Launching a New Service Line Without Eroding the Core Business

Adding a service line is one of the more sensible growth moves a manufacturer can make. The margins are usually better than the product, the revenue is more predictable, and it puts you closer to the customer than a purchase order ever will.

It also fails in a way products do not, and the failure is hard to see. A product that loses money announces itself in the cost of sales. A service that loses money hides inside the factory — absorbed as engineering time, as machine hours booked to no job, as the quality manager spending Thursdays on a customer problem nobody is invoicing for. By the time it shows up in the accounts, it has been happening for a year.

I have built a service business from nothing, and I have watched manufacturers bolt one onto a factory. The ones that work get four things right early.

Define what you are actually selling

Services do not have drawings, which is precisely the problem. A product has a specification that says what it is and, by implication, what it is not. A service launched without that definition absorbs every adjacent request a customer makes, because saying no feels like poor service.

So write it down before you sell it. What is included, what is explicitly excluded, what the customer must provide, what the response time is, and what constitutes the job being finished. "Done" is the one most often left undefined and it is where the margin goes. A repair is done when it ships — or is it done when the customer's line is running again? Those are different commitments with different costs.

This is not legalism. It is the service equivalent of a drawing, and without it nobody delivering the work knows when to stop.

Know your cost to serve before you price

The instinct is to price off a day rate that feels competitive. The number that matters is what it actually costs you to deliver a day of it.

That includes the non-chargeable time around the work — travel, quoting, writing the report, the follow-up call — plus the realistic utilisation of whoever delivers it. Someone is not billable five days a week. If you build a price assuming they are, every week makes a loss and the shortfall is quietly covered by the factory.

Decide, too, whether you are selling time or outcome. Time is safer and easier to defend. Outcome earns more and transfers risk to you, which is fine only if you understand the work well enough to size it. Early on, sell time.

Resource it so it cannot raid the factory

This is the one that does the damage. The service launches, demand appears, and it is staffed by whoever is competent and available — which means your best engineer, your quality manager, the people already holding production together.

Two problems follow. Production slips in ways nobody connects to the new venture. And the service capability sits entirely in people who cannot be spared, so it can never scale.

The discipline is to decide in advance who delivers, how their time is booked, and what happens when service demand and production demand collide. Someone has to own that decision, and it should not be resolved weekly by whoever shouts loudest. If the honest answer is that you have nobody to spare, you are not ready to launch — or your first hire is the service, not the sales.

Keep it visible from day one

Put the service on its own P&L line immediately, however small it is. Revenue, direct cost, the labour booked to it, and the gross margin that results.

Without that, the service is evaluated on feel — and feel is unreliable, because busy and profitable are not the same thing. With it, you can answer the only question that matters at the twelve-month point: is this earning more per hour of scarce capability than putting the same hour into production?

Build the booking discipline on day one. Retrofitting time capture onto people who have spent a year not doing it is a miserable exercise and the historic numbers are never recoverable.

Approvals, liability and insurance

Services carry obligations products do not, and they catch people out.

If the service touches maintenance or repair in a regulated sector, find out what approval is required before you sell it, not after — in aerospace that may mean AS9110 or a maintenance approval, and it is not something to discover mid-contract. Check whether your insurance covers advice and work performed on a customer's site or on their equipment; product liability frequently does not. And look hard at the liability terms in the contract, because a service contract can expose you to consequential loss far exceeding the value of the work.

A half-day with the right adviser here is cheap. The alternative is a clause you did not notice becoming the most expensive sentence in the business.

Measure the right things

Product metrics do not transfer. The ones that tell you whether a service line is healthy:

• Utilisation — the proportion of available time that is chargeable.
• Recovery rate — what you actually earned per hour against what you intended to.
• Response and resolution time — the thing the customer is really buying.
• Repeat rate — whether customers come back, which is the honest verdict on quality.

If recovery rate is well below your target price, your scope definition is leaking. That is a scoping problem, not a pricing problem, and raising the price will not fix it.

Frequently asked questions

Should the service be a separate company?

Usually not at first. Separate reporting is essential; a separate legal entity is a question about liability, approvals and eventual sale, and it is easier to answer once the service has a track record. Keep the option open by keeping the numbers clean.

How do we stop the service cannibalising production capacity?

Decide the priority rule in advance and make someone accountable for applying it. Book service time to service jobs so the cost is visible. If the service consistently wins on margin per hour, that is useful information about where the business should be going — but you can only see it if you are measuring.

What is the most common mistake?

Underpricing because the cost to serve was never calculated, and then discovering the shortfall is being absorbed by the factory. Closely followed by scope that was never written down.

How long before we know if it is working?

Twelve months of clean numbers. Less than that and seasonality and the early learning curve distort it. This is the argument for getting the reporting right from the first invoice.

We help businesses put the structure, controls and reporting in place around new ventures — so you can tell early whether something is working. See our startup structures work, our organisational design work, or get in touch.