Why €8,000 for startup branding is becoming harder to defend

Some economic models remain standing for a remarkably long time after the labour they once rested on has fundamentally changed, and that is exactly what is becoming increasingly visible in branding.

Good designers remain valuable, and so do good brand strategists, while the relationship between time, production and price is shifting faster today than many agency models can follow. Asking a starting company €4,000 or €8,000 for a name, positioning, visual identity, tone of voice and brand book therefore requires more and more explanation when a substantial part of the preparatory work that used to take weeks of human production can now be done in a few days.

The value of branding is not disappearing, while the arithmetic by which that value was traditionally sold is clearly wearing thin.

A startup cannot yet know its definitive brand

In its first months, a starting company has remarkably little reliable information about itself, because much of what later becomes essential to the brand still has to be uncovered by the market.

The founder has an idea, a conviction, perhaps a first product and a few customers, while important questions remain unanswered. Which customers keep coming back? Which service turns out to be truly valuable? Which words do customers use when they explain the company to someone else? Which parts of the offering grow into core activities, and which gradually fade into the background?

That kind of knowledge rarely emerges fully during a workshop, since part of it only becomes visible through transactions, conversations, failures, repetition and the slow emergence of patterns.

Startups therefore develop almost every part of their business iteratively. Products change, pricing models evolve, technology is rewritten and commercial processes are continuously adjusted based on new information.

It would be strange to treat precisely the brand identity as an exceptionally static element, as if the company already had, on its first day, all the knowledge needed to shape the next ten years.

Colours, typography, words and design remain important, since they form the first interface between an organisation and the outside world. A young company that wants to sell professionally can afford little when it comes to visual chaos, inconsistency or an unclear appearance.

That does not create an obligation to immediately build a fully crystallised brand architecture that gives the impression that all strategic answers have already been found.

The first brand identity can perfectly well be regarded as version 1.0, as long as it is sufficiently coherent, professional and recognisable to let the company start credibly.

When product, service and people convince, the brand layer can be sharpened later. A colour disappears, a name evolves, a positioning becomes more precise and a website is rebuilt as soon as it becomes clearer what customers are really willing to pay for.

What changes at that point does not have to be the correction of a wrong choice, since it is often simply information that was not yet available at the start.

Claude changes the amount of labour

The second shift is more technical and probably even more important economically, because generative AI is taking over a growing share of the preparatory knowledge work that for years was billed as human production time within branding projects.

In its 2025 global AI survey, McKinsey reported that 88% of surveyed organisations regularly use AI in at least one business function, compared with 78% a year earlier.

That figure marks an important turning point, as AI is gradually shifting from a specialised technology to a regular instrument for knowledge work.

Within branding, I notice that change above all with Claude.

Of the major generalist AI systems, Claude is currently, for me, the most convincing environment for brand design, brand structure and translating strategic choices into a coherent visual and verbal system. ChatGPT remains particularly strong in broad analysis, creation and multimodal applications, while Gemini has impressive capabilities in visual generation and integrated Google workflows. As soon as strategy, language, brand logic and design have to stay together over a longer project, I currently experience Claude as the most consistent environment.

That difference lies mainly in the way a brand can be treated as a system.

Structuring a brand interview, mapping a market, comparing competitors, summarising hundreds of pages of research, exploring positioning directions, developing naming directions, building a tone of voice, producing copy variants, documenting brand principles and structuring a first brand book now belong to one continuous workflow.

Anthropic is also explicitly moving in that direction with Claude Design, which allows existing design systems and brand guidelines to be used to keep visual output more consistently within the same system.

As a result, for me Claude shifts from a language model to something far more interesting within branding: an environment in which research, strategy, language and visual application are coming ever closer together.

Work that a few years ago required several separate phases and weeks of production time can therefore be prepared within days when someone works in a focused way on one brand.

That by no means implies that all choices must also be made at the same speed.

Technology can compress labour, while taste and judgement often need time.

Time and value increasingly diverge

Forrester reported in 2026 that nine out of ten US marketing agencies use generative AI, while for 81% of those agencies higher employee productivity is among the main objectives. In addition, 74% use generative AI to summarise documents and communication, while 70% use the technology for research and competitive analysis.

McKinsey also estimates that generative AI in marketing can create productivity value equivalent to 5 to 15% of total marketing spend, which amounts to roughly 463 billion dollars a year worldwide.

The exact percentages will keep shifting, while the underlying economic movement is far more fundamental.

Work that used to require human time by necessity is increasingly becoming partly computational work, and computational work simply follows a different cost curve than human production.

When a competitive analysis used to take two days and can now be prepared within a few hours, the economic meaning of that activity changes.

The same applies to desk research, copy variants, first positioning models, documentation, conceptual exploration and other parts of brand development that used to be directly tied to human capacity.

That does not reduce the value of the result, since it is mainly the relationship between time spent and value created that becomes less reliable.

Forrester describes exactly that tension, with agencies realising considerable efficiency gains through AI while their commercial models are still largely built around people, hours and capacity.

Those models emerged in a world where more output usually required more human labour, while generative AI is increasingly loosening that historical link.

WPP now also writes about commercial models that rely more on value, subscriptions and outcomes, because classic agency capacity can less and less self-evidently serve as the primary unit of price.

The relevant discussion thereby shifts from production speed to economic value creation.

As soon as production becomes more abundant, scarcity moves to the ability to distinguish relevant possibilities from irrelevant ones and to arrive at a defensible choice out of that abundance.

The client therefore pays less and less for the existence of options and more and more for the quality of selection.

Faster production does not mean faster decisions

At the same time a curious paradox emerges, because AI can produce an extraordinary number of variants without human decision-making having to follow the same acceleration.

A system can explore dozens of colour combinations, naming directions and positionings in a single afternoon, while that gives no reason whatsoever to decide that same afternoon which colour, name or positioning will be attached to a company for years to come.

Some choices need time because people can rarely judge meaning, recognition and conviction fully in real time.

A name has to wander around for a few days, an identity has to work on a website and at the same time hold up on a quote, presentation, building, invoice and LinkedIn profile, while a phrasing that seems particularly clever in the evening can feel surprisingly ordinary the next morning.

AI mainly shortens the distance between a question and a set of possible answers, while the distance between those possibilities and a good decision remains partly human.

That distinction becomes more important as technology gets faster.

A branding project that used to take three months can technically perhaps be prepared within a few weeks, while the quality of the final decision may actually benefit from a certain delay between analysis, observation and choice.

The computer has no need for incubation time, while people often only recognise which possibility truly holds up after some distance.

The founder ultimately has the vote

Within startup branding there is also a reality that is sometimes surrounded by a remarkable amount of methodology, while its essence remains fairly simple.

The founder ultimately makes the decision, because a young company in its first years almost inevitably carries a large part of the convictions, taste, ambition and personality of its founder.

Every personal preference of course deserves strategic testing, since a brand still has to function within a market, against competitors and in interaction with real customers.

At the same time, a brand that someone has to represent for years has little chance of lasting when that person barely recognises themselves in the language, images and design used on their behalf.

Family, employees, friends and advisers can also play a role, especially because people in one's immediate circle often sense quickly whether a name, colour or phrasing still matches the personality of the person building the company.

Their judgement does not have to replace representative market data, while it can certainly yield relevant information about recognition and credibility.

The task of a brand strategist therefore consists of connecting personal taste with market knowledge, positioning, communication and design, so that subjectivity is channelled rather than defined away.

Meanwhile Claude can generate hundreds of possibilities, while the real work only begins once that abundance has to be reduced to a direction that is substantively right.

Which possibility deserves further attention, which direction merely looks new, which phrasing can a company truly live up to, which trend can safely be ignored and which choice feels distinctive without being forced?

That is where a form of value lies that is much harder to automate than production.

As production becomes cheaper, scarcity shifts to the ability to recognise the right direction within an abundance of possibilities and to take responsibility for it.

For large organisations the problem changes completely

For larger organisations the economic and strategic context changes completely, which makes it impossible to apply the same reasoning one-to-one.

There, €10,000 for a serious rebranding can even be very limited, because an existing company never starts from a blank page.

There is history, customers know certain words and symbols, employees are familiar with specific habits, sales uses existing arguments, HR has its own interests and management works from strategic objectives that often span several years.

Sometimes there are also multiple countries, business units, sub-brands and target groups, all connected to the existing identity in a different way.

A new colour in such an environment is rarely just a new colour, since it can affect hundreds of documents, digital products, buildings, campaigns, presentations, vehicles, templates and processes.

In larger organisations, rebranding also often runs in parallel with another fundamental change, such as a merger, reorganisation, internationalisation, a changed product portfolio, a new strategy or a change in management.

Branding thereby becomes far more an organisational question than a purely creative project.

Which brand value must be preserved, which historical elements still have economic meaning, which terminology lives internally, which customers recognise themselves in the existing brand and which stakeholders stand to gain or lose something from a change?

AI accelerates a considerable part of the research there too, since large volumes of interviews, documents, customer feedback and market information can be analysed and structured much faster.

Organisational complexity, meanwhile, stubbornly persists, however tempting the idea that enough good prompts can ultimately reduce every problem to a neat diagram.

Gartner reported in its 2026 CMO Spend Survey that marketing organisations spend on average 15.3% of their budget on AI, while only 30% of surveyed CMOs say they are mature enough to truly apply AI at scale.

Buying technology and changing an organisation with it thus turn out to be two different competences, the second of which demands far more than access to good systems.

That is why the role of a strong brand strategist within larger organisations is likely to become broader rather than narrower.

That person has to understand communication, read organisational politics, assess design, guide decision-making, interpret research and at the same time have enough technological literacy to know which tasks AI can accelerate and where human responsibility remains necessary.

The economics of branding shift towards judgement

Anyone who concludes from this development that branding should henceforth be cheap confuses the falling cost of production with the lasting value of direction, selection and responsibility.

Production becomes cheaper, analysis becomes faster and the number of possible directions that can be explored in the same time increases sharply, so that economic value automatically moves to other parts of the process.

Less value lies in producing twenty pages of competitive analysis and more in understanding which competitor is truly relevant to the brand's strategic position.

Less value lies in generating a hundred slogans and more in recognising the phrasing an organisation can still credibly defend five years later.

Less value lies in accumulating billable hours and more in the quality of the decisions that ultimately come out of all that research.

Forrester warns of an important side effect here, because agencies that use AI solely to make production faster and cheaper run the risk of efficiency coming at the expense of creativity and distinctiveness.

That danger is logical, since a technology that makes mediocre production a hundred times faster can above all produce mediocrity a hundred times faster when nobody takes responsibility for selection and quality.

The real gain arises once the freed-up human capacity is used for work that remains harder to automate, such as understanding, assessing, choosing, combining and taking responsibility for the outcome.

For startups, that shift has a relatively simple consequence.

A young company mainly needs a solid starting point, a brand identity that is professional enough to give customers confidence, coherent enough to build recognition and flexible enough to change as soon as the market returns better information.

This creates a more realistic model of branding, in which version 1.0 does not represent a provisional failure, but the first usable answer to the limited amount of knowledge available at that moment.

In larger organisations the difficulty lies elsewhere, since there less and less is paid for merely producing a new brand identity and more and more for responsibly changing something to which revenue, reputation, employees, history and internal power relations are tied.

AI will keep accelerating both processes, while speed on its own has never guaranteed that an organisation is also moving in the right direction.

Price mainly follows the number of interests

The cost of branding is often presented as if it were directly tied to the number of logo variants, pages in a brand book or hours a designer needs. In reality, the price is determined far more by the number of people who influence the decision and by the interests attached to that decision.

In a startup with one founder, a choice about name, colour, positioning or tone of voice can be made relatively quickly. There is one clear line of decision-making, little history and usually little internal politics. The analysis can be thorough, while the final decision-making remains compact.

As soon as several stakeholders are involved, the character of the project changes. A marketing manager looks at brand consistency, sales at commercial usability, HR at employer branding, the board at strategy, finance at costs, legal teams at risks and local teams at their own market context.

The cost of branding therefore rises mainly through coordination, alignment and decision-making.

After all, every additional stakeholder adds new information, but also new preferences, objections, risks and dependencies. A brand project with eight decision-makers therefore requires a different process than one with a single founder, even when the final visual identity looks just as simple on paper.

AI can help structure that complexity. Interviews can be processed faster, conflicting viewpoints can be clustered and scenarios can be prepared more efficiently. The need to weigh interests against each other remains, however, because a brand decision within larger organisations is rarely only about aesthetics.

That is why I would tie branding prices less to creative output and much more to the complexity of the decision-making process.

A simple rule of thumb is that the price rises as the number of stakeholders, markets, internal interests and existing brand assets involved increases.

For a startup with one or two decision-makers, a compact project can suffice.

For an organisation with a board, marketing, sales, HR, legal teams, several countries and existing brand architectures, a much heavier project naturally emerges.

The cost then lies less and less in drawing the brand and more and more in organising agreement on what that brand should mean from now on.