Video: Toyota Uses “#LeanStartup” Methods to Develop a New Product


Since the “Lean” in the “Lean Startup” methodology basically comes from Toyota, it was fascinating to see two engineers from Toyota, Matt Kresse and Vinuth Rai, speak at the 2013 Lean Startup Conference. Matt and Vinuth (pictured below) are from the Toyota software development center in Silicon Valley, talked about their use of “Lean Startup” methods to prototype and iterate with a new Android-based in-car entertainment system.

toyota lean startup

It's really interesting to see how big companies, like Toyota and GE, are using “Lean Startup” approaches to create and refine new products – by better understanding their customer needs, creating a “minimum viable product” and then iterating based on real feedback that comes direct from customers.

I took some notes on their talk as part of my broader set of notes that I published and the video of their talk is now online:

My notes (expanded a bit for this post):

Matt Kresse @mattkresse

Vinuth Rai @vinuth_rai


  • They are part of a team in Mountain View building a “connected car”
  • Why is Toyota trying to learn Lean Startup?

    • Matt & Vinuth attended last year's conference
    • While principles of TLS are based on TPS, Toyota's software development was still a very traditional “waterfall” approach
  • Consumer expectations are changing quickly… but it takes about 3 years to get a new system out the door (safety and regulatory issues are part of the delay?)

    • They were not doing a good job of listening to customers (not getting feedback until the car is already on sale and out on the road)
  • Focused on the “customer development” process –> aiming for a better product

    • Their goal was to incorporate user feedback much earlier in the process
  • Matt read Eric's book, then Vinuth –> they said, “this is awesome” – joking: “We should do this and it will solve all of our problems, ha ha.”

    • Their idea — to experiment… let's run Lean Startup at a small scale and see what the gains are
  • Our “MVP” – looking at a “head unit” navigation system

    • Took an Android tablet wired into a car as an MVP
    • It didn't have the typical car features like AM/FM radio etc.
    • Wanted to learn more about providing mobile technology in a car
    • “We have never directly interacted with customers”  (Mark's comment: Isn't this amazing to hear, coming from  Toyota?)
    • Put out a Craigslist ad to get people to come in (300 of them, wanting to complain about their experiences)
    • Found 30 people and interviewed them
    • Chose 5 people to put the MVP into their existing car and trial was live

      • Those 5 were told you can use it for a month – then keep using it or we'll pay you $100 to take it back
      • 60% retention, 40% wanted to refer it to someone else
    • Big Question: How to connect this to the mainstream product development process?
    • They thought perhaps they'd build an MVP, establish a customer base, and the Toyota tells everyone to do lean startup? That didn't happen
  • Asked Toyota tech teams to benchmark their MVP-based product against others

    • Ran experiments together with test cars
    • Prioritized features in a matrix

      • Easy/hard and low/high impact
  • Looked at “innovation accounting” – will we stay on track? are we learning?

    • Often went back to bad habits and built what felt good to us
    • Worst – we stopped learning
    • Used google docs to track the experiments and it worked well, it's boring but it has to be done (See template shared by David Bland)
    • This process has forced us to be more creative – be creative in validation other than customers riding around with the tech in their cars

      • Cheaper ways to learn? Ideas from TLS community
      • Goal – most learning in shortest amount of time
  • TLS is not easy, but it's better because we get customer feedback before vehicles are on the road

I'm impressed (and not surprised) by the humility and I'm thankful Toyota would share this openly. It's risky to spend three years developing a product and THEN getting customer feedback (especially when it's hard to iterate once it's out in the field). It seems so simple to get customer feedback along the way (in real use, not just focus group type discussions).

What are your thoughts on this? Questions? I am trying to get Matt and Vinuth to join me for a podcast. What would you ask them?

How can your organization use Lean Startup methods and concepts to create new products or services, even in a big company?

What do you think? Please scroll down (or click) to post a comment. Or please share the post with your thoughts on LinkedIn – and follow me or connect with me there.

Did you like this post? Make sure you don't miss a post or podcast — Subscribe to get notified about posts via email daily or weekly.

Check out my latest book, The Mistakes That Make Us: Cultivating a Culture of Learning and Innovation:

Get New Posts Sent To You

Select list(s):
Previous articleGuest Post: What’s Needed in Developing a Kaizen Culture
Next articleStandardized Work? Why Are NFL Concussion Guidelines Not Always Followed?
Mark Graban
Mark Graban is an internationally-recognized consultant, author, and professional speaker, and podcaster with experience in healthcare, manufacturing, and startups. Mark's new book is The Mistakes That Make Us: Cultivating a Culture of Learning and Innovation. He is also the author of Measures of Success: React Less, Lead Better, Improve More, the Shingo Award-winning books Lean Hospitals and Healthcare Kaizen, and the anthology Practicing Lean. Mark is also a Senior Advisor to the technology company KaiNexus.


  1. I’m curious why “customer interviews” listed as the least learning method for validation? Isn’t this contradicting with customer development process?

    • Not necessarily. Systematically conducting open ended interviews around Problem Hypotheses and Solution Hypotheses can be effective. The issues with it is that most people are horrible at conducting them.

      Problem interviews, not solution monologues.


    • It seems like it’s far better to have an actual product (a prototype or MVP) in actual customer use. Interviews should probably only be a first step until you have something people can use (and better yet buy, voting with their wallets)?

      I’ve done customer interviews (this was taught at MIT back in 1997 even) and I always question the gap between what people SAY they will buy and what they will actually spend money on. That’s one reason I love the Customer Development / Lean Startup approach – trying to actually test hypotheses and not just talk about it.

      • I feel the conversation here shouldn’t be centered around “this or that” but instead “this and that” as we mix up qualitative and quantitative approaches for validating product/market fit and problem/solution fit.

        It’s ok if we disagree too :)



Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.