How to escape the price war between agencies: sell what the client can't compare
If you've lost a client this year over €100 difference in the retainer, you already know how this film ends. Your SEO proposal, the one from the agency across the road and the one from the freelancer with a shared Semrush login all say roughly the same thing: audit, keywords, content, link building, monthly report. The client, who can't assess the technical quality of any of the three, does the only rational thing from where they sit: sorts by price and pushes downwards.
That's not bad luck or clients "who don't value the work". It's a market mechanic with a name, and understanding it is the first step to getting out of it.
The mechanics of commoditisation (why it happens to you and not to your plumber)
A service becomes a commodity when three conditions are met at once:
- The buyer can describe the service without your help. "I need SEO" is already SME vocabulary. When the client knows how to ask, they know how to ask for three quotes.
- The proposals are comparable line by line. Same deliverables, same tools cited, same timelines. The comparison table builds itself.
- A known market price exists. In Spain, SEO ranges between €600 and €4,000/month depending on scope (pacoruben.com). The client who has asked for two quotes already has the anchor, and every proposal gets negotiated downwards from it.
When all three are met, price is the only variable the client can move with confidence, so they move it. And the usual defences don't break the mechanic: lowering your price puts you in the spiral (there's always someone willing to charge less than you), and "differentiating on quality" is invisible at the moment of signing, which is when it's decided. Quality is proven at six months; the quote gets compared today.
The way out: sell what has no comparison table
The price war only exists where comparison is possible. So the way out isn't to compete better within the table: it's to sell something that doesn't have a table yet.
A service escapes comparison when it's missing one of the three conditions above. And there's a whole category of service that today is missing all three: visibility in AI search engines — getting the AI to recommend your client when someone asks ChatGPT, Gemini or Perplexity; the discipline is called GEO, Generative Engine Optimisation.
Run through the conditions:
- The buyer doesn't know how to ask for it. They know "the ChatGPT thing" worries them — frequent use in Spain has gone from 4% to 28% in two years (Funcas, 2026) — but they don't know how to translate that worry into a spec. Whoever gives the problem a name and a method defines the conversation.
- There are no proposals to compare. Most agencies don't offer it yet. When your proposal is the only one on the table, it isn't compared: it's accepted, negotiated or rejected, but against the problem, not against another agency.
- There's no settled market price. This is the pioneer's advantage, and it's worth saying plainly: AI visibility has no "street price" today. Nobody will tell your client "well, mine do it for 300". That window won't last forever — no new service keeps it — but today it's open, and whoever sets prices during the window also sets the references they'll be compared against later.
And is there substance behind it, or is it smoke to escape the table? There is, and it's measurable: we tested the visibility of 9,865 Spanish SMEs across the 4 main AIs and 91% only appeared in one (the study, here). The problem is real, massive and almost nobody is working on it. For an agency, that's the definition of a virgin market.
How to anchor the value when there's no reference price
No market price doesn't mean charging at random. It means you set the anchor, and there are three solid ways to set it:
1. Anchor against the sister service. SEO already has a known range (€600-4,000/month). Presenting AI visibility as a sister discipline — "it's working your presence where people ask now, just as SEO works it in Google" — inherits price legitimacy without inheriting the war. The specific ranges that are working in Spain, by client type, we develop in how much to charge for AI visibility services.
2. Anchor against the cost of not being there. Traffic that arrives from AI answers converts at 14.2% versus 2.8% for classic organic (sector data compiled by roymo.es). Every month the client doesn't appear, the recommendations go to whoever does. That frame turns your retainer into insurance against a concrete loss, not just another marketing expense.
3. Anchor with a closed package, not hours. The agency hour is the most comparable unit there is; the productised package ("weekly tracking across 4 AIs, fixes on sources, monthly PDF report") can't be broken down into a table. Productising is also what turns the service into recurring revenue that scales: the full MRR maths is here.
And the cost structure backs you up: with Surfeo for agencies, the base account is €20/month and each client costs €35 or €79/month depending on tier, with each client getting the full product. A service charged in the point-1 ranges over those costs leaves a margin that commoditised SEO hasn't left you in years — you show it with the calculator in front of the internal client (your partner), you don't exaggerate it in front of the external one.
The strategic honesty: the window is closing
Let's be clear about the small print, because this argument only works if it's true: the absence of a market price is temporary. In two or three years, "AI visibility" will have its street range, its comparison sites and its own price war. The pioneer's advantage isn't escaping commoditisation forever; it's arriving at it with the roster built, the case studies published and the reference position from which you compete on value even in mature markets. Whoever entered SEO early in 2008 still charges above average today. The mechanic repeats.
If what holds you back is the underlying doubt — what if this is a fad? — that objection deserves its own answer with data: is GEO another fad like NFTs?
Frequently asked questions
Isn't this just "jump on the latest trend"?
No, and the difference is checkable: there's real, massive use (28% frequent ChatGPT use according to Funcas; 37.9% of the population used generative AI in the last quarter of 2025 according to the INE) and a problem measurable in your own clients' data. Fads without users don't pass that filter.
What if a competitor starts offering it cheaper?
During the window, your defence is the cases: dated screenshots of clients who didn't appear and now do. The second to arrive competes against your evidence, not your rate. That's why it matters more to start building history now than to nail the perfect price on day one.
Can I use this logic with services other than AI visibility?
The mechanic is general: any service where the client can't compare proposals escapes the price war. What makes AI visibility special is the combination of growing demand, almost no supply and low tool cost. The comparison with other new services for 2026, with numbers, is here.
How do I present the service without seeming to sell fear?
With data from the specific case, not headlines. An audit of their real visibility — where they appear, where they don't, what the AIs say about them — turns the abstract argument into their own, verifiable problem. Generic fear sells badly; your own data sells well.
The first step to selling what can't be compared is having the data nobody else brings to the meeting: run a client's website (or a prospect's) through the free AI visibility test and see what the AIs answer today.
