The growth market you already own. Digital and AI-enabled after sales, built end to end.
Your installed base is demand you have already paid for. We build both halves: the service organisation that runs differently, and the business that earns on it. Not a digitalisation project — a P&L of its own.
Most manufacturers run after sales as a cost function — steered by cost per case, not by the revenue a machine produces over its life.
In 2023 Vestas sold its turbines at a margin of minus 1.2%. Its service business, same year: plus 21. Not an outlier — the pattern. Seven out of seven groups that report service separately earn the better margin in service. Seven out of seven. An accounting mistake with strategic consequences: not knowing that number for your own house means steering your biggest profit engine blind.
The installed base is demand that has already been paid for. No acquisition cost, known machine, known address. It is the cheapest growth market a manufacturer has — and the only one it can defend, because the machines are standing at its customers' sites.
Why now — the arithmetic changed
After sales was always profitable. What is new is that the three things that made capturing it expensive are tipping at the same time. Cases that did not add up five years ago add up today.
The core market no longer carries growth
Flat unit volumes, prices under pressure, new competitors from Asia. Growth is not in the next machine. It is in the machines already out there.
Diagnosis collapsed
The expensive step in any service event was always the same: finding out what is broken. That step now costs almost nothing.
Regulation creates demand
The EU repair directive 2024/1799 makes repair an obligation. Treat it as an obligation and you pay for it. Treat it as a sales channel and you earn from it.
If you do not read your installed base as a business today, you hand it to whoever does — the independent service market, the marketplace, the dealer. And they do not hand it back.
Five collapses where the arithmetic turns. One of them is the condition for the other four.
Documents → structure
Manuals, P&IDs, bills of material, service history turned into a queryable model of your machines. This used to be a specialist project with an entry price only the largest installed bases could carry. Today it is the cheap part — and still the condition for the four that follow. Skip it and you build demos, not economics.
Symptom → cause
A description, a photo or a sound turned into a ranked fault hypothesis with a part number. Finding out was always the expensive step, and it cost the same again on every single case. That multiplication is what disappears. Models that used to fail on case volume now carry themselves.
Expert → partner network
And not for your own people — they know what to do in 95% of cases. The collapse happens one level further out: at the untrained multi-brand partner, and at your hotline, which pays for his diagnosis today. Enabling a partner network across the board was never affordable. Now it is.
Signal → prediction
Anomaly detection on condition data turns unplanned stops into scheduled work. Until now that only paid for the most expensive assets, because every machine model was its own project. That threshold is moving down. The benefit lands with the operator, so it belongs in the contract, not in the datasheet.
Contact → affordable channel
Once a large share of cases resolves automatically, a direct relationship with the end customer becomes affordable at volumes where it never was. This is the collapse where cost reduction turns into revenue.
Five collapses where the arithmetic turns. One of them is the condition for the other four.
Documents → structure
Manuals, P&IDs, bills of material, service history turned into a queryable model of your machines. This used to be a specialist project with an entry price only the largest installed bases could carry. Today it is the cheap part — and still the condition for the four that follow. Skip it and you build demos, not economics.
Symptom → cause
A description, a photo or a sound turned into a ranked fault hypothesis with a part number. Finding out was always the expensive step, and it cost the same again on every single case. That multiplication is what disappears. Models that used to fail on case volume now carry themselves.
Expert → partner network
And not for your own people — they know what to do in 95% of cases. The collapse happens one level further out: at the untrained multi-brand partner, and at your hotline, which pays for his diagnosis today. Enabling a partner network across the board was never affordable. Now it is.
Signal → prediction
Anomaly detection on condition data turns unplanned stops into scheduled work. Until now that only paid for the most expensive assets, because every machine model was its own project. That threshold is moving down. The benefit lands with the operator, so it belongs in the contract, not in the datasheet.
Contact → affordable channel
Once a large share of cases resolves automatically, a direct relationship with the end customer becomes affordable at volumes where it never was. This is the collapse where cost reduction turns into revenue.
What changed is not what is possible. It is what pays — on both sides. The build that used to take specialists and months costs a fraction. And the expensive step in daily operations, the one that multiplied with every case, costs almost nothing. The case you shelved three years ago because it did not pay: run the numbers again.
In 2017 we built exactly that at Linde: an operations engineer clicking his spare part straight out of a hand-inked P&ID. Eight plants. It paid because Linde was big enough. Today the same case carries two sizes smaller. The same asset most recently at a home-appliance manufacturer: roughly plus 8% products sold.
And we do not sell you a chatbot for it. Package the asset, not the assistant. Models get swapped in two years; the structured knowledge about your machines stays yours. What AI does not do: it lowers cost per case, but it creates no revenue. That only appears once you monetise the failure moment.
Two tracks — and we tell you up front which one is yours
The new business on the installed base
A revenue model your warranty P&L cannot host — built in a vehicle of its own, with our equity in it and without touching your dealer contracts. The defensible core is the data and knowledge asset, not the app around it. Your path if you can create a P&L, not just spend a budget. Eighteen months plus, in phases with exit points. Built: Helpify, PleaseFix, Linde PLANTSERV.
The rebuild of the running service organisation
Case flow, triage, dispatch, first-time fix, parts logic. That work produces the structured asset AI needs before it computes anything. Fee-based, inside your P&L, no new entity. Your path if you have a budget and a cost target but no appetite for setting up a company. Three to nine months to a measurable change.
Usually both — in that order
The rebuild surfaces the business case and cleans the data the new business needs. The new business creates the revenue that justifies the rebuild. Never either-or — and which order is yours is a finding of the diagnosis, not a preference of the seller.
01 — What the installed base and the purchase process are worth today
How many units, where, what a service event is worth today and what the failure moment could be worth. Alongside it, with equal weight, how the purchase actually runs: rebuy, upgrade, cross- and upsell. And a warning: never assume your installed base matters to your customer. It is your asset, not his. The sales motion has to stand on his entry point, not yours.
02 — Where revenue arises along the customer journey, and at what margin
Which revenue streams exist at which point of the journey and what margin sits on each. And which alternatives your customer has there: grown or receded, and why. Without this half you describe the organisation but not the field it plays on.
03 — Process and data diagnosis
Where the process breaks, which system holds which truth, and whether your documentation carries an asset or just a filing cabinet. This is the question that decides the path. Bad data means rebuild first — an AI business on unusable data fails on its own promise, and it fails in public.
04 — Named weaknesses with consequences
No maturity score. A list of weaknesses, each with what it costs you. A score tells you where you rank. A consequence tells you what to do.
05 — Roadmap, rebuild hypotheses and a recommendation on which path is yours
The hypotheses for the rebuild, derived from 01 to 04 — and the recommendation: rebuild first, new business first, or both in sequence. With the criteria attached, so you can check the recommendation yourself.
What comes out of this is a business. Not a concept.
We want to build new business in after sales, with the latest technology and a hard focus on real growth. Not bill projects. That is why we look at your cases through a product and investor lens: what carries itself, what scales, what we would fund ourselves. Not a transformation project on hourly rates. A business build with a number in your P&L.
What we are not
Not a consultancy
We do not stop at the recommendation. We derive the case and build the thing that earns it. If the deliverable is a presentation, we are the wrong address — there are cheaper providers for that.
Not a systems integrator
We do not take your system of record. We build the layer above it. That boundary is deliberate — it is the reason your IT can work with us instead of against us.
Not a long-run service operator
We do not take over your service for good. We rebuild it, or we build the business next to it — and after that it runs at your house. If you are looking for an outsourcing partner to staff the hotline for the next ten years, that is not us.
Plant engineering & process industry
Downtime costs more than any repair. Documentation sits on paper and in Excel, not in the ERP. The plant builder knows every component of its own plant — and often still holds only a fraction of the spare-parts business. Our first wave.
Home appliances & consumer durables
The repair costs more than the customer will pay — and in that same moment the replacement purchase is decided. On top of it, an EU repair obligation that looks like a cost block and is in fact access to the customer at the most valuable moment.
Rail & commercial-vehicle fleets
Safety-critical systems, hard availability commitments, an operator who thinks in fleet cost rather than part price. Here availability is the product — not the part.
HVAC & building technology
The installer stands between you and the end customer — and gives up neither the contract nor the data. Only a channel-safe design works here. We tested that in the field, including the findings that argued against our own thesis.
Professional & commercial equipment
The machine earns money while it runs. So willingness to pay tracks the operator's EBIT and does not wait for a failure. That is a different business from repair — and a better one.
Machinery & industrial equipment
Large installed base, distributed service organisation, spare-parts stock that is half dead in the warehouse and half missing exactly when it is needed. The classic case for diagnosis before dispatch.
Skin in the game: why bad advice costs us too
We carry part of the risk
When the diagnosis turns into a business of its own, we are co-founders with equity — not a service provider on a day rate. Our return depends on the business working, not on how long we work on it. That is why we talk you out of ideas a consultancy would sell you.
You get both jobs, not one job and a recommendation
Rebuilding the service organisation and building the business that earns on it are two different jobs carrying two different risks. We have done both — and more than once the new business fell out of a diagnosis nobody had commissioned it from. You do not get one half and a recommendation for the other.
We did not have to understand your service case for the first time
Twenty years of building products and companies: 35+ ventures, 400+ enterprise tech projects, IT and AI delivery in-house. We start from where the money is earned and build towards the technology from there. The difference does not show in the pitch. It shows in month four.
Skin in the game: why bad advice costs us too
We carry part of the risk
When the diagnosis turns into a business of its own, we are co-founders with equity — not a service provider on a day rate. Our return depends on the business working, not on how long we work on it. That is why we talk you out of ideas a consultancy would sell you.
You get both jobs, not one job and a recommendation
Rebuilding the service organisation and building the business that earns on it are two different jobs carrying two different risks. We have done both — and more than once the new business fell out of a diagnosis nobody had commissioned it from. You do not get one half and a recommendation for the other.
We did not have to understand your service case for the first time
Twenty years of building products and companies: 35+ ventures, 400+ enterprise tech projects, IT and AI delivery in-house. We start from where the money is earned and build towards the technology from there. The difference does not show in the pitch. It shows in month four.
Get in touch
If you want to stress-test an idea, unblock an initiative, or simply see whether we're the right partner, let's talk. You can book a straight-to-the-point session with our CEO Manni, or leave us a message.
30-minute call. No strings attached.
