Are you interested in getting started with pull systems?
Pull systems are a deceptively simple concept. Benefits of pull systems are easy to understand. They are more predictable and usually faster than push systems. But what do we mean by “push” vs “pull” and what do we mean by “fast”?
There are 3 primary capabilities that pull systems impact directly: lead time distribution (time-to-market), delivery frequency (throughput), and replenishment frequency (true replenishment only being possible with pull). Indirectly, pull systems impact other potential measures of success: KPIs or health indicators, and most notably various aspects of quality.
In a push system, there are no explicit limits to the work in progress in the system. Policies for how the work enters the system are not explicit. The difference between work already in progress and work yet to be started is often vague and opaque. Many things get started for many different reasons that aren’t clear. When work gets blocked, there is a tendency to start new work rather than addressing the systemic causes of blocked work. Work items wait in queues often long enough to be forgotten because they are no longer worth completing. People doing the work are often disrupted by new, urgent requests. They are constantly context-switching (commonly mis-labeled as “multitasking”). Capability of the system is continuously under strain and often overburdened. Overburdening is compensated for by people working at an unsustainable pace, such as too much overtime, weekends and pizza meals.
In a pull system, demand is balanced to capability with work-in-progress limits, explicit pull criteria policies and classes of service. Work enters the system based on realistic and rational assessment of risk. The system is replenished frequently because the system is capable of delivering frequently, allowing for constant, predictable flow of work-in-progress and therefore the frequent creation of available capacity for replenishment. People know the value of what they are working on and they are able to focus on finishing work that meets explicit quality standards. They take more pride in their work and they work at a sustainable pace because the system is never overburdened. Such systems are capable of delivering faster, more predictable, higher-quality services to customers.
Work in Progress
Its important to be clear about what we mean by “work-in-progress”. The best way to understand this is that everything you have promised to deliver and have not yet delivered to your customers is the work in progress of your service delivery system. This clarity, and the clarity of the boundaries of the system, are important for understanding, measuring and improving the capabilities of the system. In the case of system lead time, the clock starts ticking on a promise the moment that promise has been made and stops only when the promise has been fulfilled. Even if the actual work hasn’t started, it’s still work-in-progress. When work pauses on a work item (a promise to a customer), the lead time clock keeps on ticking. There is no “pause” button for lead time. The collective age of all of the promises in your system is the current load on the system, regardless of how many of them are actively being worked on and how many of them are in a “waiting” state.
How many projects do you currently have in progress? How many features? How many change requests? How many fixes? Of these, how many are currently in a waiting state? If the answers to these questions are not easy to find, then you most likely have a push system and your delivery capability is unpredictable, unreliable and untrustworthy. Your customers are likely to be dissatisfied, your executives are starting to get nervous about the long-term survivability of the business and they are looking to you for answers and results. As long as there is a push system in place, chaos will persist.
Why Aren’t Pull Systems More Common?
Pull systems are difficult because they require a new way of thinking about work and how work gets done. It requires a commitment to the long-term survivability of a business. It requires that people are not concerned only with their individual achievements and career advancement. It requires leadership maturity. That is, leaders who are focused on delivering value to customers (enabled by pull) rather than on the latest management fad and the utilization and performance of people (characteristics of push). None of these are easy to come by.
The business case for pull is simple: You can start everything (push) and finish nothing on time and with poor quality…Or, you can start a few things (pull) and finish everything on time and with high quality. Which one of these is better for your business?
You can find related content in Jerry Doucett’s article: https://berteig.com/kanban/scrum-mastery-and-kanban-part-2-limiting-wip/