What is quietly happening to communication agencies in mid-sized towns

In a town of forty-five thousand people, a six-person agency sent out a quote of three thousand eight hundred euros for a brochure website. The client never replied. Three weeks later, his site was live: he had built it himself, on a Sunday afternoon. This scene has been repeating itself everywhere for the past eighteen months, and it deserves better than another obituary for the agency business. Because that is not what is happening.

The typical agency, and how it really made money

Let us take that agency seriously. Six people: an owner who sells and steers, an art director, a front-end integrator, a part-time developer, an account manager, an apprentice. Around sixty local clients: tradespeople, two car dealerships, a veterinary clinic, the chamber of commerce.

Its model stood on three legs:

The brochure website: invoiced between three and six thousand euros, delivered in six to ten weeks.
Maintenance: an annual retainer of around six hundred euros, rarely used in full.
The rebuild: every four or five years, it reset the meter to zero.

That model held for fifteen years because it rested on one assumption nobody bothered to question: producing a website took several trades, several weeks, and therefore several thousand euros. That assumption no longer holds. It is the only thing that changed, and it takes everything else with it.

What actually broke

The problem is not that clients became disloyal. It is that the reference price collapsed in their heads before it collapsed on the market. A builder who has watched a tool produce a decent page in a few minutes no longer hears a four-thousand-euro quote the same way, even when that quote is perfectly justified.

The agency then finds itself defending its price instead of defending its work. That is a losing position: you never win an argument in which you explain why you are expensive.

Worse: what it defends in that argument is almost always production. Design time, build time, testing time. And that is exactly the part the machine has just made nearly free. The agency is defending the wrong asset.

The five trades production is absorbing

This deserves a cold look, trade by trade, because the picture is neither uniform nor total.

The graphic designer: on a brochure site, the visual identity often boils down to three colours, two typefaces and a grid. That work can now be inferred from a one-sentence brief. What resists is real brand identity, the kind that has to work on a van, a shopfront and a package at the same time.
The web designer: the layout of a brochure site has followed the same structure for ten years. A banner, an offer, a proof point, a contact form. Reproducing it cleanly, phones included, is no longer a rare skill.
The front-end integrator: this is the trade hit most directly. Turning a design into pages that render correctly everywhere was slow and thankless work. It has been absorbed almost entirely.
The developer: on a brochure site, he was called in for a form, a map, an online booking slot. Those building blocks are now plain-language requests. On a business application, by contrast, nothing has changed.
The person who keeps the servers running: the one nobody talks about, and the heaviest cost of all. A server, a domain name, a certificate, backups, security updates, and someone on call when the site goes down on a Saturday.

That last point is worth underlining: plenty of small agencies were already outsourcing hosting, often badly, and billing it at a loss.

What the machine does not take

No tool knows what the owner of that agency knows. He knows the car dealer never picks up before eleven. He knows the veterinary clinic wants to look reassuring and above all not modern. He knows which shopkeeper pays on the spot and which one will ask for three credit notes.

He also knows what a client means when he says "I want something understated". Turning a vague request into a decision is a craft. That craft has not moved an inch.

And finally, someone has to answer the phone when the site goes down. Not a support centre, not a web form: someone the client calls by their first name. Outside the big cities, that closeness is worth more than the technology.

The turn

Here is where the reading flips. If production is worth nothing any more and the relationship is worth just as much as before, then an agency has no reason to disappear. It has a cost structure to invert.

In practice, one person can now hold the entire chain, from the first visit to the client's premises to a live website with its domain name, its server and its backups. Not because that person learned five trades, but because those five trades have become a conversation.

The owner describes the project the way he would describe it to a colleague. He gets a page. He corrects it with sentences, not with software. He puts it live. The client sees only one thing: his agency delivered his website.

The economics of this model

This is where the numbers get interesting, and they are public.

With Leopar, a website with its own server, its domain name, daily backups and unlimited changes costs €298.80 a year. An agency that resells that service to its client as a €100 monthly retainer, which is still less than the same client used to pay in maintenance plus a rebuild spread over the years, clears roughly €901 of margin per site per year.

Across thirty clients, that is €27,030 of recurring annual revenue. No production, no subcontractors, no technical on-call duty. Across sixty clients, the arithmetic speaks for itself.

This is not the same business as before. It is a subscription business, not a project business. It earns less per transaction and far more over time. Above all, it does not collapse when one big pitch goes the other way.

The part nobody likes to write down

This shift has a human cost, and it would be dishonest to skip over it. A six-person agency that adopts this model no longer needs six people for production. It needs people who sell, who advise, and who hold the relationship together.

Career changes do happen, and they are real. A front-end integrator who knows the web often makes an excellent project manager, precisely because he knows what can actually be built. But they are not an option for everyone, and nobody should pretend otherwise.

So the question is not whether this shift is desirable. It is under way. The question is who, in a mid-sized town, will capture the value that remains: the trust of the shopkeepers down the road.

Nothing stops you from reselling

One last point, rarely said out loud. Nothing prevents a professional from taking out a subscription on behalf of a client, folding it into their own offer and keeping the difference. Everything produced is licensed to them for free use, modification and hosting, including somewhere else entirely.

In other words, an agency can build an offer under its own name on technology that is not its own. That is exactly what agencies have been doing for thirty years with servers, content management systems and email tools. The only novelty is that the outsourced block no longer stops at hosting: it now covers the whole of production.

Some are already doing it, quietly. They do not advertise the fact, for an obvious reason: they have no interest in their clients finding out.

Frequently asked questions

Can an agency resell a website under its own name?

Yes. Everything produced is licensed non-exclusively, with free use, modification and hosting, including with another provider. The agency remains Leopar's customer and invoices its own service to whoever it likes. The details are set out in the terms and conditions.

Do you need technical skills to manage several client websites?

No, and that is the whole point. The server, the domain name, the certificate, the backups and the security updates are included and handled for you. Changes are requested in plain English, by message.

Does this apply to complex projects too?

No, and that needs saying. A business application, a catalogue of several thousand products or an integration with management software are still development projects. The reasoning in this article applies to brochure sites and small online shops, which make up the bulk of a local agency's portfolio.

In short

The agencies struggling today are the ones that sold production time. The ones that will come through are those that were already selling something else without always realising it: the understanding of a trade, the trust of a business owner, the certainty that someone will pick up the phone.

That part cannot be automated. The rest just became so.