Think.

01

The moment

29 October 2026: that day I wrote down the five theses below with a date. Whoever thinks in loops needs something to check on the next run, and a prediction without a date cannot be checked.

These theses are my view, as of October 2026, of the next years in marketing and in the work of people like you: employed or freelance, with a feeling that something fundamental is shifting, and without a clear picture of what to sell afterwards.

02

What went wrong

What could go wrong with my own predictions is clear: without a date I could quietly rewrite them when they do not come true, and in the end I would always have been right. So one rule applies here: a thesis that turned out wrong stays on the wall, with a note next to it and an entry in the changelog. A prediction you quietly delete was never one.

Every thesis here has consequences for what I am doing right now. If I am wrong, I am wrong with my own business. My learning center failed on the fourth thesis before I had written it down.

03

What I know now

Five theses, as of October 2026. They connect, but each stands on its own.

As of October 2026

AI replaces competence itself

So far AI replaced tasks: translating a text, sorting a spreadsheet, cutting out an image. Now it replaces the knowledge of how to do the tasks. That is a different step. Whoever was paid for knowing something loses exactly that edge, and not at some point but client by client.

Consultants, agencies and freelancers feel it first, because their offer consists largely of knowledge. For you that means: if your plan for going independent is to sell what you know about marketing, you are selling something that is getting cheaper right now. I count myself among them, and that is why this thesis comes first.

As of October 2026

What sells is time taken off the client’s hands

Once knowledge is freely available, nobody pays for instructions. People pay whoever takes the work off their hands completely. The question for every offer is no longer: what do you know? It is: what do I no longer have to do myself?

I expect many offers to rebuild themselves along exactly this question over the next years. Workshops become services, consulting becomes systems, hour packages become flat fees. For you this is the good half of the first thesis: what you know gets cheaper, but what you can get done for others does not.

As of October 2026

Businesses close to people win

Doctors, physiotherapy, trades, gyms: what happens in person cannot move into a chat window. These businesses win, because AI takes over their admin and marketing while their core cannot be replaced. They get the efficiency without the threat.

So I expect exactly these businesses to become the best clients for everything that takes work off people’s hands. They have a real business, little time and no reason to become experts in data or advertising themselves. If you are looking for a niche, look here first.

As of October 2026

Only urgent, conscious problems sell

When the laptop is broken or the door is locked, you buy right away, without comparing for long. But nobody buys a solution to a problem they do not feel yet, however good it is. That was always true, but at a time when everyone is flooded with offers it becomes even clearer.

I expect offers that first have to explain a problem before solving it to have an even harder time. Anyone building an offer should start with problems the client already knows and wants solved today. My own learning center failed at exactly this point, with 0 bookings.

As of October 2026

Agencies selling hours lose to flat-fee systems

An agency that sells hours earns from work taking long. With AI it takes less time. Either the agency keeps billing hours it no longer needs, or it earns less. Neither is a good business in the long run.

The winners will be systems with a fixed fee, where the client buys a result and the efficiency stays with the provider. If you work at an agency that bills by the hour today, you probably feel that tension already. It is your hint at what your own offer should not look like.

04

Open question

What is open for me is how fast all this goes. I am fairly sure about the direction and not at all about the pace. If the theses are right but take ten years instead of two, that changes a lot about when you should act.

05

Changelog

29 October 2026
First version. Nothing here is rewritten silently: every later change gets its own entry in this changelog.