How to Stabilize your System

“It is not enough to just do your best or work hard; You must know what to work on.”― W. Edwards Deming?

In my last article, I argued that the first order of business to improve a system is to stabilize it. To stabilize a system, you must first determine why the system has more work-in-progress than it can deliver within the current measured flow time.

There are a number of reasons a system may be accruing too much WIP, and it changes for each work profile (feature, defect, risk and debt flow items). Let’s dive into that.?

Features

When feature work accumulates, it’s usually in three places:?

1. Stuck in Ideate: At the far left side of your work intake process (known as “Ideation”) waiting to be analyzed, broken down and prioritized for development.? If the create phase has capacity for new work, then this is a situation where more capacity at the product owner, manager, design or business analyst roles may be needed to flow work more quickly to be available to be pulled into create.??
2. Stuck in Create: When work is all prepped for development and waiting to enter their queue, you likely have a case of demand overproduction. This is a form of waste, as the prepped work will not be delivered within a meaningful market time frame. These great ideas will lose value and grow stale.? Some value streams can have many months (or even years!) of work waiting to start development, limited by the capacity of the developers on the value stream.
3. Stuck in Release: This is where developers have completed their work, but the feature is now waiting for the release process, e.g.? UAT, release certification, or performance and security testing.?

If work is starting but getting blocked downstream in either the Create or Release phases, it’s important to identify where and why. You should not start new work that will likely suffer the same fate.?

If there is a traffic jam downtown, we don’t want to let more cars on that will just add to this pileup. If there is an abundance of buildup of feature work, consider? working on other types of work that can be done more quickly and independently (e.g. an independent enhancement story which improves the UX experience and can deliver value quickly or a defect) while you’re working on addressing your constraint.?

Defects

When the majority of your WIP is defect work, it’s advised to investigate the quality issue generating these issues and taking capacity away from value-generating work like features. Compare the distribution of defects in your flow to good practices for your product life cycle stage. If it’s out of whack, you’ll want to course-correct.??

Risks and Debts?

If debt and risk items dominate your WIP, it’s likely these work items are never given priority to be completed in a timely fashion. This type of ongoing neglect will unduly expose your business to security and compliance risks and cripple your innovation roadmap.?

In closing, in order to understand and better utilize your Flow Metrics, it is necessary to ensure your underlying system is stable and if not, work to identify the contributing factors that are creating this instability.

I will leave you win a final quote from Deming from his classic book "Out of Crisis".

“People generally want to do the right thing, but in a large organization, they frequently don't really understand what is the right thing.”