Productize: Turning What You Know Into a Product

To productize (productise, in British spelling) is to turn something you deliver case by case into something defined, priced and repeatable that a client can buy without it being designed from scratch. For an expert or a firm, it means taking what you know, the way you diagnose a problem, the steps you always follow, the standards you will not drop, and packaging it so it no longer has to be rebuilt, or rebilled, every time it is sold.

It is an old idea that has become urgent. As long as expertise was scarce and slow to apply, selling it by the hour was a reasonable proxy for its value. AI has broken the proxy. Work that took a day now takes an hour, and a business that sells hours has just watched its unit of sale shrink.

Why the hour stopped working

The pressure is coming from clients, and it is showing up in pricing conversations before it shows up in invoices. In a General Assembly survey of 258 senior leaders at consulting, accounting and legal firms in the US and UK, published in March 2026, 79% said AI is changing their pricing conversations and 42% said clients are questioning their pricing model. In Wolters Kluwer’s 2026 Future Ready Lawyer survey of 810 lawyers, 62% of corporate legal department respondents expect AI to significantly reduce the billable hour.

Expectation is running ahead of change. The 2026 Best Law Firms survey of 4,852 US firms found that 58% say AI has had no effect on their billing practices yet. Most firms are still selling hours to clients who no longer believe in them.

That gap is the opening. If you sell time, every efficiency AI gives you is a discount you hand to the client. If you sell a defined result at a defined price, the same efficiency stays with you as margin, and the client still gets a faster, clearer deal than before. The briefing on the future of consulting and agencies describes the business models firms are moving to; productising sits underneath most of them.

What you are actually productising

The instinct is to productise the deliverable: the report, the audit, the campaign. The more valuable thing is the judgment that produces it.

Every experienced practitioner carries a method they rarely write down: the questions they ask in the first meeting, the three things they check before anything else, the warning signs they have learned to spot, the standard a piece of work has to meet before it leaves. That method is what clients are really paying for. It is also the one thing a general-purpose AI model does not have, because it was never published anywhere a model could learn it.

This is the economic turn AI creates. What is generic gets cheaper, because anyone can now produce it. What is specific to you, your way of seeing a problem, your standards, your accumulated cases, gets more valuable, but only once it exists outside your head. As what happens to knowledge work sets out, the work that remains for people is framing, verifying, integrating and deciding. Productising is how you sell those four at scale instead of one engagement at a time.

The four forms, from least to most scale

There are four ways to turn what you know into something that sells repeatedly. They are not rungs you must climb in order, but each one demands more work up front and returns more independence from your own time.

1. A productised service. You still deliver, but the scope, the price, the timeline and the method are fixed in advance. The client buys a known thing off the shelf. This is the fastest form to launch and the closest to how most experts already work.
2. A knowledge product. The method itself becomes an asset others use without you: a playbook, a template set, a diagnostic, a course. It sells your judgment once and delivers it many times.
3. A micro-SaaS. A small piece of software that does one job for one kind of customer, often your own method running as a workflow. It earns while you are not working, and it needs customers who find it without you.
4. An AI product. Software built on top of the foundation models, where the model supplies general capability and your context and judgment supply the part a model cannot. It has the most reach and the most moving parts.

Each form has its own briefing in this silo, covering what it is made of and where it breaks. This page is about choosing between them.

Which form fits

Three questions settle most of it.

How standard is the problem? If most clients arrive with recognisably the same problem, the solution can be productised. If every case genuinely differs, productise the diagnosis rather than the solution: a fixed-price assessment that ends with a clear recommendation is a product even when the work that follows is bespoke.

Does the value sit in the method or in the moment? If the value is a method someone else could follow, it can become a knowledge product or software. If it depends on your judgment applied live, to a negotiation, a crisis or a board, it stays a service, and the productised version is the frame around it: a fixed scope, a fixed price, a defined end.

How much do you want to stay in delivery? A productised service still needs you or your team every time. Software does not, but it trades delivery work for the work of finding and supporting customers you will never meet. Neither is easier. They are different jobs.

A useful rule: start with the form closest to the work you already do, and move outwards only once the method has stopped changing. Software built around a method that is still moving gets rebuilt every month.

And keep the economics honest. Software built on AI models pays for every use, so its margins look less like classic software than the pitch decks suggest: ICONIQ’s 2026 survey of companies building AI products put their average gross margin on those products at 45% in 2025. The build versus buy logic applies to your own product as much as to your tools: own the part that is specifically yours and rent the rest.

Where productising goes wrong

• Packaging what you like doing instead of what clients buy. The product has to answer a problem clients already describe in their own words. If you have to explain why they need it, it is not ready.
• Productising too early. A method you have run a handful of times is a hypothesis. Run it by hand until you can describe every step without hedging, then fix it.
• Fixing the price but not the scope. A fixed price over an open scope is not a product. It is a discount with extra risk attached. The exclusions matter as much as the inclusions.
• Building it inside someone else’s platform. If your method lives as prompts and automations inside a vendor’s tool you cannot export, you have productised your judgment for their benefit. The same vendor lock-in logic applies to your own products as to the tools you buy.

Where to start

Every one of the four forms begins with the same artefact: your method, written down properly. The questions you ask, the steps you follow, the standards you hold, the examples of work you are proud of and the mistakes you have learned to avoid.

That document is the specification for a productised service, the content of a knowledge product, the logic of a micro-SaaS and the context an AI product runs on. In the language of the AI operating system, it is the context layer: the part that compounds, and the part you most want to own. Write it once, keep it current, keep it yours, and choosing between the four forms becomes a question of packaging rather than invention.

If you want your expertise turned into something that sells without being rebuilt every time, tell Daniel what you deliver today, what you have already written down or built, and what you need the business to do for you. Book 30 minutes with Daniel.

Frequently Asked Questions

What does productize mean?

To productize means to turn a service or a piece of expertise into a defined, repeatable offer with a fixed scope and a fixed price, so it can be sold and delivered the same way each time rather than designed from scratch for every client. The British spelling is productise.

What is an example of productizing a service?

A consultant who used to quote each strategy project separately might offer a fixed-price, two-week assessment with a set list of interviews and analyses and a written recommendation at the end. The client knows what they will get and what it costs, and the consultant delivers it the same way every time.

What are the main ways to productize expertise?

Four: a productised service, where you still deliver but scope, price and method are fixed; a knowledge product such as a playbook, template set or course; a micro-SaaS that runs your method as small software; and an AI product built on foundation models, with your context as the differentiator.

Should I productize a service or build software?

Start with the form closest to the work you already do. A productised service launches quickly and teaches you what clients actually buy. Move to software only once the method has stopped changing, because software built around a moving method has to be rebuilt constantly.

How does AI change productizing?

It makes it both more urgent and easier. More urgent, because AI shrinks the hours that time-based pricing depends on: in a 2026 General Assembly survey, 79% of professional services leaders said AI is changing their pricing conversations. Easier, because AI can carry much of the delivery once your method is written down.

What should I productize first?

The diagnosis. For most experts, the most repeatable part of the work is how they assess a problem at the start. A fixed-price assessment with a clear recommendation is often the simplest first product, and it works even when the work that follows stays bespoke.