·

N° 07 · AI FINOPS · ROI AUDIT & REALIZATION PLAN

AI ROI Audit

Measure what AI returns, and close the gap. What AI has actually returned against the business case — baseline, value attribution, Hard ROI kept apart from Strategic Upside — then the plan to realise the value still missing. For the CFO, the COO, the head of AI or an operating partner.

PRICE

€12,500

DELIVERY

4–12 weeks

REVIEW

Signed by a partner

§ 01 · The problem

"How much has our AI actually earned this quarter?"

Most enterprises have AI initiatives running. Few can answer that question with a defensible number. Productivity gains stay anecdotal. Revenue uplift is conflated with market trends. Cost reduction floats in a sea of confounding variables.

This Signature gives you a defensible, board-ready number — workflow by workflow — and the plan to close the gap between what AI was meant to deliver and what it does.

§ 02 · What you get

A defensible number, and the plan to close the gap.

Five deliverables built around your two or three highest-value AI workflows, designed to extend to the rest of the portfolio.

DELIVERABLE 01

ROI audit report

Our five-step method applied to two or three pilot workflows: mapping, outcome metrics, pre-AI baseline, post-AI measurement, value attribution. Sector benchmarks included.

DELIVERABLE 02

Dashboard template

Excel and Power BI templates ready to track outcome metrics. Pre-AI baseline, post-AI measurement, value attribution.

DELIVERABLE 03

Board-ready memo

8-12 page synthesis memo, ready to present in board or Comex. Quantified value created, methodology defensibility, next quarter focus.

DELIVERABLE 04

Realization plan

The levers that close the gap between the business case and the measured return — adoption, scope, pricing, process — each with an owner, a target and a quarterly review.

DELIVERABLE 05

Extension playbook

A step-by-step method to apply the audit to your remaining AI workflows, designed for your team to run every quarter.

Deliverable excerptWhat you actually receive.
AI ROI AuditIllustrative · anonymised engagement
Business case versus actualsValue = users × active usage × value per use · cumulative €m at Q8
Axis truncated at €3m
4.5
Business case
−0.2
Timing
Go-live one quarter late
−0.2
Scope
800 of 1,000 users
−0.6
Adoption
45% active usage vs 70%
−0.1
Unit value
Gain per use −5%
3.4
Actual
Weekly active usage (%)
0%20%40%60%80%W1W4W8W12W16
Business case
Actual
Peak in week 3, then drop-off: usage settles at 45% from week 8.
More than half of the €1.1m gap comes from usage dropping off.
Usage profile consistent with Dillon et al. (2025), Microsoft 365 Copilot field study.

§ 03 · Process & timeline

Four to twelve weeks. Four sessions. A number and a plan.

From kickoff to the board-ready memo and realization plan. The duration depends on the number of workflows and the data available.

WEEK 1

Kickoff

90-min session with partner. Scoping of 2-3 pilot workflows, data access requirements, stakeholder mapping.

WEEKS 2–4

Workflow 1 immersion

2h working session + offline analysis. Mapping, outcome metrics definition, baseline reconstruction. First framework draft.

WEEKS 4–7

Workflows 2–3 immersion

2h working session covering both workflows. Cross-workflow learning, framework refinement, dashboard prototyping.

WEEKS 7–10

Consolidation

Final audit report, board-ready memo, dashboard, realization plan and extension playbook finalised.

FINAL WEEK

Final restitution

Final session with the partner: walkthrough of all deliverables, Q&A, owners of the realization plan confirmed.

§ 04 · Frequently asked

Six honest answers to common questions.

Why four to twelve weeks?

The bottleneck is not our work but the time to gather baseline data and run meaningful measurements. Two or three workflows with clean data take four weeks; a wider scope or thin data takes longer. We would rather take the time than deliver a number you cannot defend.

Does it work if we have many AI initiatives?

Yes — that is the design. The audit applies full depth to two or three pilot workflows; the extension playbook then lets your team replicate the method across the rest of the portfolio.

What data do you need from us?

For each workflow: pre-AI operational metrics (cycle time, error rate, output volume), team time logs if available, post-AI observation data, and access to 2-3 stakeholders who run the workflow. We don't need access to sensitive customer data or financials beyond what's strictly necessary.

Who delivers the engagement?

One partner, from kickoff to restitution. No junior handover, no team rotation.

Can we go quarterly after the initial engagement?

Yes. The extension playbook lets your team run quarterly measurement cycles on its own. If you want the measurement installed rather than handed over, we scope it as a bespoke engagement.

What if the framework shows our AI isn't creating value?

That's a valuable answer too. We've had engagements where the conclusion was "stop this initiative, redeploy to that one." The framework is honest — it won't manufacture positive ROI where there isn't one. That's the whole point.

The board’s questionWhat has your AI earned this quarter?
37%of companies see an AI effect on EBIT
63%cannot show one
6%draw more than 5% of EBIT from AI
Source: McKinsey, The state of AI in 2026: On the road to ROI, August 2026 — survey of 1,719 respondents in 97 countries.

§ Ready when you are

Measure what AI returns, and close the gap.

€12,500, four to twelve weeks, signed by a partner. The number your board needs, and the plan to improve it.