Two current reports touch on the same topic from different angles. One is about self-employed people and freelancers whose business models are coming under pressure from AI. The other is about the question of who actually gets the time saved that new tools generate in everyday working life. Taken together, they describe an area of tension that affects every small and medium-sized enterprise as soon as it uses AI productively.
Threatened value proposition
The starting point of the first report is uncomfortable: many business models are threatened by AI, and this expressly applies to self-employed people and freelancers too. What is affected is wherever the service essentially consists of standardisable production – texts, simple design, translations, recurring analyses. When customers can produce these results themselves, pure execution loses market value.
According to the report, however, Jimdo CEO Matthias Henze also sees many opportunities in this. This dual perspective is more than diplomacy: the same technology that devalues individual services lowers the cost for small providers to offer services that could previously not have been delivered with the available staff. A one-person operation can handle offers, documentation, analyses and customer communication at a level that would previously have required a team.
The practical consequence is to review your own value proposition and shift it to where it cannot be replaced: responsibility for the result, industry knowledge, access to data and systems, operation and liability. Anyone who instead defends the price for pure production is competing against falling marginal costs.
Who owns the time gained?
The second report turns the question around. While new tools increase employees' efficiency, the question arises as to whether and how they share in the time gained. Some tech managers answer this unequivocally with no: at Meta, the line is not to ask for additional days off, but to invest the capacity gained in more work output.
That is a distribution decision, not a technical one. Productivity gains can flow in three directions: to the company as margin, to customers as a lower price or shorter delivery time, or to employees as relief. In practice, the decision is rarely made openly; instead, whoever formulates the expectation first prevails.
For small businesses, this is more relevant than it initially seems. A corporation can skim off efficiency gains through volume. In a business with five or fifteen people, on the other hand, the motivation of those involved determines whether AI tools are used consistently at all. If employees expect every hour saved to be immediately filled with new work, a silent incentive arises not to make efficiency gains visible.
What connects the two perspectives
The sources do not contradict each other; they illuminate two ends of the same chain. Outside, in the market, the price for standardisable services is falling. Inside, in the business, capacity is being created whose use must be negotiated. Anyone who looks at only one end makes bad decisions: pure cost reduction without a new value proposition leads into price competition; new offers without internal clarification lead to overload.
A pragmatic approach consists of a few steps:
- Sort services: Which parts of your own work are production, which are judgement, responsibility and context? Only the first part is under price pressure in the short term.
- Processes before tools: AI accelerates above all workflows that are clearly described. Unclear processes do not become better through automation, but faster wrong.
- Determine how the gain is used: Decide in advance what share flows into margin, delivery time and relief – and communicate it.
- Plan quality assurance: Anyone who is responsible for results needs review steps. This review is itself part of the value proposition.
Consequences for web offerings
That the head of a website builder provider of all people emphasises the opportunities is telling. Getting started with your own online presence is becoming cheaper and faster, which shifts the gap between self-build and professional implementation. The value of agency work then lies less in creating pages than in integration, data modelling, accessibility, operation, security and further development over years.
This applies in particular to companies with grown systems – for example a TYPO3 site with connected specialist applications. There, the productivity gain does not arise from faster copywriting, but from clean structures: maintained content models, documented interfaces, traceable approval processes. Only on this basis can AI-supported steps be reliably integrated.
For web and software projects, this means: the interesting question is no longer whether AI is used, but where in the workflow it applies and who benefits from the effort saved. Projects should therefore also negotiate the process side – which steps are eliminated, which checks remain, which capacity is freed up.
Anyone who openly clarifies this distribution gets tools that are actually used, instead of an efficiency calculation that only works on paper.