Case studies

Four engagements, written from the record

Every figure below traces to a delivery artifact — an acceptance document, a commit range, or a reconciliation the client ran themselves.

There are fewer numbers here than most consulting sites carry, and that's deliberate. We publish what we can show evidence for, and nothing that depends on a client remembering favorably.

Case 01

PE-backed roll-up · 28 locations · 6 systems and Sage

Platform build → retainer

Twenty-eight locations still run their own way. They now report one way, every day.

Situation

Six systems plus Sage running as nineteen instances across twenty-eight locations. Corporate KPIs were pulled by hand once a month, and the CFO opened eight systems to answer one question. The operator who owned cash revenue built his number in a spreadsheet with no way to check it.

What we did

Three contracted phases in their own tenant, each scoped and accepted in writing before the next was priced. Every source inventoried, metrics defined in the operators' own words, then a governed model built out to a coverage figure written into the contract.

What changed

The operator reconciled the platform against his own hand-built April figure, and the arguing stopped. Operational numbers now arrive daily instead of monthly. Onboarding a newly acquired company's system fell from forty-four days to thirteen. The client chose to continue on retainer after the third phase closed.

Case 02

Travel-data SaaS · holdco portfolio company

Strategy engagement

They asked for a report on what was possible. They got a working proof instead.

Situation

A thirty-year-old platform holding decades of client data in forty-nine separate databases, which had never been aggregated across clients. Leadership knew a rebuild was needed and had known for two years, with no agreed sequence and no way to price it.

What we did

Six discovery sessions with the people who produce the numbers, finished by day ten — a week ahead of plan. Rather than write up what an aggregated view might look like, we built one against live data so the recommendation rested on something demonstrated.

What changed

A two-year-old rebuild question became a sequenced, priced roadmap the client could hand to anyone. We narrowed our own proposal before signing and said which parts could wait, which cost us scope and bought the recommendation its credibility.

Case 03

Medical-imaging lessor · 57-unit fleet

Diagnostic engagement

The useful finding was how much they didn't know they were missing.

Situation

A lessor operating a fifty-seven-unit imaging fleet, with utilization and service history spread across systems that had grown up one contract at a time. The team suspected gaps but couldn't size them.

What we did

A structured read of the data estate against the decisions the business actually makes — what a unit earns, when it needs service, which contracts are worth renewing — and where the current systems could and couldn't answer.

What changed

Gaps the team had suspected were sized, ranked and tied to the decisions they affect. For a diagnostic, a prioritized list of what to fix first is the whole deliverable.

Case 04

Auto service contracts · claims operation

Strategy engagement

A claims team was judged on how fast it answered the phone, while the cost of each claim kept climbing.

Situation

Claims leadership ran on a monthly report of 46 KPIs, assembled by hand in Excel over roughly 22 steps. Twenty-two measured the phones; five measured claims. Calls were answered in 44 seconds while cost per claim rose 60% over three years, and nobody could see where the handling effort actually went.

What we did

A five-month strategy engagement, delivered with a partner firm. We rebuilt the claims data from a backup of the claims system into a single model, then followed the handling effort by claim size, line items, touches and denial reason. We handed over the dashboard, a proposal to route claims by complexity and measure cost per claim instead of answer speed, and a twelve-month roadmap.

What changed

The diagnosis pointed at claim complexity, not answer speed. Claims over $1,000 carried nearly three times the line items and more than twice the touches, and about 45,000 denials were for lack of coverage, each worked by hand. Nobody has measured whether the roadmap was carried out, so we claim the diagnosis, not a saving.

What you'll notice they have in common

The client checked our work

In every engagement the deciding moment was someone on their side reconciling our output against a number they already trusted.

We cut our own scope

Twice on this page we narrowed or absorbed work rather than bill it. It's the cheapest way to be believed later.

Nothing depended on us staying

Each platform runs in the client's own tenant. One continued on retainer — by choice, not because the build required it.

Every one of these started with somebody unable to defend a number.

Thirty minutes, no deck. Bring yours and we'll tell you what it would take to fix it.