Case studies

What the work looks like in practice

Six engagements, described by problem rather than by sector. Client names are withheld where confidentiality applies; sector and headcount are accurate.

Logistics · 22 staff

Cutting exception-handling time by a third

Challenge: Delivery exceptions were handled differently by each of the four people who covered the desk. Nobody could state how many occurred last quarter, so recurring causes were never addressed.
Approach: Two weeks of observation produced a five-category taxonomy covering eighty per cent of real events. We then built a one-page reporting template and a weekly triage review with named owners and an escalation threshold.
What made it stick: The template was cut from two pages to one after the first week because people stopped filling it in. The shorter version survived.

Result: resolution time down 31% in eight weeks, plus the first reliable monthly exception report — produced from data the team already had.

Distribution · 9 staff

Onboarding a new starter without a handover call

Challenge: Every new hire spent their first month shadowing the operations lead, who had no spare capacity to begin with. Hiring more people made the bottleneck worse, not better.
Approach: Documented the twelve recurring weekly tasks as short SOPs, each under two pages. A new starter then followed them with no assistance while we logged every point of confusion.
What made it stick: The SOP set was revised twice from the new starter's own notes. The second revision removed a step that had been performed for years without purpose.

Result: time-to-independent from 4 weeks to 9 days, and the operations lead recovered roughly a day a week.

Fulfilment · 14 staff

Making supplier terms visible before renewal

Challenge: Supplier terms lived in individual inboxes. Two renewals had auto-rolled at higher rates because the notice window was missed entirely.
Approach: Built a single register covering 46 suppliers: contract terms, notice windows, renewal dates and a named internal owner for each. Added a monthly review where the next 60 days of renewals are read out.
What made it stick: The register lives in a tool the team already opened daily rather than a new system nobody would visit.

Result: zero missed notice windows in the following twelve months, and two renegotiations that paid for the engagement several times over.

Wholesale · 31 staff

Finding the capacity nobody knew was missing

Challenge: The team was consistently behind, and the assumption was that headcount was too low. A hiring round was planned.
Approach: Two weeks of time sampling across the operations team showed 22% of hours going into re-keying data between two systems that could have been integrated, and a further 9% into status meetings that existed to compensate for that gap.
What made it stick: The integration was scoped and quoted before any hiring decision was made, which changed the order of investment.

Result: hiring deferred by two quarters, and roughly 1.3 FTE of capacity recovered without a single new hire.

Industrial services · 26 staff

Getting job records completed on site

Challenge: Engineers completed job sheets at the end of the week from memory. Warranty claims were routinely lost because the photographic evidence had never been captured at the time.
Approach: Redesigned the job sheet around four mandatory fields that could be completed on a phone in under ninety seconds, then piloted with two volunteers before rolling out.
What made it stick: The pilot engineers shaped the field list themselves. The final version had fewer fields than the original proposal, which is why it was used.

Result: warranty claim recovery up from 61% to 94% over two quarters, with no additional administration time.

Logistics · 18 staff

One peak plan instead of twelve remembered lessons

Challenge: Each November the same failures repeated: temporary staff untrained, carrier cutoffs missed, and a backlog that took three weeks to clear.
Approach: Ran a structured retrospective on the previous peak, converted the findings into a written plan with owners and dates, and set three checkpoint reviews through the peak period.
What made it stick: The plan was owned by the operations lead rather than by us, and the checkpoint reviews were already in the team calendar.

Result: peak backlog cleared in four days instead of three weeks, and the plan was reused the following year with two amendments.

Have a comparable problem?

Most of these started with a half-day diagnostic session and one agreed deliverable.

Start a conversation