A personal trainer for your margin.
Most of us do not lack the knowledge to get fit. We lack someone who turns up every two weeks and asks what we did. Profitability PT is that, for the numbers: one-to-one coaching in costing and profitability, sold by the hour, and when it ends the model and the person who can run it are both still in your building.
your industry
every two weeks
CostCtrl, AI
on your data
Every profitability project ends. The question is what stays.
A project buys deliverables. It does not buy the habit of updating them.
Buy a large costing project and, for a few months, your margins are the best-understood numbers in the company. A team arrives, the model gets built, the whale curve goes up on the wall. Then the team leaves, and one of two things happens. Either the model goes with them, in a file only they know how to refresh, or it stays behind with nobody who can defend a number in it. Within two closes it is a photograph of a quarter that no longer exists.
That is not a failure of the consultants. It is the shape of the purchase. A project buys deliverables. It does not buy the habit of updating them, or the person who can argue with the sales director about a cost pool. Those have to be grown inside the company, and nobody sells that as a product.
We looked. As far as we can find, nobody, not the large firms, not the consulting marketplaces, not the costing certification schools, offers one-to-one coaching in costing and profitability that a company can buy by the hour. The large firms run CFO programmes, and they are good, but they are a day-long workshop, by invitation or nomination, once. The marketplaces will place a senior person with your team for a fee, which is a fine way to get the work done and a poor way to keep the skill when the person moves on. The schools teach the method in a classroom and stop at the classroom door.
Profitability PT sits in the gap. It is continuous, not a one-off. It is bought in blocks of four hours, by anyone, with no invitation. And its only deliverable is the one none of the others sell: at the end, your controller can build, refresh and defend the model without us.
Nobody hires a personal trainer because they do not know what a squat is. They hire one because on a Tuesday evening in February, somebody is expecting them.
That is the whole idea. A trainer does not lift the weight for you; the point of paying them is that you lift it. They set the programme, watch your form, and turn up. The obligation is the product, and the results compound: the tenth session is worth more than the first, because by then the routine exists and the muscle is yours.
I have sold the large project many times, and delivered it. It works when the company already has the muscle to carry the findings after we leave. Most companies do not, and no six-week sprint builds it.
What builds it is the same thing that gets any of us to the gym: somebody who has agreed to be there on Tuesday, who knows what you did last time, and who will not let you skip the hard set. Not a coach who lifts the bar for you. A coach who makes sure you do.
None of this is new to me. It is how I have worked informally with a number of companies for years: a call to help a TDABC implementation over a hump, a question about a specific piece of software, an afternoon checking whether a set of results holds up. Some of those relationships are still running. Profitability PT is that way of working turned into something you can buy directly, on purpose, instead of something that happens by chance.
So this is the most personal thing on this site. For now it is me, every two weeks, with your controller's screen shared and your numbers on it. Some sessions we build. Some sessions we argue about a cost pool for an hour. All of them end with your team one rep stronger, and none of them end with a report that has my name on it.
Miguel Guimarães
Buy hours. Set a cadence. Pick the topic. Do the reps.
Hours, not a subscription
You buy a pack of 4, 8, 15 or 30 hours. There is no monthly fee, no minimum term and no invitation. Hours are spent in sessions of 60 or 90 minutes, and the pack runs until the hours are used or twelve months pass, whichever comes first. A 4-hour pack is the smallest we sell; it is enough to scope a model and get the first cost pools right, and enough to find out whether the format suits you. If you need more than 30 hours, or want more than one person from your team coached in the same sessions, ask for a custom proposal instead of a pack.
A cadence that fits the work
We recommend one session every two weeks: long enough for your team to do real work between sessions, short enough that nothing goes cold. Weekly works for a team in a hurry; monthly works for a controller keeping a finished model honest. You set it, and you can change it.
Remote by default, on site when it matters
The remote price assumes video calls and shared screens, which is how most costing work gets done anyway. On-site sessions carry a 25% premium, with travel, accommodation and meals invoiced separately at cost. Some companies use one on-site day to launch, then go remote.
| Pack | Typical session | Sessions | Runs for roughly |
|---|---|---|---|
| 4 h | 60 min, weekly | 4 | one month |
| 8 h | 60 min, every two weeks | 8 | four months |
| 15 h | 90 min, every two weeks | 10 | five months |
| 30 h | 90 min, every two weeks | 20 | ten months |
A 15-hour pack, the middle one: ten sessions of 90 minutes, one every two weeks, spread over about five months. The gaps are the product as much as the sessions are. You set the cadence and can change it.
There is no fixed curriculum.
The pack is spent on whatever your profitability question is, drawn from what we already teach and publish here:
- Building a first TDABC model from the exports you already produce
- Activity-based costing, and when ABC beats time-driven and when it does not
- Cost-to-serve by customer, channel and order type
- The whale curve: reading it, defending it, and deciding what to do about the tail
- Capacity cost rates and unused capacity
- Pricing, minimum order values, surcharges and discount discipline
- Overhead allocation that survives an auditor
- Economic value added and product-line profitability
- Scenario simulation before a decision is taken
- Getting a model to reconcile to the general ledger
- Preparing the profitability story for a board, a bank or a buyer
Coaching by the hour only works if the hours are dense.
Ours are, because your trainer brings the same tool stack we use on every engagement, built over years of doing this and refined with AI in the last two.
A prompt library for costing
Tested prompts that turn a trial balance into a first cost-pool map, draft time equations from a process description, and flag the allocation choices an auditor will ask about. Your controller gets them, and learns to run them.
CostCtrl, connected to AI
Our own platform, with an MCP connection that lets an AI assistant read a model, run a calculation, and pull a whale curve or a cost-to-serve table in the middle of a session, on your data. Bring a workspace if you have one; if you do not, we can open one for the sessions. It is not required, and it is not sold inside the pack.
Validation, not just generation
The stack is used to check the model as much as to build it: reconciliation to the ledger, capacity that does not add up, a customer whose cost-to-serve moved 40% in a month. The speed is in the tedious parts. The judgement stays with your team, which is the whole point of the format.
In practice this means the first cost-pool map exists by the end of session one, not session three, and the hours that follow are spent on the decisions, not the typing.
You see the CV before you buy the hours.
Today the trainer on this page is Miguel Guimarães: 25 years of costing and profitability work, the Damco cost-to-serve case at Managing for Profit in 2009, SAP Education trainer across six countries, co-founder of CostCtrl. For most companies, and for the first pack, that is who you get.
Behind him is a bench of senior consultants, each with fifteen or more years in costing, management control or finance leadership, and each with an industry they know from the inside: manufacturing, distribution, healthcare, logistics, professional services. When you register, you see their profiles and choose. A controller in a hospital group may want the trainer who has run a hospital's cost model; a CFO in a food distributor may want the one who has argued about delivery cost with a fleet manager. You pick by company, industry and situation, and you can ask to change.
This is not a marketplace and you are not being matched by an algorithm. It is a short bench, chosen by us, shown to you before any money moves. That is unusual in this corner of consulting, and we think it should not be.
Every trainer on the bench works with the same method, the same tool stack and the same rule: your team does the work, and the model stays with you.
Consistency beats intensity. In the gym and in the P&L.
A model that is updated once, by an outsider, is a photograph. A model your controller refreshes every month, because somebody will ask about it in two weeks, is a capability. The second one is worth more every quarter it exists.
Everything is done by your people, on your screens, in your files. There is no handover at the end, because there was never a handover in. When the pack ends, the person who built the model is still sitting at their desk, and so is the model.
Nobody is defending their numbers against a consultant's numbers. There is only one set, and the team owns it. Findings that would have been contested in a steering committee get accepted, because the person presenting them is the person who found them.
When the tail of loss-making customers is fixed, it was the controller who found it and the CFO who took it to the board. We were the ones asking, every fortnight, what happened since last time.
No tender, no scope document, no steering committee, no invitation. A 4-hour pack is a purchase order most CFOs can sign on their own authority. If it works, you buy more hours. If it does not, you have spent four hours and €500.
Session one is slow: definitions, data, the first cost pool. By session six the team is asking sharper questions than we are. By session ten they are correcting us. That is the point.
Illustrative, not measured. Two shapes of the same spend. The project puts everything in at once and the capability fades as the people who carried it leave. A cadence puts less in each time and never stops adding, because the people who carry it work there.
Fifteen hours, one controller, one whale curve.
This example is invented to show the shape of an engagement. The company, the people and the figures are plausible, not real. Our published client results are on the case study pages, and they are labelled as such.
A building-materials distributor with €22M revenue, 640 active customers, two warehouses and its own delivery fleet. Gross margin is 27% and has been drifting down for three years. The CFO suspects the small accounts are the problem but cannot prove it, and the sales director says every customer is profitable "on average".
The CFO buys a 15-hour pack, remote, for the controller and one analyst. Sessions are 90 minutes, every two weeks.
We agree the question: cost-to-serve by customer. Using the prompt library on the trial balance, the controller has a first map of eight cost pools by the end of session one (order desk, picking, packing, delivery, credit control, returns, sales visits, and a residual). In session two we correct two of them: "delivery" was carrying vehicle depreciation that belongs to a capacity rate, and "sales visits" was hiding in commercial salaries. Homework: pull twelve months of order lines, delivery stops and credit notes from the ERP.
Capacity cost rates. The order desk costs €0.62 a minute, the fleet €1.85 a stop-minute, credit control €0.71 a minute. The analyst drafts the time equations with AI from the process descriptions and then corrects them against real timings: an order takes 4 minutes plus 1.5 per line plus 6 if it arrives by phone. We spend most of session 5 on one thing the team wanted to skip, the idle capacity in the second warehouse, because leaving it in the rates would have made every customer look 9% worse than they are.
The first whale curve, pulled from CostCtrl in the session. The top 130 customers generate 158% of operating profit. The bottom 210 accounts, a third of the customer base, lose €380,000 a year between them, almost all of it on orders under €150 delivered by the company's own vans. The sales director's average was right and useless.
What to do about it. The team models three levers: a €150 minimum order for own-fleet delivery, a €12 surcharge below it, and moving the 90 smallest accounts to a parcel carrier. We push back on one idea (dropping the accounts outright) because 40 of them are subsidiaries of profitable groups. The scenario the team picks recovers an estimated €210,000 in the first year, with 11 customers expected to leave.
The controller presents to the board. We are not in the room. The board approves the surcharge and the minimum order. The model is refreshed monthly from then on, by the analyst, in about two hours.
€1,600 for the pack, about 25 hours of the controller's time and 40 of the analyst's over five months, and one bruised assumption about averages.
A cost-to-serve model the team built and can defend, two people who now read a whale curve without help, the prompts and the workspace they built it with, and a habit of asking, every month, which customers moved.
Four packs and a custom option. The more hours, the lower the rate.
The full rate is €125 an hour, remote. Larger packs are discounted against it. On-site sessions carry a 25% premium, with travel, accommodation and meals invoiced separately at cost.
Weekly for a month. Enough to scope a model and get the first cost pools right.
Every two weeks for four months.
Every two weeks for five months.
Every two weeks for ten months.
50+ hours, or more than one team member coached in the same sessions.
| Pack | Remote | per hour | On site (+25%) | per hour |
|---|---|---|---|---|
| 4 hours · minimum, weekly for a month | €500 | €125.00 | €625 | €156.25 |
| 8 hours · every two weeks for four months | €900 | €112.50 | €1,125 | €140.63 |
| 15 hours · every two weeks for five months | €1,600 | €106.67 | €2,000 | €133.33 |
| 30 hours · every two weeks for ten months | €3,000 | €100.00 | €3,750 | €125.00 |
| Custom · 50+ hours, or more than one team member in the sessions | Request a proposal | - | - | - |
- Prices exclude VAT.
- Remote means video call and shared screen. On site means we come to you.
- On-site packs: travel, accommodation and meals are invoiced separately, at cost, against receipts.
- Hours are spent in sessions of 60 or 90 minutes at a cadence you set. Short questions by email between sessions are included, answered within two working days; anything that needs a shared screen becomes a session.
- A pack is paid in advance and is valid for twelve months from the first session.
- Unused hours from one pack carry into the next if you renew.
- No monthly fee, no minimum term, no automatic renewal.
- Need more than 30 hours, or two or three people from your team in the same sessions? Ask for a custom proposal; it is priced on the hours and the people, not on a pack.
Tell us the question. Pick the pack. We reply within one working day.
This is the whole sign-up. No call is required first, though you are welcome to book one. Tell us who you are, what you are trying to get past, and how many hours you want to start with, or tell us you need something custom. We reply within one working day with the trainer profiles that fit your industry and situation, a proposed cadence, and an invoice for the pack (or a proposal, if you asked for one). The first session is usually scheduled inside two weeks.
Questions we hear
- What topics can the hours be spent on?
- Anything this site already covers: TDABC and ABC model building, cost-to-serve, the whale curve, capacity costing, pricing and margin, overhead allocation, EVA, scenario analysis, getting a model to reconcile to the ledger, and preparing the profitability story for a board. There is no fixed programme. In the first session we agree the question, and the pack is spent on it. If your question is outside what we know, we say so in that first session and you keep the hours.
- How often do we meet?
- Every two weeks is the default and suits most teams: enough time to do real work in between, not enough for it to go cold. Weekly for a team under pressure, monthly for a controller keeping a finished model honest. You set the cadence and you can change it between sessions.
- Remote or on site?
- Remote by default. Costing work is done on screens and spreadsheets, and a shared screen loses very little. On-site sessions are available at a 25% premium plus travel, accommodation and meals at cost. A common pattern is one on-site day to launch, then remote.
- Isn't this just a fractional CFO by the hour?
- No, and the difference is not the price or the format. It is what is left in the company at the end. A fractional CFO, or any senior person placed with your team by the hour, does the work and takes the knowledge with them when the engagement ends; that is what you are paying for, and it is a fair trade if what you need is the work done. Coaching is the other trade. Your controller does the work, slower at first, and at the end the company has the model and the person who knows how to run it. If you need a finance function led for six months, hire the fractional CFO, or look at our Fractional Profitability Controller. If you need your own people to be able to do this without anyone, this is the product.
- Who should be in the sessions?
- The person who will build and keep the model, usually the controller or a senior analyst, and the person who will defend it, usually the CFO. Bring others for the sessions where their knowledge matters: operations for capacity, sales for pricing, IT for the data pull. Sessions are not lectures. The hours are worth most when your people are doing the work on screen while we watch.
- What happens when the pack runs out?
- Three things people do. Most buy another pack, often a larger one at a lower hourly rate, and move to the next topic; a team that wants more than 30 hours, or more than one person coached, asks for a custom proposal at that point. Some drop to a monthly session to keep the model honest. Some stop, because the team no longer needs us, which is the outcome the format is designed for. Unused hours carry into a renewal; a pack is valid for twelve months from the first session. There is no automatic renewal and nothing to cancel.
- How is this different from the Fractional Profitability Controller, or from Advisory?
- Three products, one method, different answers to "who keeps the model". The Fractional Profitability Controller is a monthly subscription: a senior controller, CostCtrl and AI embedded in your finance function for three to six months, doing the work alongside your team and handing over at the end. Advisory is the ongoing retainer once a model exists: reviews, scenarios, pricing. Profitability PT is bought by the hour, has no subscription, and your team does the work from the first session, so there is nothing to hand over at the end because nothing ever left. It is often how a model comes to exist in the first place, or how a team gets good at one topic without a project. If you are not sure which fits, the first conversation is free and we will say which one we would buy in your position.
- Can we choose who coaches us?
- Yes. Miguel Guimarães is the default trainer and, for the first pack, the usual one. When you register you also see the profiles of the senior consultants on our bench, each with fifteen or more years and an industry they know from the inside, and you can choose by company, industry or situation. You can ask to change between packs.
Your team builds it. We turn up every two weeks. The model and the skill stay with you.
If you know the question, the form above is the whole sign-up. If you would rather talk first, the first conversation is 30 minutes, with our team, and costs nothing: tell us the profitability question you cannot get past, and we will say whether a pack of hours is the right way to buy the answer, or whether you would be better served by a fixed-scope audit, a subscription, or a free Profit Check. If the honest answer is that you do not need us, that is what you will hear.
- Reply
- Within one working day
- Who answers
- A partner, not a bot
- Commitment
- None. The smallest pack is four hours.
Miguel Guimarães, Founding Partner. Cost and profitability practitioner for 25+ years. Presented the Damco cost-to-serve case at Managing for Profit (Amsterdam RAI, December 2009), on the same programme as Robert S. Kaplan. Co-founder of CostCtrl. Your sessions are with Miguel or with another senior trainer from the bench, depending on the industry and the preference you give at registration.
Proof
A distributor in New Zealand. €1.335M of cost-to-serve made visible, then halved, and 830 loss-making customers brought down to 295.
Read the case study →Who you would be talking to
Miguel Guimarães, Founding Partner
Cost and profitability practitioner for 25+ years. Presented the Damco cost-to-serve case at Managing for Profit (Amsterdam RAI, December 2009), on the same programme as Robert S. Kaplan.
Call +351 910 313 731
