From hourly billing to value billing: what AI means for your pricing as a freelancer

A few weeks ago I built an online booking system with dynamic pricing logic as a TYPO3 content block. Different services, different time periods, different prices. With Claude Code as a tool, the whole thing was implemented in just a few hours. A year ago, it would have easily taken two to three days.

When I was finished, I asked myself a question that I now come across all the time: What would this be worth on an invoice?

In my workshops, in the community, in the Business Roundtable, I hear this very question more and more often. Freelancers and agencies who realize that AI makes them faster, but their billing model penalizes them for it. They deliver better results in less time and earn less.

Welcome to the dilemma of hourly billing in the AI age.

The hourly rate paradox

The math is simple: if AI cuts your working time in half and you bill by the hour, your revenue is cut in half. With the same result. That makes no sense.

For agencies, the problem is multiplied: if your team delivers twice as fast thanks to AI, not only does the revenue per project halve. It also becomes more difficult to explain to clients why a TYPO3 relaunch now suddenly only takes half as many hours as last time.

Nevertheless, many freelancers and agencies cling to hourly billing. Because it seems transparent. Because the customer "sees what they are paying for". Because it has always been done this way.

But transparency does not automatically mean fairness. And it certainly doesn't mean that you are paid appropriately.

Why AI is forcing the issue onto the table

Until now, the topic of value billing has been rather theoretical for many. Yes, it was debatable whether hourly billing was the right model. But in practice, it worked because the time worked and the value of the service correlated to some extent.

This correlation is now disappearing.

With tools like Claude Code, Cursor or GitHub Copilot, tasks that used to take days can be completed in hours. TCA adjustments, content blocks, configurations. Everything is much faster. But the value of the solution remains the same. Or even increases because the customer gets a result faster.

Those who continue to stubbornly charge by the hour are penalizing themselves for efficiency.

Four models beyond hourly billing

So what to do? Here are four approaches that work in practice.

1. value billing: What is the solution worth to the customer?

Instead of asking "How long did it take me?", ask: "What does this solution do for the customer?" An online booking system that enables the customer to sell appointments or services directly via the website has a concrete economic value. This value does not change because you built it in three hours instead of three days.

An example: You build the booking system in four hours. At an hourly rate (let's say 120 euros), the invoice shows 480 euros. But what is this solution worth to the customer? A system that enables online bookings saves on personnel, reduces errors and directly generates sales. The value is more likely to be between 1,500 and 2,500 euros. And that is exactly the price you should calculate.

This assumes that you understand the customer and their business model. But that is part of your expertise.

2nd know-how price: 10 years of experience in 2 hours

A less experienced developer might need two days for the same task. You can do it in two hours. Not because you type faster, but because in 10 years you have learned which architecture works, where the pitfalls lie and what the customer really needs.

This knowledge is the product, not the time you spend sitting at the computer.

3. the entire problem-solving process

Writing code is only a fraction of the actual work. Before a line of code is written, you define the goal, plan the architecture, think about security, usability and accessibility. After the code comes the review, the testing, the documentation.

All of this is part of the problem-solving process and must be included in the calculation. If you only calculate the pure implementation time, you are giving away the majority of your service.

4. package prices: Three options, clear service

Instead of quoting an hourly rate, define three packages with different scopes of services. Basic, standard, premium. The customer chooses what suits their budget and requirements.

Experience shows: Most customers opt for the middle or higher option. And you have a clear calculation that works regardless of your working hours.

For agencies, package prices can be implemented particularly well via maintenance contracts and service levels. Instead of billing individual support hours, you can offer monthly packages with a defined scope of services: Updates, security monitoring, performance checks, content support. The customer gets planning security, you get recurring revenue.

The vibe coder trap

Of course, there are now people who click together complete applications with AI without really understanding what is happening. This may even work for simple websites without complex functionality.

But for complex projects with API connections, for applications in which security plays a role, for anything that goes beyond a standard website: that's where it becomes critical. Agencies in particular are familiar with this: customers come with the expectation that "with AI, everything should be faster and cheaper now". What they don't see is the effort involved in quality assurance, architecture and the responsibility you take for the result.

AI-generated code often works. But "works" and "is safe" are two different things. I see this in my own projects and hear it regularly from developers in the community: AI tools generate code with potential security vulnerabilities. These are not automatically detected or closed. Only when you look specifically, because you know what is important, do they become apparent.

And that's exactly the point: this "knowing what matters" is what distinguishes you as an experienced developer. No AI can replace that. Not yet. And whether it will look different in 2 or 5 years' time is something none of us can answer with any certainty today.

At the same time, vibe coders are creating price pressure on the market. Customers see that "someone with AI can do it too" and ask themselves why they should pay more. This is exactly where value billing protects you: you are not selling code, but a functioning, secure, well thought-out solution. That's fundamentally different from clicked-together output without quality control.

Don't sell yourself short

If you have several years of experience, satisfied customers and demonstrable results, then you have every reason to charge reasonable prices.

Customers who are only looking for the lowest price are rarely the best customers. They negotiate every item, scrutinize every hour and are still dissatisfied in the end. Customers who are willing to pay fair prices are usually more appreciative. For your work, your time and your expertise.

This does not mean that you should charge exorbitant prices. It means that you realistically assess and communicate the value of your work. And that you allow yourself to turn down clients for whom the ratio is not right.

What you can do now

Take your next quote and do the math for both models: What would you charge by the hour? And what is the solution actually worth to the customer? This comparison alone will open your eyes.

You can then proceed step by step:

• Test on a new project: you don't have to change everything immediately. Try package prices or value billing for the next suitable order.
• Calculate the entire process: conception, consulting, review, security check. All of this has value and belongs in your offer.
• Communicate the added value: Explain to the customer what they will get, not how long it will take you. Fast delivery is an advantage, not a reason for a discount.

And if you are unsure whether the new approach will work: talk to others who are facing the same question. It's difficult to change pricing on your own. It's much easier when you talk to people who are working on similar projects.

Conclusion

AI does not change the value of your work. It changes the time you need for implementation. And that's a reason to rethink your billing model, not to lower your prices.

Your expertise, your experience, your eye for the big picture: these are the things that customers pay for. The mere line of code is only the visible result of a much larger process.

The insights in this article come not only from my own practice, but also from many conversations with freelancers and agencies who are faced with precisely these questions. We regularly have this exchange in my Business Roundtable Community: open, direct and without a sales show. If you find yourself in this article, you're in good company.