Stage four · orchestrate

AI implementation tells you whetherthe technology works. This tells youwhether the investment works.

£5,000 a month, twelve-month initial term. We stay commercially involved while your technology partner builds — and afterwards, when the question changes from “is it live?” to “did it pay?”

The problem it solves

“It went live. Nobody can tell us what it returned.”

A build can succeed technically and still fail commercially. The partner delivers what was asked for. It passes testing. It goes live on time.

And then, twelve months later, nobody can answer the only question the board actually cares about. Usually for two reasons. Nobody wrote down what performance looked like beforehand. And somewhere during the build, small technical decisions quietly changed what the thing does.

Neither of those is anyone’s job to prevent. That is what this is for.

How value leaks

The baseline was never taken. Once the new system is live you cannot go back and measure the old one.

The design drifted. Reasonable technical choices, made one at a time, add up to something different from what was approved.

The measures were never built in. Nobody asked for them during the build, so the data to prove the return does not exist.

Nobody owned the question. The partner’s job ended at acceptance. Finance moved on. The board stopped asking.

Two phases, one engagement

During the build, and after it.

It is one product because the commercial thinking should not disappear while the technology is being built.

Phase A · during the build

Is what gets built still what the board approved?

We do not manage the technical programme. Your partner remains fully responsible for delivery. We check commercial integrity at four defined points.

Phase B · after go-live

Is what we built delivering what was approved?

The question changes. We measure what actually happened against the case, find where value is not showing up, and bring decisions to the board.

Phase A · four commercial checkpoints

Four gates. Not open-ended project management.

Defined points, not a standing seat in every meeting. That is what keeps this affordable and keeps us out of your partner’s way.

Gate 1

Mobilisation

Has your partner correctly understood the new process, the operating model, the business case, the edge cases and the measures?

Gate 2

Design integrity

Are technical decisions quietly changing the commercial outcome the board approved?

Gate 3

Pre-go-live readiness

Are the processes, controls and measurement mechanisms actually in place before it goes live?

Gate 4

Value baseline

Can performance now be measured against the case? This is the gate that makes everything afterwards possible.

Phase B · the Value Tracker

Three kinds of evidence, joined up.

Most organisations have the first. Some have the second. Almost nobody connects either to the third. The point is not another dashboard — it is the chain between them.

AI performance

Usage, activity, adoption, and how the models or agents are performing.

Operational performance

Cycle time, throughput, exceptions, how often a person still has to step in, and what the process now produces.

Economic performance

Cost taken out, capacity released, revenue and margin effect, cost avoided — against the case that was approved.

AI activity → process performance → economic outcome. Measured against the baseline taken before anything was built.

We own
The measurement method
The commercial data model
The measures and the economic assumptions
What the numbers mean commercially
Your technology partner owns
The technical integrations
Making the source data available
Any instrumentation needed to expose it
Whether the technology is working technically

We are not a software company. The Value Tracker is a measurement method and a commercial data model, set up wherever practical inside the systems you already have — usually your Microsoft estate, alongside the line-of-business systems the process actually runs on. Your technology partner exposes the data; we decide what it means commercially.

Why this stage exists at all

It is the only stage that reaches “measured”.

Everything we tell you carries an evidence status. Three of the four can be reached before anything is built. The fourth cannot.

Research-backedA pattern we see in your sector.
Practitioner-validatedConfirmed by people who run this work at comparable firms.
Client-validatedConfirmed by your own people, during the Accelerator.
MeasuredCounted against a baseline. Only here.

Without this stage, the best you will ever have is a well-argued expectation.

How it runs

A closed commercial loop — and it never runs through the build.

This is not a delivery lifecycle. It is a loop that starts at the decision your board made and comes back to it.

Phase A · protecting what was approved

01Board-approved valueThe case that was signed off, with a baseline attached.
02Commercial specificationWhat your partner builds to, agreed commercially first.
03Implementation assuranceFour gates. Is the approved change surviving the build?
04Baseline at go-livePerformance can now be measured against the case.

Phase B · proving what it returned

05Actual against baselineThe three kinds of evidence, joined up.
06Variance and value at riskWhere the return is not showing up, early enough to act.
07Commercial interventionScale it, fix it, optimise it or stop it. A decision, not a report.
08Realised valueMeasured, not claimed. And it points at the next opportunity.

Back to the start — the evidence identifies the next thing worth doing, and the loop runs again.

Technical implementation and operation sits outside this loop — assured at four gates, measured afterwards, never delivered by us.
What the board actually receives

One quarterly session where the numbers are on the table.

An AI portfolio view

Every material initiative in one place — owner, business case, and what it has actually returned so far.

Variance analysis

Expected against actual, in economic terms rather than usage statistics.

Value at risk

Where the return is not materialising, surfaced while there is still time to do something about it.

A quarterly executive review

A proper decision session with the Chief Executive, Finance Director and Chief Operating Officer — not a status update.

Portfolio decisions

Every initiative classified: scale, optimise, correct, stop, or look at next.

An annual review

What the portfolio returned over the year, what was learned, and where next year’s money should go.

What it is not

This is not a second project manager.

Not technical programme or project management. Not development or sprint management. Not solution architecture. Not tenant administration. Not technical testing or security work. Not support or managed services.

Nor does £5,000 a month buy unlimited access to consultants. It buys four defined gates, a measurement model, a quarterly executive session and portfolio decisions.

Your technology partner keeps delivery, support and the ongoing revenue. We are not competing for it.

Who is accountable for what

Formalus. The integrity of the commercial design, the measurement, the evidence, spotting where value is slipping, and steering the decisions.

Your management team. The business outcome. Acting on the evidence, resourcing the change, running the business.

Your technology partner. Technical delivery and operation.

We measure and steer. We do not promise you a number — nobody honestly can.

The baseline has to be taken before the build.

A baseline taken beforehand is evidence. One reconstructed afterwards is an argument. That is why this normally starts at the close of the Accelerator, not after go-live.

Book a free assessment