Skip to main content

Search...

A Productized Service Needs No Hourly Rate

A requirements spec in two weeks for 20,000 euros: this productized service takes its price from the project risk. Nobody negotiates hours.

• • 8 min read
Cover of the expert talk on 'A Productized Service Needs No Hourly Rate' with Maik Pfingsten and Richard Seidl.

A productized service is a service sold like a product, with a clearly defined result, a fixed process and a price set in advance. That price comes from what the result is worth to the client; for a requirements specification, that is the project risk it reduces. Hourly rates no longer come up in the negotiation.

Key Takeaways

  • In one example, Maik Pfingsten prices a requirements spec at 20,000 euros from project risk: 30 percent of 5 million euros, a third of that, his 20 percent, five releases.
  • Hours never come up when a productized service is sold, so the client does not care about the effort; for the requirements spec it ended up under 30 hours.
  • Maik Pfingsten scaled through four other freelance systems engineers who ran jobs on his process. Nobody billed by the hour there either, so he saw no problem with false self-employment.
  • A documented process shows step by step what a person does, what gets automated and where an AI agent helps, for example with sorting the documents a client hands over.
  • A productized service can also be twelve months of mentoring at a fixed price, with the stated goal that the client’s team carries on alone afterwards.

Getting Off the Time-for-Money Treadmill

Trading time for money has a built-in ceiling for freelancers. While the contract runs, the account is full; when it ends, you sit at home and wait for the next one. A productized service sells a clearly defined result at a fixed price instead. How many hours it takes you is your business.

Maik Pfingsten has lived both models. He set up his own engineering firm in 2005 and then spent years as a freelance troubleshooter, rescuing large international projects. The money was very good, but he calls it a “golden time-for-money hamster wheel.” His wife had a simple rule for it. If Maik is on the road, there is money in the account and the fridge is full. If Maik is at home, there is a problem.

In 2010 he stood on the Rhine embankment, frustrated, looking for another way. Role models were hard to find. The reputable online entrepreneurs he asked in the US knew the consumer business; engineering services for mid-sized B2B clients were foreign to them. So he experimented on his own until his service had become a product.

How Do You Turn a Service into a Productized Service?

In four steps: reduce, focus, systematize, scale. Maik originally packed this path into two workshop days for another freelancer. On day one, the one to three “rough diamonds” are picked out of a sprawling list of offers and polished. On day two, the service gets a fixed process, a price tag and rules of the game.

His own first product came out of a March morning in 2015. The participant’s train broke down, day two was cancelled, and Maik applied the prepared method to his own engineering firm instead. By evening he had the result in front of him, a requirements specification in two weeks.

The work itself was nothing new. He had already written requirements specifications as a troubleshooter, because in a crisis project he needed the client’s wishes and requirements on the table within roughly two weeks. For that he had developed the System Footprint, a visualization method. Hardly anyone loves the job, by the way. If you want to put engineers into a deep sleep instantly, Maik says, give them a requirements spec to write.

What was new was a process planned from end to end, which Maik calls his “star chef’s recipe.” It starts with the inquiry and the question of whether client and offer fit. Once the inquiry becomes an order, five phases follow: capture, sort, fill in, review, sign off. After two weeks the client holds a complete, signed-off requirements specification.

Then comes the follow-up. The invoice is part of it, but this is also where Maik places small details that produce a wow effect on the client’s side. They double as a “moat.” Anyone who copies the recipe and cuts the steps that look superfluous takes away the very thing that makes the service special.

How Do You Derive the Price from Project Risk?

From the value the result has in the client’s project. Maik walks through the math in the first meeting, using the basics of project risk management. His example is a project with a budget of 5 million euros.

StepCalculationResult
Project risk30 percent probability that the project fails, on a budget of 5 million euros1.5 million euros
Value of the requirements specificationone third of the risk500,000 euros
The provider’s share of that value20 percent of it100,000 euros
Price for the first versiondivided by five releases20,000 euros

The third comes from three measures that lower the risk up front: a very experienced project manager, an experienced development team, or someone who writes down all wishes and requirements properly. Each gets an equal share.

Of the specification’s value, Maik assigns 80 percent to the client, because the content comes from them. He sees himself as the methodologist who turns it into a professional document. A project that runs for about a year and a half produces three to five versions. Since he usually writes only the first, he divides by five once more.

Many of his clients heard for the first time in that calculation what such a document is worth, and they ordered right away. Depending on project size, prices mostly landed between 10,000 and 25,000 euros. For anyone who works in testing, the parallel is obvious, because testing is risk mitigation. The better the testing, the lower the product risk.

Hourly Rates Never Come Up

With a fixed price, time is off the table. From the first conversation on, Maik never talks to his clients about hours or hourly rates. The client then has no reason to care how much effort sits behind the result.

That left him free to keep refining the recipe. In the end Maik was down to under 30 hours of effort per specification. Two or three times a year a client wanted to start right away. For those cases he had a fast track, where the start date moved forward and the price doubled.

In the summer of 2015 he attached an invoice for 25,000 euros to an email and, by his own account, stared at the monitor for fifteen minutes before he clicked send. The reply came two hours later. The invoice had already been approved for payment, and the client was looking forward to starting in two weeks.

What the model added up to only dawned on him two years later. His tax advisor announced 40,000 euros in back taxes, and Maik remembered a summer that, in his memory, he had spent with his laptop in the garden and with his kids.

After that he scaled through other “star chefs.” Four freelance systems engineers took on orders using the same recipe, once Maik had qualified the client and accepted the order. Nobody billed by the hour among themselves either, so false self-employment was no longer an issue in his view.

Where AI Agents Fit into the Recipe

A written-out process shows step by step where a human is needed, where plain automation is enough and where AI fits. Maik went through his old specification recipe with that question, together with the owner of the engineering firm he has since sold his business to.

In the capture phase, the kick-off workshop, almost everything stayed with humans. One agent only checks the System Footprint against the conversation.

Sorting the documents the client supplies offered more. The first attempt, dumping all documents into one agent at once, failed. It only worked once the agent received one document at a time and held each against the System Footprint, the way Maik used to read and file them himself.

In a trial run with the specification for a wallbox charger, the agent reported back that the spec for a diesel shut-off valve probably didn’t belong in there. It now goes by “Dieter Dokumentenfresser,” roughly Dieter the document eater. A second agent, “Conny Consistent,” handles the consistency check and smooths out the requirements grammar.

What Variants Exist Besides the Requirements Specification?

Maik distinguishes two forms. The pure variant delivers a finished result for a problem the client has no time, no interest or no expertise for. The mentoring variant accompanies a client team over a fixed period, with a clear roadmap. That second form is productized consulting with a defined end.

He had an offer of his own for it. For twelve months he worked as a virtual mentor with development teams that wanted to introduce systems engineering, again at a fixed price.

“The goal of the mentoring is that after twelve months you can walk on your own. I make myself redundant from day one.”

(Maik Pfingsten)

That sets him apart from consultants who settle in at the client for the next five to ten years. A development project at Hilti that he mentored this way was finished four weeks before the start of production and needed less budget than estimated.

From the IT world, Maik names two examples from his mentees. A freelancer from cyber security offers a productized service that probes the AI agents a company has in use and shows where security gaps might be. An IT company takes over data protection for telecommunications groups as a full service across a whole year and delivers a report at the end.

That annual package also contains smaller services, such as persistent data access requests from individual customers, which the telecom hands over completely.

Share this page

Related Posts