value attained is focused. If you include a time limit you are testing adoption, and as we learn from Constellation Metric’s Three Categories … adoption is different from value.