When googling Hoshin Kanri, you will sooner or later come across an X-Matrix. It is a visually very impressive tool, but I am in serious doubt about its usefulness. It focuses on the creation of the Hoshin items, but to me this approach is overkill, and – even worse – may distract the user from actually following the PDCA, especially the Check and Act parts. While the article is highly critical, I hope reading it and understanding the shortcomings help you better understand how Toyota thinks.
Setting the right goals and filtering them through the organization is important in Hoshin Kanri. In my first post I talked in detail about this as the “to-do list.”
These simple to-do lists can be modified to be quite elaborate, eventually leading to the X-Matrix. They are also sometimes called x-matrices (instead of matrix) or target-mean matrices. The origins of this matrix is a bit fuzzy, but it seems that Japanese professor Yoji Akao (1928–2016) and Bob King (owner of a quality lean-related publishing house) were involved.
In any case, looking at the matrix, I have the strong feeling that this was not developed by a practitioner in the field but more likely by an academic or consultant who has the need to impress others with fancy methods. But before I go into details of the criticism, let me briefly show how the matrix works.
The Fields in the Matrix
The X-Matrix has a number of fields. To fill it out, you start at the bottom (usually called “South”) with the long-term objectives. Next comes the annual objective (left, or “West”), then the top-level priorities (top, or “North”), and finally the targets to improve (right, or “East”). At the end on the very far right (running out of compass directions here …) comes the people responsible for the different tasks. Let’s have a (critical) look at the steps. In the corner intersections, you mark down if the points are related, or even how related they are.
- South: Long-Term Goals: The first step is the long-term goals. What is the overall direction you want to move your company (department)?
- West: Annual Objectives: Out of the long-term objectives the annual objectives are developed. What do you want to achieve this year? In the matrix between the long-term goals and the annual objectives, you mark which long-term goal is aligned with which annual goal.
- North: Top-Level Priorities: Next you develop the different activities you want to do to achieve the annual results. In the matrix in the corner, you again connect the previous annual objectives with the different priorities to achieve these objectives.
- East: Targets to Improve: Based on the top-level priorities, you create (numeric) targets to achieve this year. Again, in the field between the top-level priorities and the targets, you mark which priority influences which target. Some examples also have a matrix that connect the targets to improve back to the long-term goals. (Which may lead to inconsistencies if an entry in a step 1 long-term goal going through steps 2 and 3 does not influence any entry step 4 targets to improve, but the matrix leading back to step 1 sees a connection. Best not to use a matrix here.)
- Responsible: On the very far right you add which person is responsible for what top-level priorities.
Like the “normal” Hoshin Kanri, this document is done at different levels in the hierarchy, starting with the top executive. These are named rather straightforward as top-level matrix, second-level matrix, and third-level matrix.
From my point of view, there are a lot of things flawed with the X matrix.
- Long-term goals not long-term enough: The long-term goals I see online usually describe long term as three to four years … which for me coming from Toyota is too short. I would much prefer to have here the truly long-term visions or the foundations of the corporate philosophy here.
- Often redundant focus on numeric goals: Most examples I have seen online are overly numeric. The sequence could be as follows (actual example): 1)Long-Term Objectives: Gain market share to 40% -> 2) Annual Goal: Increase market share to 25% -> 3) Top-Level Priorities: Develop growth strategies -> 4) Targets to Improve: Increase market share to 25%. For this I do not need a fancy matrix! In this example, 2) and 4) are absolutely identical. In many examples I see online, the connection between the different fields feel forced and often redundant. In the original Hoshin Kanri, the focus is much more on having a process rather than having numeric targets.
- Diluting responsibilities: The Hoshin Kanri I know are always documents for one person. THIS person is responsible for the document, and has to work on implementing it. While he may work with his subordinates, the Hoshin document is his responsibility. This is different in the X-Matrix. Already as part of the design, the responsibilities are handed out right away to others. The person who made the X-Matrix is already no longer responsible. This to me is a very wrong mind-set!
- Where’s the PDCA?: Probably the biggest gripe I have is that the X-Matrix distracts from the PDCA! In the original Hoshin Kanri at Toyota, the PDCA is clearly part of the process. The X-Matrix, on the other hand, is devoid of any hints of the PDCA. While many articles about the matrix mention PDCA, it just feels like it is not there. Some articles see the X-Matrix as the first step to get the items to fill out a proper Hoshin Kanri, but they rarely go into detail for the actual Hoshin Kanri afterwards. Even if the X-Matrix is in preparation for an actual Hoshin, many of the fields feel redundant with the Hoshin. But again, if the PDCA is nonexistent, then the approach will have serious problems implementing it.
When to Use the X-Matrix?
Personally, I would try to avoid the X-Matrix, as I think it is introducing unnecessary overhead while losing some of the power of the original Hoshin Kanri. However, there are a few instances where it may be of use:
First, if you are already using it and it works for you. If that’s the case, then keep on using it. If you manage to use the X-Matrix successfully (that means with a PDCA) and you want to continue, then I am not going to stand in your way. To me, lean is whatever works, not a set of dogma-like methods.
Second, you may be in the unfortunate position that your boss or your client wants flashy and fancy methods. In this case, the X-Matrix may be a nice thing to bedazzle your client or boss. It won’t really make the works easier, but it looks sooo much cooler than the normal Hoshin Kanri to-do list and PDCA crossover. Just make sure that the implementation actually happens and the PDCA includes the C&A … but then, with this kind of boss or client, a fancy presentation with colorful slides may be substituted for actual progress. I don’t like it, but then … some people do want to be lied to.
But, if you are not yet using it and your boss is at least somewhat reasonable, then my advice is to avoid the X-Matrix and rather put the effort into the classical Hoshin Kanri. In my next post I will tell you a bit about the history of Hoshin Kanri and how the Kanri Noryoku program saved Toyota. Until then, stay away from overcomplicated tools, do the normal Hoshin Kanri, and go out and organize your industry!
P.S.: Many thanks to Isao Yoshino for his input!
18 thoughts on “Hoshin Kanri – Part 4: The X-Matrix?”
It is interesting to call it a “weapon”, but I guess it can be used that way to silence someone else with overly complex tools. Nice comment!
I agree that the X-Matrix is a shiny object. I’ve seen organizations do a great job with Hoshin Kanri without an X-Matrix in sight. It’s more about the mindsets and visualizing things in a way that makes sense to people. I’ve seen people really recoil at the X-Matrix. I remember being really confused the first time my Lean Six Sigma director in a manufacturing company showed me one (he had used it at Danaher and was a big fan). I’d say if the X-Matrix is being a barrier… maybe remove that barrier. I don’t think that’s “dumbing down” anything.
I’ve seen the impact of X-Matrix and the use of this tool as a weapon. I assume it’s rooted in the originators philosophy.
Dr Roser articulates TPS key points very clearly and accurately.
At Toyota, we feel simple, clear, smooth flow is critical to our success. This applies to both process and information
Christoph – Thank you for this series on Hoshin Kanri / Strategy Deployment. I am in agreement with you about the X-matrix. Most organizations and individuals I’ve worked with have found the x-matrix to add confusion rather than clarity. Though, if the tool helps support the thinking, then of course use it.
Isao Yoshino gave a seminar to a group of ~100 leaders in the San Francisco Bay Area last summer on hoshin kanri.
As I wrote about in a blog post following the event: “Many people at our seminar for AME were surprised to learn that Mr. Yoshino had never seen an x-matrix at Toyota and that there was not a set A3 template.”
What alternatives to the x matrix have people used? My company uses this, and for my team I just translated it to a simple PowerPoint table for easier consumption
Hi James, I would just use the normal Hoshin Kanri (see previous blog posts in this series).
I originally learning Strategy Deployment using the X-matrix approach. At the time we found it to be a great tool to keep the conversations focused on longer term objectives, what portion of objective was actionable in the current year, and how we would measure success. As an engineer I loved the way it all fit together in logical progression. Non engineers hated the format. But that is another discussion.
When we were done with the initial planning we found a room full of flip chart paper with all sorts of really important information hung on the walls, scattered throughout the room, and some on the floor. Embedded in all these flip charts was the thinking which was ultimately distilled into a few X-matrices we were left with. As a going away present from the top tier session each VP would get their very own to keep on the wall in their office. Of course, we made a commitment to have someone type up all the notes we could keep them for future reference. The reality was it never happened, or if it did no one every looked at them again. Another reality was that months later we forgot all the thinking that went into the process and only had the numbers on the matrices in front of us.
As will happen in a MBO culture those numbers would start to take on a life of their own and drive thinking and behaviors that were never intended during the planning stages. An example of this was an objective for a healthcare organization to increase patient capacity. In discussion we talked about efficiency improvements that would lead to a 20% increase in capacity. A year later it became an expensive construction project to increase floor space by the requisite amount. This was never the intention or the thinking that went into the original plans but all of that ended up in the trash with the rest of the flip charts that had our notes on them. We celebrated the on time completion of new bed space and moved on.
Later on an A3 approach to Strategy Deployment became popular and this was a tremendous breakthrough for us. We now had a a method of capturing our collective thinking in a focused and concise framework that would not end up in the trash a week after the meeting.
However just as hardly anything is ever all bad or all good there is a compromise to a pure A3 approach. It lacks the cross functional rigor that the X-matrix forced on us. If you seek to grow top line performance 20% and anticipate that a 5 points of that will come from organic growth in existing markets, another 5 points will come from innovation, and 10 points will come from an acquisition, an X-matrix will outline all of that quite nicely. Visual tools will flow out of that that keep track of everyone’s areas of responsibility and the measurements that will be used to determine success. In short it keeps track of the math in a practical manner that is useful for review throughout the year and years afterwards. HOWEVER THE THINKING BEHIND ALL THESE NUMBERS ENDS UP IN THE TRASH WITH FLIP CHART PAPER!
So over time we have evolved, and now we teach a hybrid method. We utilized the A3 approach to develop the thought process and reach thoughtful, practical conclusions in a collaborative manner, AND, we also use the X-Matrices to keep track of the numbers with rigor. With us it’s not a choice of either or. It’s both as each has their strengths and shortfalls. It’s a little more work but much more useful to the whole organization.
Hi Joe, thanks for the detailed and insightful comment. It still looks a bit complex, but if it works for you keep on using it! As you so nicely stated, any method is useless if it ends up in the trash 🙂
I implemented the X matrix at Danaher in the early 90s and it is still in use there today. I think this article puts too much emphasis on the evils of the matrix. It is only one tool of many in the Strategy Deployment process. There are several points in the article that I take issue with. First, I believe that Strategy Deployment drives breakthrough performance. This cannot be done by an individual or single department. It must be performed by a multi-functional team. Additionally, it is important to segregate Daily Managememt KPIs from Breakthrough targets. There is a whole discussion about the architecture of Strategy Deployment beyond the scope of this post. If you would like to discuss, please contact me.
Christopher. I can’t comment on the value of the X matrix because I haven’t used it. In the early 1990s I was sent to Japan to work with the TQM Promotion Division (the department that Nemoto created and led) to explore how to introduce Hoshin Kanri to the North American Toyota companies. The X-matix was not one of the tools I was shown and it was not part of the implementation effort while I was involved in until 2000. Mark Reich with LEI would be able to describe what was used at the North American level when he led strategy deployment in the early 2000s through the North American manufacturing headquarters. A tool we did use was the Cross-Function Responsibility Matrix and it was very useful in helping North Americans “see” their responsibility to either lead, support and coordinate with activities to execute company business priorities rather than assuming a priority was an individual function’s responsibility.
Thank you for this excellent article.
I learned something new with each job that you published.
Prof. Roser, thank you for such an insightful article. I do agree with your review of X-Matrix.
Lean Philosophy is about making things as simple, while we see X-Matrix is complex but without much worth. X-Matrix is not an original lean tool practiced by Toyota, and it seems driven and promoted by some consultants to impress with numbers.
Thank you for such enlightenment.
My Best wishes.
I tend to differ here on some points. First, we don’t need to follow exactly the original Hoshin for something to work. What if things have evolved greatly? Won’t Hoshin has to evolve to make an impact? Second, X Matrix if used alone can be what is said here. But X Matrix supported by backbriefing A3 and Experiment A3 tools is impactful in the PDCA cycle including the C-A part of the cycle. This means we should also see things in the eyes of others. Of course if not implemented properly like any other tool or methodology, one ends up pulling a blank.
Hello Joseph, first of all, if it works for you, keep on doing it! But to me the X-Matrix is way overdone.
I just read your article, and have one question, what would you recommend as a first time Hosin Kanri tool?, at this point I see more simple and beneficial starting only with PDCA for the main KPIs on yearly basis, or also I was thinking a combination with X matrix to define priorities based on long terms goals, which is my next step, I guess, and then follow each target with PDCA,
what are your thoughts?
Hello Luis, the PDCA for improvement is the start. I wouldn’t do a PDCA on a KPI, but I would pick a project for improvement, and do the PDCA on that. Hoshin Kanri builds on PDCA, but first you should make sure that your improvement activities actually improve something. Don’t do the X-Matrix, this is overkill.
I like the X Matrix simply as it makes you look at what is missing and where you are putting your resources that you probably shouldn’t be. As it moves into lower level ones, pet projects and ‘what do we think we should do this year that has no alignment’ are removed.
Its a little complex initially but it is a good visual tool to get everyone on the page and look at clashes at business level. I don’t often see a full suite of measures that come from the first phase but this does pick them up when you start to see it. Just my thoughts.