We create monsters and then we can’t control them. -Joel Coen
Remixed Photo Credit: croweb |
Complexity minded people spend a lot of time talking/thinking about detecting weak signals, but sometimes big scary company killing problems are obvious to some, but unseen by others. The signal isn’t weak, but it isn’t traversing the social network well enough to get addressed.
Individuals in the Kanban community have taken to referring to problems like these as “Intangibles.”
Weak signal problems require attuning our senses to the present to “feel” subtle changes in what is happening. Intangible problems are caused by a disconnect between those who understand that addressing what is most painful today, may leave us exposed to what may be VERY painful in the future.
“Intangible class items may be important and valuable, but there is no tangible cost of delay associated with them in the near future….
-David Anderson Kanban“
Intangibles have a power law curve for future costs |
“Company Killers” as I like to call them are a special subset of Intangibles. They are large, monolithic, complex problems that hang around the company. These problems are elephantine, apparently docile if unprovoked, but prone to rampages once awakened. Company Killers are too frightening to approach and are shunted off never to be spoken of again.
Defusing Company Killers
Remixed Photo Credit: donsolo |
Defusing Company Killers is a two step process.
First, expose the Killers to management by “drawing them a picture”, so the Killers can be seen clearly.
Second, illustrate the amount of WIP vs Slack, so management can understand how their focus on the near term is effecting their risks in the long term.
In this post I would like to explain an exercise I have successfully used in engineering departments for “drawing management a picture”, in order to flush out Company Killers. This exercise has been particularly helpful for me when “launching” Kanban (or other WIP Limiting frameworks) efforts as it illustrates risk, WIP and their relationship.
Exposing the problems visually to management can help them “see” risk and then actively manage the reduction in risk, without invoking the Ostrich Algorithm.
The Exercise
Ingredients
Like any good visualization exercise you are going to need:
- Sharpies
- 3 Colors of Stickie (Yellow, Pink and Green work well)
- A White Board
- Team members
- Managers
Step 1. Get It All Out
Setup:
Begin getting the Team Members together. You will need the managers later, but not now. Place one yellow sticky and one blue sticky on the white board to create a legend, label them “Work in Progress” and “Work Someone Expects Me to Do.”
Distribute the yellow and blue stickies to each of the participants.
The instructions for the participants:
“We are going to build a picture so that we can show managers how their expectations are effecting our work.
On the yellow stickies, write any work that you are currently working on, or have started but not finished. On the blue stickies, please write any work that you are aware of and that you believe someone expects you to do. Please spend 5-10 minutes writing down EVERYTHING you can think of for each of these to colors.”
Step 2. See the WIP
Setup:
When the team member are done at the board, draw an X & Y axis, using the full height and width of the white board. Label the X axis “Complexity” and the Y axis “SIZE”. Put a red dot or an X mark in the top right hand corner. Write “Fix a Typo” at the origin of the graph.
“We are going to do an Affinity Diagram that clusters relatively similar items together across this space. Please cluster the items by size and complexity. We aren’t looking for an exact location for each sticky, just place each in relative size and complexity to it’s neighbors.
Down here at the origin of the graph is where you put the simplest task possible, like fixing a typo. Hopefully, none of you have anything that small on your plate.
Up here in the top right [point to red dot or X] is something as huge and complex as creating a replacement space craft for the decommissioned Space Shuttle.
Please go to the board one at a time and briefly explain each task as you place them on the graph.”
Anonymized Result |
In my experience, when giving extreme examples like “fixing a typo” and “replacing the Space Shuttle”, it has not been necessary to explain Simple Complicated and Complex. The extreme examples –something way too small and something far too large –give the team “space”. By making the extremes outside of the team’s normal expectations they are less likely to place their stickies on those ends of the graph. They are more likely to use a wider area between the axis.
Step 3. Find the Killers
Setup:
Place one pink sticky on the white board to finalize the legend, label it “Company Killer”
Distribute pink stickies to each of the participants.
The instructions for the participants:
“We need to find the projects or ideas that are going to kill the company. Spend five minutes thinking of projects or tasks that, if not completed in the next 6-12 months, would kill the company.”
After 5 mins.
“If there are any tasks already on the board that you know of that would kill the company, please write those on a pink sticky, as well. Then put the new stickies on the board relative to the other tasks we’ve identified. If you have a pink sticky for a task already on the board, please remove the old one and replace it with the pink one.”
Anonymized Result |
Step 4. Explain Yourself
Step 4.1 Help the Team See
The Churn Zone
The Churn Zone |
Nobody likes fixing typos over and over –this is “the churn zone”. Work here is busy work. It’s small, simple work. Large amounts of work items here indicate a high likelihood of multitasking. The team could be fragmenting the work too much if they identify too many items that fit here. Questions to ask: Who is assigning this work? What tasks are highly repetitive? Can we automate any of them?
The Strange Nature of Company Killers
The Kill Zone |
The tasks in the “Kill Zone” area –large, complex Company Killers, have an odd nature. Team members can identify them, but the risk is that they APPEAR well understood and encapsulated. One of the ways that Company Killers hide is in Cognitive Ease. They are large and unfragmented and everyone agrees on them even if they ignore them. The problem is, that when they are actually examined, Company Killers tend to take a “quantum leap” into the Unmodelled area skipping right over Ordered and Complex and hurtling into the chaotic. As the team unpacks the problem, they realize that they don’t really understand it as well as they thought they did.
Company Killers’ Quantum Leap |
Company Killers need to be decomposed carefully and early. Identifying test projects, agile “spikes”, and safe-fail experiments are critical to helping the teams and management truly understand the scope of these issues.
Step 4.2 Show Management
It is critical to have all the managers who are responsible for assigning work to the team members come in and examine the visual aid the team has created. Ideally, a team member (or two) can explain the image, how it was made and the implications. Show the balance between The Churn Zone and The Kill Zone. Make sure that managers leave with an understanding of:
- WIP
- Intangible Class Items (and how to manage them)
- The relationship between WIP, Slack and Intangibles
Finally, this is the time to explain to management how a Kanban system can help limit WIP and focus teams on balancing near term and long term efforts.
if you made it this far you are a trooper, thanks! Maybe you want to click a +1 or tweet about this post?
A really nice article. I am curious why the Kanban list is not addressing these company killers using real options and feature injection. Use feature injection to gather the information so that you know how long it takes to address the issue. Also find indicators to detect that an issue is becoming a company killer.I have experience of one project that was a company killer (It wasn't really a threat to the company but it almost killed me). What experience do you have with company killers? The real death of any project is caused by the "thousand cuts". All of those "it will only take 5 minutes fixes" that take several days.The use of "Intangibles" is possibly the worst name for these items. Intangibles has a specific meaning to accountants. Telling some one that you have a problem with intangibles is bound to cause confustion.
Chris, I am positive that many people on the Kanban list would use Options thinking to resolve these company killers. Finding indicators to detect issues becoming company killers seems very Cynefinish, maybe I should write about that. I would agree that another form of killing is through highly fragmented unseen work, your thousand cuts. I think in the WIP Model those are likely to be more toward the chaos fragmented area, in the Space Shuttle diagram we would expect to see those in the Churn area. As an ex-consultant I have too many years of fire fighting. In general I was probably less clear than I should have been about this exercise I think it is probably most useful when you are trying to "get a feel" for a team that is in trouble or wants to transition to a Kanban board. I like the exercise because it creates a shared understanding of the work between the team and management. The exercise itself offers no insight into how to fix the problems (though I think Kanban and Real Options do). I am going to write up another post to work on the Intangibles name.-jabe