Decision Library
Find the constraint before you redesign the work
Operations decisions look tidy in a memo and meet reality on a Monday morning. A new shift pattern, a centralized team or an outsourced function can improve one number while quietly breaking another downstream. A Decision Room compares the redesign with the current way of working, which has known flaws but also known costs.
The options for a process
Keep the current process and fix its worst failure. Redesign the flow. Add capacity with a shift, a hire or overtime. Outsource the function. Automate one step. The current process stays on the Decision because its failure modes are known and measured, while those of the redesign are still guesses.
Two old ideas that still settle arguments
In The Goal (1984), Eliyahu Goldratt argued that a system's output is set by its one tightest constraint, so improving any other step adds cost without adding throughput. Find the step where work waits longest before changing anything else.
Little's Law, proved by John Little in 1961, ties three measures together: average work in progress equals throughput multiplied by the average time each item spends in the system. If a team finishes 50 tickets a day and 300 are open, the average ticket spends about six days in the queue. Cutting the backlog shortens the wait even if nobody works faster.
Evidence from the floor
Measured cycle times, overtime records, defect and rework rates, incident logs, and short interviews with the people who do the work each day. A process map on a slide shows how work was meant to flow. The logs show how it actually flows, and the gap between the two is usually where the decision lives.
How the room argues a redesign
Each option gets a Champion, including the current process. The Craftsman tests whether the new design can be staffed and run, shift by shift, with the people who exist. The Red Teamer runs the bad week: two people out sick, a system outage and a volume spike together. The Diplomat maps whom the change touches, from the night team to the customers who notice the handover.
The Verdict names the leading option, the constraint it addresses, and the measure that would reverse it after launch. The operations lead records the Human Decision.
What to weigh
Effect on the constraint
Whether the option moves the step that limits the whole system's output.
Error cost
The price of mistakes the new arrangement introduces, including rework and refunds.
Staffing reality
Whether the people, skills and schedules exist to run it every week of the year.
Changeover disruption
Customer-visible delay and lost output during the transition.
Resilience
How the system behaves when a key person, site or tool is unavailable.
Where it goes wrong
Hero dependency
The plan works only while one experienced person never takes leave.
Metric gaming
Teams reach the new target by moving work out of sight rather than finishing it.
Handover loss
Context disappears between shifts or teams, and errors appear where work changes hands.
Night coverage for support
A support team receives about 220 tickets a week between ten at night and six in the morning, and the median first reply for those tickets is seven hours. Option one adds a local night shift of four agents at about $70,000 each in total employment cost, or $280,000 a year. Option two contracts a partner team in another time zone for two seats across 56 night hours a week at $34 an hour, about $198,000 a year. Option three keeps the current on-call engineer and adds automatic routing of urgent tickets. The ticket logs show most of the seven hours is time spent waiting for morning, not time spent working. With weights of constraint effect 30, error cost 20, staffing 20, disruption 15 and resilience 15, the partner team leads. The Verdict names what would favor the local shift: refund errors at the handover rising above today's rate during a three-month trial.
Questions
- Can I bring our process map and ticket data?
- Yes. Paste the process description, volumes and timings. The engine works the sums from your figures, and the room compares the options against the constraint those figures reveal.
- Who approves an operating change?
- Your operations leadership. The Verdict lays out the reasoning, and the choice made, with who decided and who approved, is recorded separately.
Further reading
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.