How to buy this, and how to tell when you should not.
The five questions people actually ask before engaging, answered properly rather than in a FAQ accordion. What it costs and why there is no rate card, which model fits which question, and the criteria for deciding this is the wrong purchase.
Most consulting websites make you infer the commercial terms from tone. This page states them. If you read it and conclude the answer is no, that is a good outcome and it cost you eight minutes instead of a discovery call.
1. Which engagement fits your question
Engagement details →Scope is set by the shape of the question, not by a package you pick off a menu. Four models, and the honest criteria for each.
30 minutes, free
Fixed fee, 1–2 weeks
Fixed fee, 2–4 weeks
Retainer, 3-month minimum
The sequencing rule: audit first, always. Roughly half of those conversations end with a recommendation not to hire an advisor at all. Starting the wrong engagement is the most expensive mistake available to either party, and thirty minutes is a cheap way to avoid it.
2. What it costs, and why there is no rate card
The same word covers a two-week pricing module and a three-region expansion diagnostic. Publishing one number for both would be a marketing decision, not an honest one, so the fee is quoted after the audit, when the question has an actual shape.
What is fixed before any work starts:
- A single fee, agreed in writing before the engagement begins
- No hourly billing, so a longer meeting never costs you more
- A named deliverable set, listed in the SOW rather than described in adjectives
- A change-control clause, so scope movement is a conversation with a price attached, not a surprise on an invoice
If the fee is wrong for your situation, you will hear that in the audit rather than in a proposal three weeks later.
What moves the number
- How open the diagnosis isA named constraint costs less to work on than an unnamed one. Most of the cost sits in D1.
- The state of your dataWhere customer, funnel or margin history has to be reconstructed before it can be analysed, that is real work and it gets scoped separately rather than absorbed silently.
- Number of decision-makersA single founder and a five-person exec team with competing views are different engagements, and the second one is mostly alignment.
- Whether execution is includedA diagnosis and a roadmap is one purchase. Driving it through two quarters of cadence is another.
3. This, a large firm, a hire, or nobody
Four ways to buy the same judgment. Three of them are sometimes right.
A large firm
Better when the work needs sustained parallel capacity across many workstreams, a brand-name signature for a board or regulator, or bodies on the ground in several geographies at once.
Worse when the person who wins the work is not the person who does it, and the leverage model means analysts learn your business on your budget.
A permanent hire
Better when the need is continuous rather than a defined question, and the role will still exist in two years. Institutional knowledge compounds in a way an advisor's cannot.
Worse when you need an answer this quarter. Hiring well takes a quarter on its own, and the wrong senior hire is far more expensive than the wrong engagement.
This practice
Better when the question is defined, the answer will change what you do next quarter, and you want the senior person doing the analysis rather than reviewing it.
Worse when you need more than two workstreams running at once. At most two engagements run at a time, which is a real capacity limit and not a positioning line.
Nobody
Better when the constraint is already named and evidenced and what you actually lack is execution capacity, not judgment. Buying analysis you do not need delays the work you do.
Worse when the team has been circling the same argument for two quarters. That is rarely an information problem and almost never resolves internally.
How big you need to be
Clients run from roughly $1M to $1B in revenue: seed-stage companies, scaleups, mid-market businesses, PE portfolio companies and divisions of larger firms.
The revenue band matters far less than the shape of the question. The three tests below decide fit better than any threshold.
The three tests
- A decision is actually pending. Not "we should think about pricing at some point." A choice with a date on it.
- The answer changes next quarter. If the recommendation would sit in a drawer regardless of what it said, the engagement has no route to value.
- There is enough signal to work with. Some customers, some revenue, some history. Pre-revenue companies with no customers have a validation problem, and validation is not what this practice does.
If the honest answer might be unwelcome and you would still want to hear it, that is the strongest single indicator of fit on this page.
5. What happens in the free audit
Book it →Nothing required
No pre-work, no questionnaire, no data request. A P&L summary or funnel report sharpens things faster if one is to hand, but it is not a prerequisite.
The situation, your words
You describe what is happening and what you have already tried. No interruption, no discovery script.
Live problem framing
The same discipline that opens every paid engagement, run on your situation: turning a described problem into a precise, answerable question.
The likely constraint
An initial read on where the binding constraint probably sits, and, more usefully, what evidence would prove it wrong.
A straight answer
Whether this is worth paying anyone to solve, and if so which model fits. Roughly half of these end with a recommendation not to hire an advisor.
A written recap
One page: the question as reframed, the hypothesis, and what would confirm it. Yours to keep and forward, whether or not anything follows.
Where AI sits in the work
Two roles, and conflating them is how buyers get sold the wrong thing.
Internally, AI is a cost structure. Model builds, market scans and document synthesis compress from weeks into days. That is why a sprint runs in two to four weeks rather than two to four quarters, and why the fee is what it is. It is not a client risk and it is not a deliverable.
As a service, AI enablement is its own capability. Assessing where agentic workflows genuinely pay, standing up governance, rationalising tools, and getting adoption to survive the second month. That is bought deliberately, not bundled.
What does not happen: your confidential material going into a public model. The AI-use clause is written into every SOW alongside the scope boundaries and IP terms, and you can read it before signing anything.
What is in scope, and what is not
Always included
- The named deliverable set, listed explicitly in the SOW
- The senior advisor doing the analysis, not reviewing someone else's
- A gap register: what could not be established, why, and what it would take
- Full ownership of every deliverable produced for you, with a licence to use, modify and share internally and with your board or investors
- A written recommendation even when it is that you should not proceed
Never included without a separate scope
- Implementation delivery: building the software, running the campaigns, managing the team
- Reconstruction of customer, funnel or margin history where it does not usably exist
- Interim executive cover or line management of your people
- Regulated financial, legal or tax advice
- Work outside genuine competence. If it is not a fit you get a straight answer and, where possible, a pointer to someone better suited
Methodology and templates stay with the practice. You own the outputs; the underlying frameworks remain proprietary. That split is written into every SOW rather than left to be discovered later.
When not to hire this practice
Five situations where the answer is no, stated here so you do not have to spend a call discovering them.
- You need capacity, not judgmentIf the constraint is already named and evidenced and what is missing is people to execute, hire people. An advisor will produce a better-argued version of what you already know.
- The decision is already madeEngagements commissioned to validate a settled decision produce expensive agreement. If the real need is board air cover, say so and it can be scoped honestly or declined.
- Pre-revenue with no customersWithout customers there is no retention, pricing or funnel evidence to analyse. That is a validation problem, and it needs a different kind of help.
- Nobody will own the outcomeA roadmap with no named owner inside your business does not survive contact with a full quarter. If no one has capacity to own it, the engagement has no route to value.
- You want the unwelcome answer softenedThe most valuable thing this practice sells is being told, on page one, with the evidence, that the thing you believe is not supported. If that is not wanted, the fee buys nothing.
Common follow-ups
How quickly can an engagement start?
Usually within two to three weeks of the audit, sometimes sooner. At most two engagements run at a time, so availability is genuinely limited rather than artificially scarce. If the timing does not work you will be told the actual date rather than strung along.
What do you need from us during the work?
Access to the data that exists, an hour a week from the decision-maker, and thirty to sixty minutes each from perhaps six people who see the problem daily. The heaviest ask is the decision session in D2, which needs the leadership team in one room for half a day. That session is where the engagement either produces committed choices or does not.
What if the diagnosis contradicts what leadership believes?
You get told, on page one, with the evidence behind it. That is the reason to hire an outsider at all. An advisor who confirms what you already believe has added a fee and no information.
Can we start with something smaller?
Yes, and it is often the right call. A Focused Module on one discipline is a genuine test of whether the working relationship is useful, at a fraction of the commitment. Several sprints have started that way.
What happens if the engagement is not working?
It gets said out loud at the weekly check, not saved for a post-mortem. Fixed-fee scoping means there is no incentive to extend a piece of work that has stopped producing value, which is precisely why the model is fixed-fee.
Related
Thirty minutes, and a straight answer either way.
If this page has not resolved it, the audit will. Including the possibility that the answer is no.