Control Work in Progress to Cut the Backlog and Finish Projects Faster

It seems like common sense: start sooner, and you’ll finish sooner. The problem is that everyone starts sooner! Increasing the volume of work in progress only heightens confusion and conflict—and reduces real productivity. Having a lot of work does keep people busy. However, while everyone is busy, the true picture of the overall project is obscured until deadlines approach, when the failure to complete the right tasks becomes all too visible.

The more work you have stacked up, the more opportunities you have to multitask and to work on the wrong thing (out of sequence, too soon). Which task should you prioritize? How many tasks do you have on hand? How many tasks should you have on hand?

The extra work complicates priority setting. If you have ten tasks on your desk and no guidance on priorities, what criteria do you use to choose? Do you pick the task you’re really interested in, or the one you don’t understand that will require the most concentration? Do you pick the one with the closest deadline or the one you can finish the fastest?

Most of us will automatically choose the task that interests us most. Unfortunately for those of us who get bored easily, this has as much impact on projects as delays do; if the wrong task is chosen, the wrong work gets introduced into the system. Visibility can help with this issue—if you can see what the other team members need from you to accomplish the ultimate goal, then you can prioritize accordingly—but without visibility, there’s no way to know which task you should start with. The right work lags, and the wrong work moves ahead. Once again, the next link in the chain will move on to a different task instead of waiting to receive the completed one that they need—and the delay cascades down the chain, growing exponentially again. The project team gets loaded up with work that shouldn’t be in the system, and then there’s expediting and sorting to be done in a project or process that shouldn’t need it at all. Then, we’re back to multitasking, and the process begins all over again. It’s a vicious cycle, one that creates extra work for team members and additional management overhead to address it.

Just as adding more cars to the highways doesn’t mean everyone gets there faster, putting more work into play will never accelerate progress. Starting more work does not equate to finishing more work. Starting more work means finishing less work.

Secondly, without control, new tasks are added to the system prematurely. If prerequisite tasks, such as design, or external requirements, such as component completions, are not completed before initiation, resources are misdirected to tasks that are impossible to complete – or that are completed under incorrect assumptions, requiring rework further down the line. This again creates additional unplanned work and cost, and increases project duration.

To control work in process, managers establish and enforce pre-release criteria to align work releases with the rate of work completion and ensure no work is started that cannot be finished.

What Controlling Work in Progress Looks Like

A Houston-based engineering team created a visual board to see all the work in the system. This board became a mechanism for managing work in progress: the total work in the system, the rate of completion, and upcoming work. The engineers, drafters, and even an engineering intern used the board to guide their decisions and activities, determining the amount of work in the process and taking steps to limit levels and ensure work was not released before all information was available – thus ensuring rapid completion.

They created “ready to release” and “preparation for release” categories of work, from which they monitored inbound work to ensure timely, high-quality (full-kit) release into the system.

Knowing what was in the system also helped with major decisions. When customer requirements changed, they were able to respond quickly, avoiding the typical chaos such changes had created in the past. The team relieved the engineering bottleneck by moving it downstream into the supply chain and assembly, increasing total system output by 21%.

Organizations with maturity in controlling work in progress:

• Have clear criteria for releasing work (nothing is released that cannot be completed)

• Use a clean-start or full-kit process to manage inbound work

• Have established targets for total system WIP

• Managers and leaders meter work (projects) into the system, staggering releases to align with global priorities and resource loads. The full model is in Why Do Projects Succeed or Fail?