The Modern AnalyticsReturn on Intelligence
How it runs

You always know what happens next.

Scope in writing before anything starts, a named deliverable at every phase, and a review against the baseline ninety days later whether or not I am still engaged.

01 / The approach

The goal comes first. The technology is whatever serves it.

Every engagement starts from a goal the business has already committed to, and works backwards. Not from a platform, not from a vendor shortlist, and not from a technology anybody is excited about.

That order matters, because it is the only one that produces a number a finance director can still find twelve months later. It also means I will sometimes tell you the answer is a licence you already own, a process that should be retired, or nothing at all.

I have built and shipped platforms, so I know what building actually costs and where a specification leaks. I do not sell it. Development is not on my services list and it is not going on it.

That is deliberate, and it is the part worth reading twice. The moment building is on my price list, every recommendation to build becomes suspect, including the correct ones. When the answer genuinely is build, I help you scope it, choose who does it and hold them to it. The invoice for the build goes to somebody else.

The second thing worth reading twice is that I do not run projects. A project ends. The decisions it was meant to serve do not, which is why so much delivered work quietly stops being used in the year after closure. Everything I do is scoped to leave a product behind: a named owner still in post, a measured baseline, and a review that happens whether or not I am in the room. The argument for that is here.

02 / The shapes

Three ways this runs, depending on what you actually need.

Most people arrive assuming it is the first one. It often is not.

01

The 7D Diagnostic

You need one decision made properly, and made now.

Fixed scope, fixed fee, the five phases below. A portfolio scored, a platform reviewed, a business case tested, a roadmap sequenced. It ends with a decision, a measured baseline and a named owner, which is what lets the work continue as a product after I have gone.

Runs for
Four to twelve weeks
You get
A decision you can defend
02

Ongoing

The decisions keep coming and nobody senior owns them.

An agreed number of days a month against a named remit. Governance, value cases, portfolio sequencing and independent challenge in the room where it is decided. The senior data and AI judgement on your leadership team without carrying the headcount.

Runs for
Six month minimum, thirty days notice
You get
Judgement on tap, not a body
03

Fractional and advisory

You are building something technical and the senior call is missing.

Part time technical leadership for founders and small leadership teams. Architecture and technology choices, technical due diligence, hiring and assessing engineers, vendor selection, and the calls that are expensive to reverse later. I will take a named advisory seat and be accountable in it.

Not a body on your build team, and not equity in exchange for delivery. Advisory, on a retainer, with the independence that keeps it worth having.

Runs for
One to four days a month, rolling
You get
A senior technical second opinion that stays

The phases below describe a diagnostic. Ongoing and advisory work starts with a short version of the same thing, because a remit without a baseline is just attendance.

03 / The phases

Five phases, each with something you keep.

No open ended discovery, and no invoice you did not see coming.

01

Scope

A written statement of work before anything begins: what is in, what is out, who is involved, and what you receive.

You getStatement of work, fixed fee
02

Evidence

Interviews, systems review and data assessment. I talk to the people doing the work, not only the people describing it.

You getInterview findings, readiness heatmap
03

Assessment

Every candidate scored against the seven gates, with the ones worth pausing marked as clearly as the ones worth starting.

You getScored portfolio, value cases
04

Decision

A working session with your leadership team, not a document drop. I present, take the challenge, and adjust in the room.

You getLeadership pack, agreed roadmap
05

Review

Ninety days on, we check the baseline against reality, whether or not I am still engaged.

You getValue review against baseline

04 / Your side of it

What this costs you in time, not money.

The question nobody asks on the call and everybody worries about afterwards. Here is the honest answer, because an engagement that quietly eats your team is a failed engagement even if the recommendation was right.

PhaseWhat I need from youYour time
ScopeOne conversation to agree what is in and what is out45 minone person
EvidenceSix to ten interviews with the people doing the work, and a walkthrough of the systems8 hrsspread across the organisation
AssessmentNothing scheduled. Occasional questions by message when something does not add up2 hrsad hoc, no meetings
DecisionOne working session with the leadership team who will act on it90 minleadership team
ReviewOne session ninety days later against the baseline we agreed60 minwhoever owns the outcome
Total across the whole engagement~2 daysof your organisation

If an engagement starts drifting past this, that is a signal something is wrong with the scope, and I will raise it rather than bill it.

05 / The fit

I would rather lose the work than take the wrong work.

A bad fit costs you a quarter and costs me a reference. Both of us are better off finding out in the first fifteen minutes.

Worth a conversation

  • You have real data volume and real margin pressure
  • You have an ERP and some form of BI, and data you cannot fully use
  • Your IT team has no data scientists in it
  • Nobody currently owns data and AI strategy as their actual job
  • You have more ideas than capacity and no agreed way to sequence them
  • You would rather hear an honest no than a comfortable yes

Probably not me

  • You want a co founder taking equity instead of a fee
  • You want a body shop rather than someone accountable for the outcome
  • You have already decided what to buy and want a business case written backwards to justify it
  • You need a team on site full time from next week
  • You need the answer to be yes before the work starts

06 / The terms

The commercial terms, in plain words.

These are the same on every engagement. Nothing here is a negotiating position, and none of it is buried in a schedule at the back.

  1. Mutual NDA as standard. Signed before the first working session, not after.
  2. Fixed scope and fixed fee. Agreed in writing before anything starts, and it does not move without your written agreement.
  3. You own the output. The scoring, the value cases, the models and the working files. Not licensed to you. Yours.
  4. Nothing auto renews. Ongoing work ends when you stop asking for it, with thirty days notice either way.
  5. The ninety day review is included. It happens whether or not you renew, and whether or not the news is good.
  6. Your data stays yours. I do not take copies off site, and I work in your systems where practical.
  7. I declare the conflict before you have to ask. If my assessment points at a build I could profitably deliver, you hear that from me in the room, and you are free to take the work elsewhere.

07 / Questions

The ones I am always asked.

How is this different from hiring a consultancy?

A consultancy sends a team, produces a recommendation and leaves. I hold the role part time and stay accountable for what happens to that recommendation, including presenting it to your leadership team and answering for it at review.

What does it cost?

Assessments are fixed scope and fixed fee, agreed in writing before anything starts. Ongoing work is monthly against an agreed number of days, on a six month minimum term.

I quote against the specific engagement rather than publishing a rate card, because the right scope varies considerably. You will have a written figure before any work begins, and it will not move without your agreement.

Do you build things?

Yes. I have built and shipped platforms, and where building is genuinely the right answer I can do it or manage whoever does. What I will not do is let the build decision be made by whoever benefits from building, which is why the assessment is always a separate piece of work.

We already have a Head of Data. Is this still relevant?

Often not, and I will say so. If you have that capability, what you usually need is support for the person doing it rather than another voice above them.

Where it does help is at leadership level, on governance, value cases and independent challenge. That is a much narrower engagement.

How quickly can you start?

I work with a small number of clients at a time, so there is sometimes a wait for ongoing capacity. A briefing or an assessment can usually start within two to three weeks.

What happens to our data?

I work under a mutual NDA as standard. I do not take copies of client data off site, and I use your systems where practical. Full details go in the statement of work before anything starts.

Still sounds like a fit?

No deck and no pitch. If I am not the right person for this I will say so on the call and point you at who is.