In this guide you’ll learn:

• How AI assistants decide which ceramic brands to recommend – Why AI visibility is a positioning problem before a technical one.
• The content and brand signals AI systems can actually cite.
• How to measure whether AI recommends you — and fix it when it   doesn’t.

The shift is measurable and fast. AI-generated overviews already appear in about one in four Google searches (Conductor, 2026), and referral traffic from AI assistants is multiplying fivefold year over year (Similarweb, 2026). Behind those numbers sits a stranger one:

AI platforms read thousands of pages for every visit they send back (Cloudflare Radar, 2026). Influence now happens without a click — your brand is being evaluated, summarized and recommended (or omitted) in conversations you will never see in your analytics.

What decides the recommendation is not advertising spend and not even rankings. AI systems synthesize what is published: they favor brands whose identity is unambiguous, whose claims carry verifiable evidence, and whose name appears consistently across sources the system trusts. In other words — they favor brands that have done their branding homework in public.

That is the strategic reframe this guide rests on: everything that makes a brand recommendable to a machine is what makes it credible to a human. The work is the same work. AI just removed the places to hide.

Ask an AI assistant about a well-positioned competitor and it will repeat their positioning almost verbatim. Ask it about a brand that never wrote its story down and it will improvise something generic — “a Spanish manufacturer of ceramic tiles” — or say nothing at all.

AI does not invent your differentiation. It repeats the one you published, or replaces you with someone who did.

2.1. The written positioning

One clear, distinctive answer to “who is this brand, what does it make, for whom, and why does it matter” — present on your site, in your about page, in your export materials, everywhere. This is verbal identity doing commercial work: naming, message architecture, a point of view a machine can quote and a buyer can remember.

2.2. Consistency across the web

The same company described three different ways in three directories reads, to a machine, as three weak entities instead of one strong one. Audit how your brand is named and described everywhere it appears — fairs, associations, press, distributor sites — and align it. Tedious, unglamorous, decisive.

2.3. Structure for machines

Schema markup, clean page hierarchy, data with visible sources. Your developers implement it in days; what they cannot invent is the clarity it encodes. Structure carries the positioning — it never replaces it.

AI systems assemble answers from citable statements: definitions, data with sources, frameworks, honest comparisons. A brand that publishes real knowledge becomes raw material for thousands of answers; a brand whose site only says “quality and innovation since 1985” gives the machine nothing to work with.

For a ceramic manufacturer, citable expertise looks like: application guides architects actually use, technical content that answers real specification questions, project stories with facts, and market points of view backed by data. It does not look like a calendar of generic posts — the same content discipline that wins specifiers, covered in our guide to digital marketing that works, is exactly what wins citations.

One honest note on craft: AI answers are text, but the buyer’s journey is not. The architect who gets your name from an assistant lands on your site thirty seconds later, where photography, catalogue and identity take over the argument. Machine visibility opens the door; design culture decides what happens after it opens.

Machines trust brands the way markets do: by cross-checking. A claim on your own site is a statement; the same claim reflected in sector trade press, fair directories, association listings and distributor sites becomes evidence. Brand mentions across trusted sources now weigh more than links alone in how AI systems select who to name.

This is communication strategy doing new work. The sector map that built reputations for decades — Ceramic World Review and its peers, Cersaie, Coverings, Obra Blanca, CeramBath, Unicera, the associations — is precisely the map AI systems read. Earned presence there was always brand-building; now it is also machine evidence. The disciplines did not change. The audience gained a very literal reader.

You cannot manage what you never ask. Once a quarter, run the same battery: ask the main AI assistants which brands they would recommend in your category, for your applications, in your target markets. Ask in English, and in the language of your priority export markets. Record who appears, who does not, and what the machines say about you.

Then treat gaps as diagnosis, not mystery. Not mentioned at all: a legibility and publication problem — back to Steps 2 and 3. Mentioned with a generic description: a positioning problem — the machine has nothing distinctive to repeat. Mentioned behind competitors: a corroboration problem — their footprint outweighs yours. Each gap maps to brand work, which is the point of this entire guide: the answer you want the AI to give is the brand you have to build.