Where we help

Six situations, and what happens in each.

Engagements here start with a situation rather than a specification. If one of these sounds like the inside of your week, the first conversation will be useful whether or not it goes further.

01

I can’t tell if what I’m being told is true

Reporting has decoupled from reality, and everyone who could confirm it has a stake in the answer.

“I get a status report every two weeks. Each team reports accurately on its own piece, and somehow the program is four months late. I cannot get a straight answer on what is actually blocking it.”

Programs rarely fail in one moment. A workstream goes amber for a quarter and amber stops meaning anything. We read the artifacts, talk to the people doing the work, separate the real blockers from the reported ones, and hand over an honest re-baseline — including “stop” if that is the answer.

When the status report stops meaning anything →

What you get

ShapeFixed fee, 4–6 weeks
Ends inA written re-baseline you own
CoversReal blockers, dependencies, date
02

I’m about to make a decision I can’t take back

Build or buy, continue or stop, stay or re-compete — and everyone advising you benefits from one of the answers.

“It is seven figures either way and I get one go at it. The integrator says continue. The vendor says buy. Neither of them is wrong exactly, but neither is disinterested.”

These calls get made once and lived with for years. There is no software to license here and nothing gained if the answer is “build”, which is the only reason this view is worth more than the one already in the room. The output is a recommendation with the reasoning attached, so you can defend it to a board.

The second market is where the architecture decision gets made →

Typical questions

Build or buyAgainst your actual workflows
Continue or stopWith a real cost to finish
VendorRetain or re-compete
03

Nobody here has run something this size before

Good engineers, a capable vendor, and no one who has seen what goes wrong at month seven.

“We are not short of people. We are short of someone who has delivered a program this big and knows which conversations to have before they become problems.”

Fractional technology leadership — two or three days a week for a few months, running the program the way a permanent technology director would. Governance, vendor management, executive reporting and the decisions that keep stalling. For organizations that need that seniority on one program and cannot justify it on the payroll.

What it looks like

Cadence2–3 days a week
DurationA few months, defined up front
OwnsGovernance and reporting
04

The vendor is running the show

Scope, quality and timeline have moved to the other side of the table.

“Our systems integrator owns the delivery. Every change is a change order and I cannot tell whether we are being managed or handled.”

This practice does not replace the integrator. It restores the client side of the table — governance that produces decisions rather than status, scope and quality control that hold, and outcomes rather than hours. Two years were spent on exactly that side of the relationship, owning a 300-application estate across more than seventy countries and deciding what got bought, built and switched off.

Worth knowing

IndependenceNo licences, no reseller margin
OutcomeCan conclude “keep them”
BasisBuyer-side portfolio experience
05

Our systems disagree with each other

The interfaces work. The data still cannot be trusted.

“The integration went live months ago and three people in finance are still reconciling spreadsheets by hand.”

That is a governance problem wearing an engineering costume. When two systems hold a different value for the same customer, somebody has to have decided in advance which one wins, under what conditions, and who is allowed to change the rule. Without that, an integration propagates the disagreement at machine speed. The master data work is not a line item on the project — it is the project.

Your integration project will not fail on the interfaces →

Three questions

SourceWhich system is right?
ConflictWhat happens on disagreement?
OwnerWho can change the rule?
06

It launched, and now nobody owns it

The build team left, the warranty period ended, and the platform has a phone number instead of an owner.

“It works, mostly. But nobody is watching it, nobody has patched it, and I am not confident we could restore it.”

Managed cloud operations — monitoring and alerting, patching, backup and recovery, capacity and cost management against a defined availability commitment. Or, if you would rather run it yourself, the handover work to make that genuinely possible: documentation, runbooks and knowledge transfer that leave you able to operate it or put it back out to tender.

Either way

Run itManaged, with an availability commitment
Or hand overDocs, runbooks, knowledge transfer
NeverA dependency you cannot exit

None of these quite fit?

Describe it in your own words. If it is not something this practice should take on, you will be told that in the first conversation.