Agile LEGO – Toyota Kata an alternative to Retrospectives


My name is Håkan Forss and I’m a Lean/Agile Coach at Avega Group in Stockholm.

In this short story I want to introduce Toyota Kata as an alternative or as a complement to agile retrospectives.

But let me first introduce you to the team.


This is a tightly knit, cross-functional and very experienced team. They have been working with agile for a few years, mostly using Scrum.

The team consists of developers, testers, operations and business representatives.


The team has made many big process improvements over the years.

After great resistance from the office manager and human resources the team could last year be fully collocated.

They have adopted pair programming and test driven development.

But lately the team seems to have plateaued and is not improving any more.


At the end of every sprint the team does their agile retrospective

They reflect on what worked well, what did not work as well and what could improve.


They collect suggestions on improvements and add them to the already long and growing list of improvement suggestions.

They prioritize the improvement list and what improvements should be done in the next sprint.


Maybe it is time to try an alternative?


One such alternative may be Toyota Kata from the book with the same name.

The Toyota Kata book is written by Mike Rother and it describes two behavior patterns, or two Kata-s


The two Kata-s are:

  • The Improvement Kata
  • The Coaching Kata

In this story we will touch on the Improvement Kata.


The Improvement Kata guides us in a very focused way from our Current Condition towards our vision. The path goes through a number of intermediate Target Conditions in an iterative manner.


The Improvement Kata contains four steps

  1. Understand the Direction, so you know where you are going
  2. Grasp the Current Condition, to get your reference point
  3. Establish the Next Target Condition, that is your next step on the path towards the vision
  4. Plan-Do-Check-Act Toward the Target Condition


Having a clear vision, a compass heading, a true north is very important.

This is our guiding star that will make us all go in the same direction and not lose our focus.


As important as our vision, or maybe even more important is to grasp our Current Condition.

And I mean REALLY grasp our Current Condition.


To really understand you need to go to the place where the real work is really done. You need to talk to the people really doing the work.

You need to measure and collect data to grasp you Current Condition.


When we understand where we are and we have our long term vision it is time for the team coach and leader to set a challenging but not too challenging Target Condition.


A Target Condition describes the state of the process when the target is achieved. It does NOT describe the steps needed to get there.


When the Target Condition is set it is the teams turn to come up with experiments, Plan-Do-Check-Act cycles, that will take the process towards the Target Condition.


The team tries different experiments using Plan-Do-Check-Act cycles until they have reached or are close to reaching the Target Condition.

Then the Improvement Kata starts over and a new Target Condition gets set that is a little bit closer to the long term vision.

Now that the team has tried this for a few months, how are they doing?



With Toyota Kata the team can now be more focused on continuously improving their process. They practice the Improvement Kata every day and gets better and better at hands on problem solving. We have created a true learning organization.

Toyota Kata rocks!!


This story was presented as an ignite talk at Agila Sverige 2012 the 23rdof April.

The Toyota Kata and related material is developed by Mike Rother and his team. You can find more information here

A special thank you  goes to Andrea Chiou Adaptive Collaboration for help with my poor spelling.

70 thoughts on “Agile LEGO – Toyota Kata an alternative to Retrospectives

  1. Tobias Fors 2012-04-25 / 11:20

    Toyota Kata may rock, but your Lego storytelling style rocks even more! /Tobias

  2. Bill Costantino 2012-04-25 / 18:51

    Hello Hakan,
    I will be in Sweden May 2-16. During part of that time I will be leading a 3-day Toyota Kata public training event for PLAN, co-hosted at the Volvo plant in Gotenburg. During the latter part of the visit, I will be spending some time in Stockholm. I would enjoy the opportunity to get together to share ideas and experience applying Toyota Kata to different industries and situations. Please let me know if you have any interest.

    I agree with Tobias, GREAT JOB telling the TK story with the Legos. Mike Rother was particularly excited by the slide with the 3 TK Ninjas. Thanks for sharing.
    I look forward to hearing back from you soon.
    Best regards,
    Bill Costantino

    • Håkan Forss 2012-04-25 / 23:23

      Thank you, Bill for the nice comments. I would love to meet and discuss Toyota Kata in knowledge work.

      I’m happy to say that I will attend the 3-day Toyota Kata training at Volvo.


  3. Jacob Hamacher 2012-04-25 / 20:34


  4. Jeffrey Uitenbroek 2012-04-25 / 23:28

    I really like this.

    Who is the person(clown)outside the window on slide 17?!?!

    • Håkan Forss 2012-04-25 / 23:34

      Who knows? I don’t know how he got there🙂


    • inventionaddict 2012-06-27 / 15:00

      This is awesome. I intrepreted the clown as someone who does not get process improvement and is on the outside looking in. I’ve run into people who are dead set against change and try to derail the teams efforts. Well done!

  5. Marcus Hammarberg 2012-05-01 / 11:23

    Loved this talk. Loved this blog post.
    This is so well thought out and well taught out🙂

  6. Mike Wroblewski 2012-05-01 / 23:37

    I love your lego presentation on Toyota Kata! Very well made indeed with a nice storyboard style.

  7. PM Hut 2012-05-02 / 17:21

    Hi Hakan,

    I wonder how much time did it take you to actually do this post and create all these legos!

    I’d say 2 days, even more!

    • Håkan Forss 2012-05-02 / 22:43

      All in all I think I have invested something like 4 days. This include developing the manuscript for talk the post is based on, getting hold of the LEGO mini figures and some of the other LEGO pieces, building the set(including the handmade mini post-it’s), taking the photos and then all the practice runs of the talk. Timing the manuscript to the 15 sec automatic advance was hard but fun.

      / Håkan

  8. Petra 2012-05-08 / 21:01

    Vilken presentation, vilken kreativitet! Ser fram emot den 14:e.

  9. Frode Odegard (@Odegard) 2012-05-24 / 10:14

    I love this!!! We’ve been teaching this in the sw industry for years, but never illustrated so beautifully. My hat’s off to you.

    Frode Odegard

  10. Timo Mulder 2012-06-01 / 15:50

    Looks cool with the LEGO and stuff but Toyota KATA is similar as in A3 thinking. That’s not new…😉

    Doing Agile retrospectives with LEGO serious play is something that we do at our company and with our clients. Will shortly post a blog about it.

    Cheers for the blogpost though!


    Co-founder & Agile/ Lean Coach at ValueFirst

    • Håkan Forss 2012-06-01 / 16:30

      Hi Timo,
      Yes Toyota Kata and A3 thinking are in many aspects similar. It is not something new as it has been used at Toyota for decades🙂

      Looking forward to read your blog on how you use LEGO Serious Play in your retrospectives. I’m planning to learn more about LEGO Serious Play and start using it in my client engagements.

      / Håkan

  11. Arthur Richard 2012-06-27 / 08:11

    Hello Håkan, extraordinarily effective way of communicating a practical technique. Thank you for sharing with us!

    Arhut Richard

  12. Pixy 2012-06-27 / 08:42


  13. TomJ 2012-06-27 / 11:08

    Absolutely bloody brilliant🙂

  14. Susan Mortarelli 2012-06-27 / 11:55

    Truly Remarkable! Thanks for a whole new approach to teaching and communicating simple concepts.

  15. Ed Francisco 2012-06-27 / 14:59

    Excellent presentation of process improvement in a simple approach.

  16. Cat Lewis-Tubre 2012-06-27 / 17:28

    Just fabulous! Wonderful innovative teaching technique!

  17. Jennie G 2012-06-27 / 17:49

    I just love it … Thanks so much for doing this and putting on a blog to share!

  18. Connie Tolman 2012-06-27 / 19:06

    Could you provide more informaiton on how you collected data on the “current condition” phase. I’ve had trouble with that when it isn’t a mfg. process that has yield data and throughput times, etc.

    • Håkan Forss 2012-06-27 / 22:04

      Hi Connie,
      Understanding your current condition is important. To create understanding you need by collecting process data and facts not gut feelings or opinions. The basic facts I usually try to collect are:
      • Lead time. Often grouped based on work type or bucket sizes
      • Lead time variation. Also this grouped as above
      • Some type of quality metric, like number of bugs escaped to production
      • Throughput
      • How work flow through the process often using a CFD
      Based on what’s your current challenge or your next target condition you may want to collect appropriate data that give you insight if you are improving or not. This will be very context dependent.

      Most of the time it is not important to get the collection of data to be exactly 100% correct. It is much more important to just do your Improvement Katas. As you get good at doing the Improvement Kata you can fine-tune the collection of data.
      / Håkan

      • Connie Tolman 2012-06-27 / 23:03

        You can get caught up with collecting too much data, I guess, but how long do you spend collecting data in relation to the other steps? My problem is that the processes are not continuous and therefore, the data is extremely variable. I guess I should just start the Improvement Kata, like you said.

      • Håkan Forss 2012-06-27 / 23:19

        The length of the PDCA cycles, according to Mike Rother and Bill Costantino, are in hours not days in manufacturing. This is how quick the iterations are.

        I collect basic process data every day in a continuous fashion so usually there is no long data collection phase. In one case I spent the better part of two weeks collecting detailed data for one process step but that was a big exception.

        As Mary and Tom Poppendieck says, “Result is not the point”

        It is much more important to get good at improving and leaning than get precise data how much you improve.

        / Håkan

  19. Håkan Forss 2012-06-27 / 21:35

    Thank you all for your kind comments. I’m working on a longer story so stay tuned.

    / Håkan

  20. Butcher 2012-06-27 / 23:01

    Love the legos. This is going to basics; The more simple way we explain things the more clear way to undesrtand the concept.

  21. Jo 2012-07-09 / 18:48

    I think the team members need to be careful to not drink too much! Ha! Seriously I loved the presentation method and look forward to your next one.

  22. olaflewitz 2012-07-30 / 15:52

    Great post, great presentation, awesome pictures:-)
    Two questions:
    – why doesn’t the team define the target condition? What qualifies the leader to set that for them? Or is leader just the “in the moment” function of any person suggesting that target?
    – what is the context of the Poppendieck quote you give in the comments: “result is not the point”? Because I think (but that might just be viewed from a different angle) there’s no other point besides results. We should focus on improving results: the outcome and impact of what we do.
    Thank you for the inspiration!

    • Håkan Forss 2012-07-30 / 20:21

      Thank you Olof!

      Two great questions. Let me try to answer them as best I can.
      – What is the role of a leader? In my opinion the role of a leader is to know the direction and the capability of the team. I also think it is every leaders responsibility grow the teams capabilities by coach them in this growth. The leader’s role is to set a target condition that is just beyond the team’s current capability to challenge them to learn and grow. It is the team member’s role to come up with all the experiments that potentially brings them closer to the target condition in short cycles.

      – Let me start by this Chinese Proverb “If you want 1 year of prosperity, grow grain. If you want 10 years of prosperity, grow trees. If you want 100 years of prosperity, grow people.” At Toyota and other mature Lean companies the actual results is not the main point as a leader. The main purpose of the leader is to grow the people. If you grow your people to every day strive for perfection and continuously improve you will get results. Alas – result is not the point for a Lean leader.

      / Håkan

      • olaflewitz 2012-07-31 / 15:51

        Interesting points, thanks for the clarification.
        I agree on what is being done (re leadership) I disagree with the “job” of a leader assigned to a specific person. This assumes an organisational structure that IMO is not appropriate anymore in this century, and only adds value within a dysfunctional context (that of a hierarchical organisation). With continual improvement, such an organisation will reach a limit on its effectiveness if folks do not start questioning this structure.
        I think Bob Marshall is right with his questioning of leadership in general, and agree we should aim for replacing it with fellowship.

        Re Results: I think we’re talking two sides of the same coin here. Surely we need to grow people (everything else would be more than short-sighted for a knowledge work organisation) but we need to grow them so that they can focus on results. Growing the people is a means, not an end (it might be to the individual, yet not for the organisation). Only when the individuals’ growth is aligned with the organisation’s goals the organisation and the individuals can all get what they really want. Which means, given a hierarchical organisation, the leader should care for the people so that the people can care for the results.

        To summarise:
        Results are all what I care about, and I believe people will love their work most (and in a sense, grow best) if they can fully focus on results, be fully engaged in teams which achieve awesome results.
        I do not care about leaders, and I only nurture leadership as long as we don’t have a better alternative—and I see that coming.

      • Håkan Forss 2012-07-31 / 21:23

        I think it is a novel idea of organizations without leadership assigned to specific persons. I have not seen or read of this to working in any organization of size. I will put myself in the pragmatist camp on this topic. I will promote what is known to work. Leaders that lead through coaching at the gemba are what I currently promote.
        And now to the question of results.

        I will start with a quote from John Seddon’s book Systems Thinking in the Public Sector

        Unfortunately and counter-intuitively, focusing on cost drives costs up. Managers need to learn to focus instead on value.

        By focusing on the actual results you will not achieve the full potential of your organization in the long term. You need to focus your attention on perfecting your capabilities that generates value for your customers and let the people grow. Growing people is not the means. Growing people should be the goal of every leader in your organization, not the result. As you work your organization towards perfection and individual mastery you will get results.

        / Håkan

      • Andrew Bishop 2012-12-19 / 19:19

        Just another thought on results: Rother’s analysis of the final step is focused on LEARNING not results, as I’ve heard him present it. “When should I check back to see what you have learned”, not “…what you have accomplished.” It is an important distinction, since sometimes we try something that takes us in the wrong direction. As long as we learn (and the PDCA cycle is appropriately small) improvement will continue. This echoes many stories I’ve heard from Toyota: “No, you haven’t failed – look how much you’ve learned.”

  23. Panama 2012-09-20 / 17:13

    The Improvement Kata talks about a daily cycle of looking at the current actual condition, in light of the current target condition, understanding the obstacles explaining the gaps between the actual and the target, and urging us to choose one of the obstacles and work to address it in small experimentation steps using the PDCA cycle (Plan Do Check Act). On top of this approach sits the Coaching Kata with Five Questions that are aimed at coaching people on using the Improvement Kata. The aim is for managers to coach their people in their improvement work.

    • Håkan Forss 2012-09-20 / 20:06

      Yes that is correct. In my experience the PDCA cycles of the Improvement Kata has to be put in relation to the cycle time of the process you are applying it to. If the cycle time is in minute’s, daily Improvement Kata PDCA cycles are appropriate. When the cycle time is in a day or more, a weekly or biweekly Improvement Kata PDCA cycles are more appropriate.

      What do you think?

      / Håkan

  24. Diana Larsen 2012-11-05 / 22:16

    Hi Håkan,

    A very interesting process and, as you say, a compliment and additional activity to add to one’s Agile Retrospectives toolkit. (Which, by the way, is much more than listing WWW/DD or putting sticky notes under 3 emoticons despite popular misconception.) Actually it tracks quite well with the AR framework – Set Stage/Understand Direction; Gather Data & Generate Insights/Grasp Current Condition; Decide What to Do/Establish Next Target Condition; and Close the Retrospective. Repeat improvement steps iteratively/PDCA Experiments. It’s also similar to the Adaptive Action Cycle of Human Systems Dynamics.

    These are all great ideas that show self-similarity, and all parts of a larger improvement pattern that works well in complex adaptive systems. Toyota Kata adds to the breadth of iterative continuous improvement techniques.

    Thank you for this clear & entertaining description!

    Diana Larsen

    • Håkan Forss 2012-11-06 / 00:15

      Hi Diana,
      I agree that an Agile Retrospective is very similar to Toyota Kata when done as you describe in the book.

      What I have found when using Toyota Kata is a more problem solving and focused mindset. Given a challenge the team gets focused on solving the obstacles stopping them from meeting that challenge. The emphasis on gathering real process data and facts over gut feel and opinions also brings a high focus. The goal of Toyota Kata is to build a continuous improvement capability of the organization. If possible the Improvement Kata and the Coaching Kata should be run every day all the time.

      Thank you for your comment and for your great books. I bought my second copy of Agile Retrospectives just a few weeks ago for my Kindle.

      I have posted a more detailed presentation on how Toyota Kata can be applied in software development. Take a look at my Kanban Kata presentation here.

      / Håkan Forss

  25. Ellen Gottesdiener (@ellengott) 2012-12-28 / 19:47

    Thank you Håkan for sharing this fabulous, creative presentation and the links to the Toyota Kata.

    It is a deepening of, and orthogonal with, a number methods, tools and practices – some of which can be integrated into the improvement kata.

    I see the improvement kata as adding depth to how i have seen many lean and agile teams using retrospectives. (Although Diana and Esther’s wonderful Agile Retrospectives book provides the framework for the ongoing, deeper improvements, most teams not yet using them systemically).

    Among the practices i see related to this improvement kata are: Appreciative Inquiry, Complex Adaptive Systems/human systems dynamics as mentioned by Diana above, Goal-Question-Metric, lean startup, Tom Gilb’s value management (Planguage for quantifiable value), Quality Improvement Paradigm (for software), Solution Focus, various change and learning models (e.g., Kolb, Dreyfus, Satir, Kotter), Technology of Participation (ToP), Skilled Facilitator approach (ala Roger Schwarz), and Kanban implementation practices as guided by David Anderson.

    Well…that’s all i can think of at top of mind in this moment😉

    In my experiences thus far: 1) practices need to be adapted to context and culture and 2) we benefit by combining both qualitative and quantitative practices which integrates and honoring both our mind and heart.

    You have shared a valuable method for deep sustaining change. Thank you.

    warm regards,
    ~ ellen

    • Håkan Forss 2013-01-09 / 17:54

      Hi Ellen,
      I’m glad you liked it.

      I agree that Toyota Kata is both a deepening of, and orthogonal with, a number methods, tools and practices. Some can be integrated very well and some will replace one or two of the main Katas.

      The Coaching Kata, that I did not touch on in this post, is a very important part of Toyota Kata.

      I highly recommend Mike Rothers book. More or less all examples are from manufacturing but the method really applies to any industry.

      / Håkan Forss

  26. Matt Bryan 2013-01-05 / 11:27

    Hello Mr Forss,
    We are just getting started with the kata. We are working in a hardware development area using lean product development. Have you tried the kata in hardware development? I am trying to think of what the differences might be. Any ideas or pointers you could give me?

    On what to watch out for……

    • Håkan Forss 2013-01-05 / 14:08

      Hi Mr Bryan,
      No, I have not used it specialty for hardware development. But I don’t see why it would not work. It is a tool for continuos process improvements and a process learning. Just like in software development the longer cycle times may pose a problem. I currently recommend at least one new experiment a week and at least one experiment should be running at any time.

      The important part is to get started. Think of using Toyota Kata as an experiment for continuous improvements. You will learn as you go and you will need to adapt.

      Good luck, Hakan Forss

  27. Corinna 2013-05-06 / 12:03

    Great storytelling. I am very impressed by the simplicity of the idea. Others spend thousands on videos, etc.

  28. Franco Martinig 2013-06-19 / 14:06

    I am impressed by this post where good content meet nice form. I have seen that you do presentation in conferences about Toyota Kata. Would you be interested in writing an article about this topic for the Methods & Tools magazine? If yes, just contact me to discuss the details.

  29. Pingback: Agile LEGO
    • Håkan Forss 2015-07-09 / 14:44

      Hi Peter,
      Thanks for reading and pointing me to this article.

      The Toyota Kata method by no means guarantee high quality by itself. This management practice is about creating a learning organization that can learn from it’s mistake and improve over time. The more interesting question for me based on the article you refer to is if Toyota has learned from it’s mistakes and are now developing safer and higher quality software.

      • peterwilliams97 2015-07-10 / 07:20

        Dear Håkan,

        Most of what I know I about Toyota Katas is what I have read in your blog post and on Wikipedia so I may have misunderstood them.

        I, apparently incorrectly, thought the point of Katas was do them regularly over a long time to gradually improve your work practices and eventually make something good.

        In Toyota’s case, I expected years of Katas to eventually lead to good software. However Toyota ended up with bad software.

        I didn’t think the point of Katas was that it was okay to end up making something bad after years of work, as long as you learned from the experience of making something bad.

        However, I am new to Katas, so I probably misunderstood.

      • Håkan Forss 2015-07-10 / 08:13

        Dear Peter,
        I don’t believe any management system, including Toyota Kata, or quality improvement system can prevent all mistakes. Your argumentation is guilt by association to me.
        You don’t need to trust my word, or Mike Rother’s research, if Toyota Kata works or not. Try it!
        Apply the scientific thinking and try if it works, or not, in your context.

        Form your hypothesis.
        Make your prediction what will happen if you try Toyota Kata for some time.
        Run the time limited experiment.
        Closely observe what happens.
        Analyze the delta between your predication and the actual outcome and see what you can learn.

        I would love to hear your results.

  30. peterwilliams97 2015-07-12 / 07:26

    That sounds like a good idea,.

    I remember this quote from Sherlock Holmes.

    “It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.”

  31. Michelle 2016-10-04 / 20:08

    What a great way to present a topic that is not complex, but difficult to grasp.

    I, also, want to know more about the sneaky clown.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s