An Inherent Problem

This is a chapter from Amplio Development: The Path to Effective Lean-

Agile Teams

Most organizations are structured in hierarchies to manage the work. But the work flows across the organization. Management attends to the efficiency of these hierarchies instead of attending to the delivery of value. This has adverse effects.

Managing within a hierarchy

Hierarchies tend to have those responsible for the hierarchy see if the people under them are:

1. working on the right things
2. working well
3. fully occupied

This seems so natural that we don’t even think about it. However, it also means that the focus is on the silo, not across the workflow. This results in delays when handoffs are made between the silos. It is common for silos to prioritize work within the silos over cooperation between them. In fact, along with the hierarchical structure are reward and compensation systems that set these silos up to compete, instead of cooperating.

Reflect for a moment on projects you’ve seen in the past. Let’s look at the people doing the work and the work itself (when it is being worked on and when it is waiting to be worked on). See if what you notice matches the following table. Feel free to add to it.

 When we make these observations, we often see that no one is managing the work itself but are instead managing the people. Work is seen where it is, but the time it spends waiting is not usually seen.

When we look at the people doing the work, we focus on local steps in the workflow. This is the most significant difference between waterfall and Flow/Lean. In waterfall, we try to maximize the efficiency of the work at each step. The presupposition is that we can do this. Flow and Lean have us look at how long it takes to go from concept to consumption. A project that would take one month if all the people working on it worked on it alone can easily take six months to accomplish when they are busy working on many things.

The actual cost of multitasking.

Many people point to multitasking as a significant problem. It is a significant one, but not the key one. A significant cause of multitasking is people working in multiple value streams. People are being interrupted and are interrupting others. Correlated with multitasking is work waiting in queues. While multitasking may cause a 10-20% drop in efficiency, work waiting in queues can create unplanned work in the form of bugs; working on the wrong things, rework, etc., is much larger.

Focus on delays – not eliminating waste

A common Lean mantra is “eliminate waste.” The problem is that mantra comes from looking at manufacturing. In this context, you can see waste. A car is being built, and errors are visible. Planning can be considered waste because you already know what to do. In knowledge work, the situation is different. First, you can see everything. You often don’t know when you have an error or not. Also, things like design and planning are not wasteful. Overdesign and planning are, but these are still necessary functions. On the other hand, re-doing requirements, working from old ones, building unneeded features, fixing bugs, overbuilding frameworks, duplicating components, and integration errors are wasteful.

These wasteful activities create delays in getting the information and using it or when making an error and detecting it. This is yet another reason for quick feedback and reduced delays. The mantra needs to change from “eliminate waste” to “eliminate delays in the workflow.”

It’s not bottom-up or top-down, it’s attending to the value stream

Many people recognize that hierarchies can be problematic. Many Agile proponents suggest starting at the team. Most are now recognizing that getting executive to buy into Agile adoptions and suggest that a top-down bottom-up approach is required.

While top-down bottom-up is better than just bottom-up, this is still a focus on hierarchies. It often is a “starting at the team with guidance of management approach” – not a true value-oriented perspective.

A direct focus on value streams is essential. Otherwise we’re working on improving local optimizations.