Making Work Visible
Authors: Dominica DeGrandis
Publisher: IT Revolution; МИФ (перевод)
Length: 2017
Dominica DeGrandis on exposing hidden work, reducing overload, improving flow, and making operational decisions from visible evidence.
Original
TranslationBrief
The essential idea
Making Work Visible identifies five major time thieves: excessive work in progress, dependencies, unplanned work, conflicting priorities, and abandoned work. They create hidden queues and context switching, make commitments unreliable, and allow urgent requests to displace valuable work without a deliberate decision.
A board should mature from a simple To Do/Doing/Done view into a model with work classes, explicit policies, WIP limits, owners, dependencies, commitment points, service expectations, and flow analytics. The objective is not visual management theater but a pull system in which aging, lead time, cycle time, and queue growth drive concrete decisions.
Decision lens
Key takeaways
Visibility converts reactive status chasing into systematic improvement of the work system.
Large WIP increases queues and cycle time even when every individual appears busy.
Dependencies and unplanned work must be represented explicitly or delivery dates remain accidental.
Feature work, defects, maintenance or risk work, and expedite items need visible classes and policies.
Limiting unfinished work improves predictability more than accelerating isolated individuals.
A useful board evolves through visualization, constraints, measurement, and correction rather than being designed perfectly upfront.
Workplace experiment
Apply it at work
- 1
Visualize the current end-to-end flow as it actually operates, including waiting and unplanned work.
- 2
Place a WIP limit on the stage where the queue grows fastest and enforce it for one review cycle.
- 3
Create a separate expedite lane with a strict limit and an explicit entry policy.
- 4
Review aging work weekly and cancel items whose value has disappeared instead of finishing them by inertia.
- 5
Add cycle-time and lead-time trends only after the team's status definitions are consistent.
Choose one action, define the observable effect, and keep the first test small enough to reverse.
Evidence
Sources and further reading
Additional sources
Channel, aggregator, and commentary links confirm the work; they are not the primary source.