Daniel T. Jones has contributed to arguably some of the most important management books of the last 30 years. In “The Machine that Changed the World”, and “Lean Thinking: Banish Waste and Create Wealth in Your Organization” Dan and Jim Womack describe the principles and practices behind the amazing growth of Japanese car manufacturer Toyota.
Dan’s thinking journey of the last few years have been quite interesting as he has been applying Lean thinking to the advent of digital. The latter has indeed been one of the main topics at the center of his interests: you can tell from the last conferences Dan gave. I personnaly find it interesting as this is somehow the opposite journey as the one I have made, from digital (collaboration platforms, agile …) to Lean. Note that I am not pretending here that my work is comparable to his invaluable contribution to management thinking.
Dan is a busy man. Yet he somehow managed to allow us some of his precious time. We are so grateful and honored to discuss the intersection of Lean thinking and digital with one of the men who invented the word Lean and deciphered Toyota system for us.
Digital has had a lot of traction lately. You have been talking quite a lot about it in the conferences you’ve made in the last couple of years. How does it articulate with IT and why do you think it is so important? How does digital change the game of IT as we knew it?
Digital brings the power of the computer in the hands of the user, which IT never did. It brings pervasive communication: between the company and users and provider and among different providers (What Clay Shirky calls the change of communication paradigm: from one-to-many standard top down diffusion to many-to-many networked communication – Cecil). IT was basically an old architecture that the web has now actually replaced. Digital brings fundamental new capabilities. But the seeds of this were visible much earlier, take for example what Rolls Royce was doing in aerospace twenty years ago as they track engine performance in real time because they sold Power by the Hour rather than selling engines. It takes a while for company to learn to build a two-way dialog with the customer rather than the one way push.
This is something you already had envisioned ten years ago in your book Lean Solutions
We tried to define value from the customers’ perspective. In a sense we were doing what you would today call Customer Journeys analysis. We were extending the logic of lean from understanding the production and delivery process to the customer journey. At that time, we thought this could be done with a sample of customers, but now we can do it with all customers thanks to the power of big data.
However, we must never forget the personal emotional part of the process: the data can’t tell the whole story. Yet, it’s very useful.
Organization is an embodiment of a way of thinking: command and control on one hand and systems designed by experts on the other.
In the last Lean Summit, you said that IT has been used to enforce command and control throughout the organization. Can you please elaborate? How and why according to you this has happened?
Command & control was top priority of senior managers in traditional organizations. As GE told us, you manage by the numbers, you hire smart people and give them targets. You don’t care how they achieve them. Not surprising management wanted big systems that gave them better control and compliance. However these big systems never fulfilled their promise, they introduced enormous rigidities and were very expensive to maintain. The web demands a much faster response, which means a different system architecture, like modular Service Orientated Architecture, which can be updated every day.
Do you mean the platform technical architecture has dictated the change?
This technical architecture reflects the change that is happening, led by digital companies: it focusses on rapid, incremental, accelerating improvements punctuated by occasional leaps. This is a major business perspective change as opposed to what has been done so far.
In traditional models, you push technology and products out to the customer and you bet your sales people will actually sell these. It is not so easy in the digital world. The rule of the game is different, power is in the hands of the customer.
What is it that big companies are missing on the digital topic according to you? Do you think they are ready to tackle the digital transformation?
Many traditional hardware companies are facing the integration of hardware and software, forcing them to think more like a digital company. Google has developed agile teams – much easier than for established companies. The latter have social stickiness that prevent them from moving fast enough. I’ve always believed that changing social organizations is slower than exploiting the opportunities opened up by new technology (check out #hypertextual blog post on that very idea). Social organization is the break. And lean is relevant because it helps create more agile organizations, more able to cope with the digital challenge.
If you try to automate processes when they’re not stabilized and streamlined, you just automate waste and chaos.
Organizational structures and forms are the biggest obstacles for big organization to cope with digital transformation – both in mindsets and organization structures. Organization is an embodiment of a way of thinking: command and control on one hand and systems designed by experts on the other. Lean changes the focus to improve the capabilities of the whole organization: engaging 80% of the people in continuous improvement rather than 15%. It takes a fundamental change in mindset to make organizations more responsive to the customer.
You said that we need to pull IT capabilities as opposed to be in a push IT environment. What is the distinction between both? What should be the strategy for digital on that very issue?
The current perspective of IT departments is they are the experts able to redesign the whole organization and its processes. Of course sitting in your function does not tell you what the real needs of the existing processes are. And the success rate of doing this is very low. So we need another solution: collocate software developers with user teams and work together to understand how users are struggling with existing, broken processes so they can first streamline the process and then automate it. The required mindset of IT in the digital world is that the teams have to be dispersed not centralized. Software is now so much part of everybody’s job, it is a key part of continuous improvement. 20 years ago we realized we needed to collocate design, developers, and business operations teams. My plea would be to break the IT department and disperse it throughout the organization so that business teams could pull IT capabilities as they need it.
Pierre Masai in Toyota did this: he established some guidelines for how they will solve business problems by responding in a pull mode to business needs. However, they work with them to do that only when the business teams has streamlined and stabilized their process. If you try to automate processes when they’re not stabilized and streamlined, you just automate waste and chaos. And if you do it remotely, you probably design the wrong solutions and generate streams of rework tickets. I hear many support teams saying they get many rework tickets because the product is complicated. If the IT teams were with the user teams they would see what the problem is and develop more robust solutions.
We built our companies with this pyramid structure with knowledge embodied in experts while digital is spreading this knowledge throughout the organization.
You said this stunning thing at Lean IT Summit 2014: “What Diderot did for the craftsmen in the 18th century, Google has done for the Knowledge Workers in the 21st century ». Could you please elaborate on this quote?
Diderot wrote down the craft knowledge that was protected by the guilds. The craftsmen were angered by this because their place in society was acquired through their knowledge. What Google is doing is making knowledge widely available on the web. The position of experts will change from knowing many detailed things to how to use them in a particular context. They no longer have a monopoly of technological knowledge. For medical knowledge, patients can go on the web finding papers, exchanging experiences with patients with similar pathologies: they now have different questions for their doctors expecting them to advise them on which path to follow. Doctor no longer has monopoly of wisdom but co-manage the knowledge with the patient (what french philosopher Michel Serres calls “Présomption de compétence”[FR] – Cecil). It is fundamentally changing the role of experts and it is a good thing because if you look at the literature on the success of experts, they are often wrong because they come with point answers to more complicated topics which span different fields. On the other hand, expert working with other experts from different fields are creating new knowledge. We built our companies with this pyramid structure with knowledge embodied in experts while digital is spreading this knowledge throughout the organization.
What can Lean thinking bring to the digital enterprise?
Lean is a learning system and at its core is a product development system. Toyota Production System is a way of delivering knowledge. Their aim was to develop knowledge with less investment and more quickly. In doing so they also organized the work visually and reviewed progress on a daily and weekly basis. This goes a long way to resolving the cumulative delays with a waterfall approach. Agile tries to do the same thing, but is often used very loosely with little attention paid to capturing reusable knowledge. The much richer business context in Toyota Product Development has still not been understood.
You also gave in this Lean Summit talk the example of Spotify, a company you have visited. You were quite impressed with their organization but said they had no improvement system. How would you explain that?
Spotify is an extremely interesting example. What is interesting is that they’re embodying three different dimensions in their teams. First, everyone in a team has different technological capabilities, but they are also part of their tribe where they get together and deepen their technological knowledge. Second, they have an Agile coach to help team manage the work and the visual management of the process. Third, someone takes responsibility for the product. Thus they are able to combine vertical knowledge (functional), horizontal learning (process) and product learning. This is a very interesting organization for the future (refer to this Henrik Kniberg video describing it). My questions to them was around the degree to which they capture reusable knowledge and, building on that, a learning system that captures the standardized knowledge that comes out of the process. From what I’ve seen, I missed the XP dimension where the focus is on the quality of the software and endlessly improving it. We learnt about that from Kent Beck at Lean IT Summit 2015.
Discussing this triggered a debate within Spotify about that. Yes they haven’t been focusing on this topic. Organizing systems with productive environment for software craftsmen but not explicitly on reusable knowledge. The other thing I questioned them about is “Can you extend this across the business?” and they’re trying to do that (sales, market, support functions …). It’s easier in IT but not so easy in finance. Now, that model doesn’t need to be replicated. It’s a piece of the answer and not the whole story. People would be foolish to copy it (Well I somehow did – Refer to my story where we have implemented something very similar at a software vendor house – Cecil). They must figure out what to learn from it. Just like copying Toyota blindly without understanding what it involves. Spotify talk at Lean IT Summit 2014 was very interesting.
If management don’t change and learn to find problems, there will be no learning organization.
Everybody is talking about the learning organization. Yet no-one (apart from Lean practitioners) are really clear about learning process/system. How do you explain that?
Michael Ballé, Jacques Chaize and Orre Fiume are writing a book on this to be published next year. The core idea is that while everybody has got very excited with problem solving (PDCA) as the core of the learning organization, it’s only half the story. Again, problem solving is fine just as long as you solve the right problems. The management task in the learning organization is to find the right problems and frame them so that teams dedicated their efforts to solving the right problem for the organization. We need a different way of thinking to find the right problem. The question is how do you see the big underlying problem you couldn’t see before. Answer is: on the Gemba, whether the user’s, provider’s, software development or operations Gemba, where value is actually being created. That is typically not what upper management does. If management don’t change and learn to find problems, there will be no learning organization. Most organizations are not convinced about this approach and delegate learning to training, which is completely different.
Last week I was with an executive team and we were discussing. They talked about cost and competition. We then discussed the Toyota story and we went to the shop floor. They really looked with open eyes and open minds. Then I asked them what was the problem? They said it is management. They found this out on shop floor, while discussing with people. They realized what most people were struggling with: broken process they designed and forced compliance with. If things are going to change it is by unlocking the obstacles to change. It’s management vision and a deep understanding of the dysfunctional nature of the process they put in place that is the key to change. Management’s role is business improvement.
Lean has outlasted many other change movements and I believe it will continue to stand as a very robust organization system and management practice.
The whole management consulting and literature is vibrating around agile (or even worst, Scrum which actually is a project management method) which is highly derived from lean, while the actual lean literature doesn’t enjoy so much coverage The remarkable success of lean startup tends to embody this trend. Same with Holocracy. How do you explain that?
I think people are looking for shiny quick answers because they are consultants or managers. You look for easy answers rather hard ones. Lean is hard and people don’t want simple thing. If you look at the agile community, most of them were inspired by Toyota one way or the other. It takes deep knowledge to put the pieces together. There is no easy answer and they don’t have the discipline and depth of understanding. As far as Holocracy is concerned, this is pure nonsense and it will disappear very quickly. Agile is fine until you realize there’s more to it. As people really study business as opposed to finding a consulting method, you realize that Toyota is a rich source of knowledge and inspiration. It has outlasted many other change movements and I believe it will continue to stand as a very robust organization system and management practice.