← Back to Journey Map Phase 06 · Validation & Experimentation
Phase 06 · Validation & Experimentation
MVP — Minimum Viable Product
The smallest thing you can build to learn something important
Frank Robinson (2001), popularised by Eric Ries · 2001/2011 ★ Must Know

An MVP is not the smallest product you can ship — it is the fastest way to learn whether your most important assumption is correct. It is defined by the learning goal, not the feature list.


Before full product development. Whenever you have a significant assumption about customer behaviour, willingness to pay, or product value that hasn't been validated.


  1. Identify your riskiest assumption — the one that, if wrong, makes everything else irrelevant
  2. Define the specific learning you need to validate or invalidate that assumption
  3. Design the minimum experiment that will produce that learning
  4. Build only what is needed to run the experiment
  5. Measure the specific metric that tells you if the assumption holds
  6. Decide: persevere, pivot, or stop — and document what you learned

🎵 Spotify

Assumption: users want a personalised weekly playlist curated by algorithm. MVP: manually curated playlists sent to 50 Spotify employees each Monday, tracking how many songs they saved. Result: employees saved 3x as many songs as from their own playlists. The learning justified building the algorithm — Discover Weekly was validated before a single line of recommendation code was written.

📊 Trade Surveillance

Please contact the author for more information on these examples at linkedin.com/in/kshitijrege



Frank Robinson (2001), popularised by Eric Ries 2001/2011


← Return to Product Journey Map