Concept
Learning Milestones
Learning milestones are the three checkpoints of innovation accounting (establish an honest baseline with an MVP, tune the engine toward the model in your plan, then decide whether to pivot or persevere) that let a startup prove real progress before revenue can.
Early on, revenue is too small for the usual financial reports to say anything, so it is easy to mistake activity for progress. Eric Ries's answer in The Lean Startup is innovation accounting, which moves through three learning milestones. First, you use a minimum viable product to measure where the business really stands today. That means numbers like conversion, retention and referral, even if they are embarrassing. An honest baseline is the whole point. Second, you make only the changes aimed at moving those numbers toward the model in your plan. You check whether each change actually moved them. Third, you look at the trend. If tuning is bringing the numbers closer to the plan, persevere. If it isn't, it's time to pivot. The milestones turn "we're working hard" into a claim you can test.
A real-life example
Leila launches a $15-a-month tutoring-match app with a bare-bones MVP. Her baseline: 4% of visitors sign up, and 20% of those are still paying after two months. Her plan needs 40% retention. For six weeks she makes only changes aimed at retention, such as a first-session guarantee and a progress email, and she measures each one. Retention climbs to 24%, then stalls. The stall tells her it's time to consider a pivot. Her feelings don't get a vote.
How to use it
- 1Before changing anything, run your MVP and record your few key numbers exactly as they are. This is your baseline.
- 2Write down the target each number must hit for your plan to work. Then make only the changes aimed at closing that gap, and measure each one separately.
- 3After several rounds, compare the trend with the plan. Steady movement toward it means persevere. Flat lines despite real effort mean pivot.