Interview with David Anderson, Part 1: Triage Tables: developing the tool for prioritizing a workflow in a WIP limit environment. The story behind the scenes.

Triage Tables were brought to light, as a solution, during the lock-down in July, 2020 from David Anderson’s office in Bilbao, Spain. Triage Tables can be called a breakthrough in the Agile world. They are the only known tool, with a solid mathematical foundation, to solve the problem of prioritization for a limited capacity system based on an understanding of the Cost of Delay. 
In this interview we talk with David about the development of Triage Tables. Asking him about his biggest inspirations, challenges, and milestones.

I am not sure we ever thought about the size of our audience. What we do know is that we have tens of thousands of people who have taken Kanban training and almost the same amount of people who follow our newsletters. What we could not predict is the percentage of people that might be interested in the Triage Tables Poster. However, two thousand people was very good result and we think it was worth producing it.  We hope there are a lot of people that we helped.

Tell us more about behind the scenes – how did you come up with the idea, what inspired you? When and where did it happen?

In summer of 2016 I worked out the mathematics for the Cost of Delay problems along with Andy Carmichael. It is quite a challenging mathematical thing to figure out; what is the possible impact over delay in decision to start working on something? There are two sets of unknowns, how long is it going to take us to finish the item once we start working on it? and much the bigger unknown – what benefit we are going to get and when does this benefit occur?

Since I worked it out in the summer of 2016, I started sharing it at some conferences and workshops. It turned out to be really enlightening for a lot of people. But the reality is that this complicated math involves a convolution of two curves – the lead time and lifecycle value acquisition curves and that is why it is not very accessible for people without a lot of knowledge in this field.

Somewhere in 2017, Dragos Dimitrius, who was involved in the beginning of the Microsoft XIT case and now  is a consultant and trainer, was visiting me in our Seattle office. I was showing him all of these studies and he said: “This is great, but most people can`t use it. Product managers can`t do this kind of mathematics. What you need is something like scuba dive tables!”.

And it was it. Because in 1950 scuba divers worked out how long you can stay under the water, how deep you can dive, and the impact it has for the body. They came up with a mathematical solution for professional diving for people who do sea diving for their living as well as military people and the different risks they are willing to take. But for people who want to do recreational scuba diving, they could not rely on them doing the mathematics themselves. To make this material accessible for the wider audience they published a set of look-up tables that could tell how long you can dive at a certain depth, and depending on how long you dive, how long you have to wait before flying home, etc. Dragos thought it a good idea for us to have an equivalent of that.

However, it is easy to suggest, but it’s much harder to figure it out. It took another 3 years and an additional piece of inspiration – to combine the ratio of the range of the lead time with lifecycle of the value acquisition. If you think about it as something that may take up to 1 year to develop and you are going to make money from it for 5 years, then that is a 5:1 ratio and so on. It turned out if we thought about the problem using these ratios, it enabled us to have a look-up table solution, and the Triage Tables were born from that. It was a suggestion of can we have an equivalent of scuba dive tables, but it was 3 years later before the epiphany, that if we use ratio it would enable us to create the tables. Then it was just a case of mapping out all the convolution functions and figuring out how to put them into the tables.

Where were you working on the tables?

The original idea and the original mathematics were done in 2016 and the suggestion about the scuba dive tables was in 2017, I was in Seattle office then. But the Triage Tables were mostly worked out during the winter tour 2019 -2020 and I finished them when I was in Lima, Peru in February 2020.

Why and how did you decide to apply triage approach to these studies?

There wasn`t really a decision. The need for triage happens naturally as soon as you accept that you have limited resources. Triage originates from emergency rooms, even from field hospitals. In military conflicts when there is a limited number of operating tables, surgeons, and hospital beds, there is a scarcity of resources. If the field hospital become overwhelmed, if there is too much demand and they do not have enough doctors, surgeons, beds, etc. , then they must decide who gets attention and who does not. It naturally creates the problem of now, later, or not at all. These are the three categories of triage. Who gets treatment immediately, who gets it later and who does not get it at all.

When we use Kanban systems for any kind of service delivery work it creates a constraint – the limit of work. Therefore, the Kanban system WIP limit of 10 means you can have only 10 things in progress. So, which 10 things are you going to do now? Because everything else must wait until later. And perhaps it may wait till forever because you`ll never get around to it: you get too much demand compared to your available capability. That is why you really need to have the 3 categories of now, later, and not at all. Triage might have this in three categories, but the decision making of now, later, or not at all must be based on the Cost of Delay.

Originally with field hospitals when wounded soldiers have been brought from the battlefield the nurse had to decide which category this soldier went into. The decision was based on the severity of his injuries. And quiet literally of how fast he is “bleeding to death”. If he is bleeding to death too quicky – there is nothing we can do, so he is in “not at all” category. If he is bleeding to death quickly, but if we act immediately, we can save his life – he is in the “now” category. And if he`s bleeding much more slowly, he can wait until “later”.

What needs to happen in professional service knowledge worker industries is that we need a metaphorical equivalent for “what`s the rate the patient is bleeding to death”. We need to understand the probable impact is if we suffer one day, one week or one month of delay. So, the cost of delay is the method we use for deciding between these three triage categories. The Triage Tables put triage and cost of delay analysis capability together into accessible tool, that does not require mathematical knowledge and presents a cost of delay problem as a series of simple questions that an experienced Product Manager can answer. 

What was the hardest and the most challenging in developing the Tables?

Some of it is just hard work. All together it`s 55 combinations in a current table, each one of those involves a convolution of mathematical functions. So there`s a lot of hard work. But probably the most challenging was the idea in the very first place, the epiphany, the innovation. The Triage Tables are a combination of 10 years of work, but it’s probably only 1 month of work that was spread over 10 years of time. Because the hardest thing was figuring out the problem and how to construct a tool that would be accessible and easy to use.

What are the requirements for Kanban system or business to use it?

The whole point of Triage Tables is to help take a basic flow system Maturity Level 2 and advance it to Maturity Level 3 or 4, to enable proper WIP limits. As soon as you enforce the WIP limit to a Kanban system, it creates a triage challenge of now, later, and not at all. So there needs to be a method of deciding on now, later, or not at all. Triage Tables provide that method. The requirements are of course that you have a Kanban system or at least you have a service delivery that has a restrained capacity, a WIP limit. And the Triage Tables provide the rest of the solution to enable that. Before we had Triage Tables many organizations struggled to implement Maturity Level 3 or 4 Kanban solutions, they got stuck at level 2. The goal with Kanban adoption is to make it easier to get to levels 3 or 4.

What type of business do you think should adapt the approach?

All modern professional services businesses. They need to be thinking of themselves as a service business. They need to be managing their services and recognize that people doing the work have limited capacity. Kanban helps them with all of that.

What type of professionals were you having in mind, while developing The Tables?

Product managers and anyone who is involved in scheduling the work: Project Managers, Service Delivery Managers, and anyone who`s job is to figure out what should people work on and when should they start it.

What about Menta Triage DS – the application that works as Triage Tables, but calculates the result in seconds as a calculator. Who inspired you to make Triage Tables solution digital?

The posters are all well, and we encourage people to print them out and put on the wall where they regularly hold meetings to decide what will they work on next, where they schedule their work. Of course, a lot of people will use the same meeting room all the week for this kind of meeting. But the poster still requires people to do the work manually and it just seemed so much simpler if there would be some software, that would prompt you for the questions. So you would answer the questions from dropped-down menus or just type-in some information and it calculates the answer. So rather than looking up the Triage Table manually, the application calculates the answer for you. It seemed to make more sense because as a mobile App, because people can have their phone in the pocket, they can use it while walking around a workplace, in different meeting rooms or in their private office. And if they are asked question, they can answer that with confidence just by inputting a few basic pieces of information, like using a calculator, and it provides them with the answer. So, it seemed like the obvious step forward. The poster is useful, but Menta Triage App gives you confidence in the palm of your hand.

This is the first application developed by one of the entities of your company, Mauvisoft. Did you feel comfortable going into this field?

I have lots of experience in the software industry. The critical thing for a business like ours is if we are going into a world of software development and providing applications, we had to choose to do something, that would add value. There is no point in competing with existing software solutions. We wanted to find something we could develop quickly and that would add value immediately for people, providing the functionality they could not get anywhere else. And the Menta Triage application using Triage Tables provided that opportunity.

Do you think people should learn information on how to use the Triage Tables beforehand?

The idea is that you don`t need specific education. However, Menta Triage DS or the Triage Tables still require you to answer some questions. I think those questions are fairly basic: from historical evidence how long it should take to develop this piece of work; once you launch it on the market, how much time you think you could make money from it; and do you think you`ll make the most of the money early in the lifecycle, half-way through or later in the lifecycle. These are not difficult questions. But what we discovered since we launched the Tables and the, is that there is a remarkable number of people who struggled to answer these questions. And therefore, perhaps they would benefit from some education, some training to help them understand the concept.

The Triage Tables topic is covered in our Kanban for Design and Innovation (KDI) class and in our Enterprise Services Planning (ESP) classes, where we also give people a deep dive into the mathematics behind the concept. Therefore, for the people who just simply want to know how to use the Tables or the App, the KDI class is enough, but for those, who want to understand more about how it works underneath, we explain it during ESP classes.

Are there any accredited KU trainers, that teach classes, that include triage studies?

Anyone who is qualified to deliver the Kanban System Improvement (KMP II) class. For the ESP separately, there is a class that is called ESP trainers. However, there are only handful of them.

Learn more about Triage Tables in Kanban Maturity Model book or access it online using kmm.plus.

Download the Triage Tables and Dependency Management Posters here.

Take a training at David J Anderson School of Management or find your local trainer at Kanban University.