Kanban: can we follow the learning from Pied Piper in ‘Silicon Valley’?

Youu may have seen the rather clever and funny HBO sitcom ‘Silicon Valley‘ exploring the lives of five white young male software engineers build a startup called ‘Pied Piper’. In the show, the team at ‘Pied Piper are developing their new IT products, and you may be familiar with the whiteboard often in use when they are in the midst of a massive amount of code programming. (This massive amount of code programming I believe is also sometimes known as a sprint, I think they are loosely modelling an agile improvement approach in the programme).

The whiteboard seems to represent a really simple process. At the fictional ‘Pied Piper’, this consists of three columns, one for ‘to do’; one for ‘work in progress’ and ‘completed’ (or similar wording). In each column is a number of post-it’s with a task on, all starting in the ‘to do’ column. A programmer picks one post–it from the ‘to do’ column, accepts it as his (and in this programme they are largely all male) and moves it across the board accordingly. This simple process allows everyone at a glance to know who is working on what, and how many tasks still need completing. The team of ‘Pied Piper’ plucky engineers love celebrating after their frequent all nighters when all of the post-its that detail each programming task have moved from the left hand ‘to do’ column to the ‘completed’ column.

Increasingly, I have noticed that this type of board is becoming ubiquitously followed and used in offices and other settings often referred to as a ‘kanban board’ or sometimes a ‘personal kanban board’ which increasingly has made me feel a little uncomfortable. Why uncomfortable? Because in these examples, kanban is being viewed as a simple visual management process for managing workload, and yet for me this doesn’t really resemble kanban at all, and only very partially meets the principles of kanban. In short, whilst these have undoubtedly good intent and are helpful, these types of whiteboards seem to be more workload tracking systems, than kanban systems. And I’m not even certain that a kanban boards is a good name for them. This is because for me, they lack fidelity to the practice of kanban and these boards could be more effective work organising systems through more careful application of kanban principles.

So, if we can’t follow ‘Pied Piper’ to learn about kanban what can we learn?

First, a kanban is a cardboard card, rather than a whiteboard. Each card corresponds to an amount of work/work task. So at ‘Pied Piper’ the kanban card is each separate post-it. So far so good you say? That still seems like kanban? Yet each post-it at ‘Pied Piper’ may not equate to the same amount of work… one might be a 30min task, another a 3 hour task. So how does the ‘Pied Piper’ system really help us to visualise workload?

Second, A kanban system is a card system that materialises information to ensure we can visualise flow, both material and information. Again… so far so good at ‘Pied Piper’, we can see tasks flowing across the board… but how much information can we see? Is it clear how the tasks connect?

Third, There are two main sorts of kanban cards (Balle et al, 2017):

1) withdrawal cards: these are mini ‘purchase orders’ for each task, tasks/information cannot flow to the next step without a withdrawal card to trigger the movement. This card represents demand.

2) production instruction cards: when a task is triggered by a withdrawal card, a related production card it is sent to trigger the start the the production or service task as a request to start, e.g to the programmer. This card represents the production response to the customer demand, is an extremely visual way, and helps to see blockages.

Can you see these two cards and their functions on the ‘Pied Piper’ whiteboard system? I can’t. I can only see one, which might be ok, if they have done something else in their system… but I can’t see it. I usually can’t see two systems one for demand and one for production in office systems either, yet work in offices is rarely ‘produce by only one person’ so surely some kind of ‘production order’ is still needed in response to the customer demand.

Kanban (in lean) is used as the basis of just-in-time and one piece flow, and was based on the idea of supermarkets where customers pick what they want and shelf stackers fill in the space on the shelf after they have picked their product. Kanban imposes discipline to ensure no more than needed is produced, in demand order according to the customer and kanban supports other processes linked to getting it right first time. (at ‘Pied Piper’ the programmers correct all their programming errors at the end, rather than as they go).

The purpose of kanban is to have the least possible number of cards possible in use at any one time. This minimises work in progress thus the cost of inventory and stock holding. For example, if an employee has a whole pile of task post-its, or a whiteboard has more post-its in the work in progress column than it has staff working, then this defeats the purpose of using kanban. The post-its (cards) are not being used to drive one piece/task/decision flow.

This is important because studies show that there is as much as 20% lost time as employees switch between different tasks without completing them and the risk of errors and mistakes increases. Kanban when used correctly enforces a limit to the number of tasks anyone person is doing at a point in time. This also has an added benefit for employees as it reduces overburden and the interpersonal conflicts and frustrations that can arise from resource allocation decisions.

Nevertheless kanban is also difficult to learn and use, I am still learning its intricacies, and how kanban principles can be applied in healthcare, and yet it is deceptively simple to look at and consequentially mis-understood and imho, mis-applied.

The book ‘The Lean Strategy’ (from which I have drawn heavily in this post) suggests that a way of managing office based workloads such as that in the ‘Silicon Valley’ programme might need 2 whiteboards (Balle et al., 2017, page 101). One containing ‘to do’ tasks without prioritisation, and another containing a grid sized for capacity, and new tasks can only start when an aisle is freed up (see photo below).

Photo from The Lean Strategy, Balle et al., 2017, page 101

This process controls the flow of work tasks, limiting batching and work in progress, and even more importantly, reveals where problems and overburden in workflow are. For service tasks such as programming, designing, policy making and product development, the daily practice of reviewing flow by all team members and leaders together to solve problems, reveals holds ups, errors, delays and where team members need help (do we see that at ‘Pied Piper’ or most office based systems?). It also illuminates where there are areas of over processing and overwork, leading to simpler and more elegant designs and can accelerate office-based productivity.

So whilst, I can see the attraction of simple whiteboards used to visualise workload tasks… for me at least, these are not really kanban boards, because these 3 column whiteboards fail to benefit from the potential of kanban by failing to apply the principles that underpin it. For me, the name of the board and the language used matters. Fidelity of use matters, because if other learn to use ‘kanban boards’ that do not hold true to the principles of kanban. And learners perceive them as ‘best practice’ or similar, then the benefit of such visual systems will not be felt and misunderstandings of what kanban is, and isn’t, and what it can be used for, may fail grow further. Possibly perpetuating messages of that ‘improvement’ stuff doesn’t work.

So I am not going follow ‘Pied Piper’ and use a similar whiteboard workload tracker from but I am going to keep learning to use improvement tools like kanban with as much fidelity as I can, to keep benefitting my customers, and in healthcare, our patients. If we are going to practice improvement, is it not important that we try to practice it well?

What’s your view, does fidelity matter? Am I being too technically picky? Does method really matter that much for this? Is this just a difference between agile and lean approaches to improvement? Do you disagree? Love to know!

Advertisements

About joyfurnival

Interested in TQM, lean, Quality Improvement, Healthcare, Regulation, Accreditation, Inspection Improvement Science, QI, improvement capability.
This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

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

WordPress.com Logo

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

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s