It is not uncommon to find ourselves needing to answer such questions as:

  • How much documentation should we write?
  • How much of our code needs test coverage?
  • How much work or how many projects can we work on at one time?

There are no stock answers to these questions. The right answer for you may be different than the right answer for the person next to you in the line for coffee in the morning. There is a parable from “The Way of Testivus” that explains this quite well.  There are some of us that want someone to tell us what to do so we can just go do it and not have to figure it out. But, there are things that no one outside of your context can tell you – they can only give you pointers and guide you in the decision-making process.

To that end, I’ve come up with a very rudimentary A3-like template that I call the Lagom Discovery Canvas for guiding you through a thinking process that helps you decide where you want to operate in a spectrum of too much or too little of something. It does this by focusing on outcomes and consequences of operating in different levels of the idea you are considering. The canvas is available in PDF form currently and has two pages: one with prompter questions for each section to help guide you and one that is blank for you to fill in.

Let’s walk through the canvas and how to use it…

The Discovery

The first step of most endeavors is to gather and understand the current condition and our target condition. The first few steps guide us through some discovery so we can form our experiment.

Step 1: The Decision

The best time to use the canvas is when you need to make a decision about how to operate. (Read about ODIM to learn how outcomes should drive our decisions.) When you have the decision clearly in mind, you record it on the canvas. For this post, our decision will be “How much work should we start at one time?”

Decision

Step 2: The Outcomes / Conditions

Once the decision is recorded, it is time to think about what conditions should look like when you operate at various points on a spectrum related to this decision: doing too little work at one time, doing a lagom amount of work (just the right) amount, and doing too much work at one time. Let’s fill out the canvas starting with lagom as that is the state we want to get to – our target condition.

Lagom: When we work with an optimal amount of work-in-process (WiP) we believe we will see fast delivery due to reduced waiting time for each piece of work, better quality due to increased focus, and better sustainability because our team members are working at a sustainable pace and are challenged without being overwhelmed.

Too little: With too little work, we’ll have fast delivery but workers will be bored and unchallenged, which can lead to complacency and poor quality. Stakeholders won’t be happy because our throughput will be lower than it could be and when management walks around, they see people idle and think we’re overstaffed. While busyness isn’t what we want to optimize for, there is such a thing as doing too little work.

Coaches note: What’s interesting is that if we only measured cycle time (speed) then it would look like we’re doing great when we have too little work, but a balanced set of metrics shows a different story.

Too much: When there’s too much work it creates dissatisfaction in spades. Our delivery times will be much slower than desired. Our throughput will be extremely low. Quality will be poor due to context switching and lack of focus. Team members will experience burnout and turnover could be high. There’s no fun in the land of too much WiP.

As we discover this, we’ll capture it on a canvas. You can be very quantitative in your discovery. My example for this post is very general and qualitative.

outcomes

Step 3: The Natural Tendency

Once we have put some thought into the different outcomes we might find on the spectrum of choices, the next step is to determine where you naturally operate when left unmanaged. Most of us operate in the land of doing too much work at one time, so I put an X in the box in that area on the canvas.

tendency

Your natural tendency is important to consider because it will help us understand what state we need to push away from and that informs our hypothesis and experiment, which is what the rest of the canvas is for.

The Experiment

Just like an A3 canvas, there is a section for experimenting with countermeasures.  Let’s break it down…

Step 4: The Hypothesis

The first step in experimentation is to create a hypothesis. Here is where you determine what action can be taken to push you away from your natural tendency towards a lagom state. My opinion is to take small or moderately sized steps to avoid a pendulum swing to an opposite, and equally unhealthy, operating state. For our example, moving from 30 to 20 things in process at one time seems like it would help us operate at a more lagom state.

hypothesis

Steps 5-7: PDSA

Once you have a hypothesis, you are ready to begin the PDSA cycle.  First, you make a plan. How will you test your hypothesis? What will you do specifically? How will you know you’re successful? When will you look at the results? In this example, our plan is to enable WiP limits on our virtual board and set them to 20. We plan to stick to these WiP limits, the only exception will be stop-the-line events. We will review throughput, cycle time, quality, and team satisfaction metrics in 30 days.

At the 30 day mark we come back to see what the findings are. It looks like we’re on the right track. Burnout seems to be a bit better and throughput is higher. But, we haven’t moved the needle on improving quality and we only were able to adhere to our WiP limits 60% of the time due to excessive stop-the-line events.

We decide to continue with our current hypothesis but reduce the WiP to 15 to account for some of the excessive stop-the-line events until we can reduce how many of those we have. Then, maybe we can increase it again. We’ll check again in 30 days.

On the canvas, it might look like this:

PDSA

For each iteration of the PDSA cycle, I’d attach a new canvas on top, keeping track of all past iterations.

How you can help me help you

I would love for people to try this out to see if it actually helps guide them in making the right decisions. Tell me what works. Tell me what’s wrong. Tell me what’s missing and what’s superfluous. Send me your stories and pictures! I would love to put them on the website. You can contact me via the contact form, the comments on this post, or via twitter (click on the feed to your right).