What expansion teams do with Claro, one sentence at a time

What every use case has in common

Every use case starts from the same input. You type one sentence in any language, or you upload the brief you already have as a PDF, Word, Excel or PowerPoint file, or as a photo of a printed page.

Claro shows the criteria it read as chips before you look at any result. If a chip is wrong, you correct it in words.

Each criterion then gets its own verdict on each premises: met, not met or not stated. The evidence sits beside the verdict. It is the listing's own sentence, a measured figure with its source, or a note that the listing does not say.

“Not stated” means the listing does not say. Claro never reads silence as a yes or a no; a listing that does not mention a smoke outlet (salida de humos) is marked not stated, not not met.

Location figures (footfall, residents, age, income, spending power, places nearby) come from a Casafari group company and are shown where available. How Claro counts premises and keeps each fact with its sentence is set out in the methodology.

• Input: one sentence in any language, or a file of up to 10 MB.
• Reading: the criteria shown as chips you can check and correct.
• Answer: met, not met or not stated per criterion, with the evidence beside it.
• Location figures: from a Casafari group company, where available.
• Access: one free search with no sign-up; access adds more searches, every result, dossiers and side-by-side comparison.

Which use case is yours?

Find your situation below. Each line reads: your situation, then the use case, then an example brief to copy and edit.

If you are writing a brief from scratch, the guide on how to write an expansion brief has a template.

• QSR or burger chain adding units in a city it already serves → Keep away from your own stores → “Burger restaurant in Madrid, street level, 150–250 m², smoke outlet, not within 500 m of our own stores, no burger restaurant within 300 m”
• Gym operator → No competitor nearby → “Gym in Valencia, 600–1,200 m², ceiling at least 4 m, no gym within 500 m, at least 15,000 residents within a 10-minute walk”
• Coffee group → Footfall and people that fit → “Coffee shop in Málaga, corner unit with a shop window, 60–120 m², at least 12,000 people a day walking past, at least 25% of residents aged 15–29”
• Pharmacy or clinic → Underserved areas → “Pharmacy in Zaragoza where many people live and there is no pharmacy within 600 m, street level, step-free access”. Claro applies distances and residents; pharmacy opening rules are not among its checks.
• Fashion retailer → All of Spain in one brief → “Fashion store anywhere in Spain, street level, 150–300 m², frontage at least 8 m, shop window, at least 20,000 people a day walking past”
• Franchise network → Upload the local tipo (the network's standard premises spec) → upload the file, or type “Street level, 80–150 m², open plan, licence for cafetería, Valencia or Alicante, rent under €4,500 a month”
• Broker or consultant searching for clients → Facts from the listing itself → “Restaurant in Seville, street level, 150–300 m², smoke outlet, food-service licence, rent under €6,000 a month”, one brief per client

How to read the example verdicts

Every verdict, quote and figure on the cards above is an example. The cards show the format of an answer, not real premises.

Each verdict has its own mark: a tick for met, a cross for not met and a dashed circle for not stated. The evidence beside it says where it comes from: the listing, location data or your store list.

• Competitor distance: 1–1,000 m in a straight line, for nine segments: gym, burger, fast food, coffee, bakery, pharmacy, supermarket, fashion and pizza. With access, the top 20 results are re-checked on foot. More on premises with no competitor nearby.
• Underserved areas: premises are sorted by residents within a 10-minute walk per similar business within 600 m, and those inputs are shown on every result. “None found” is a finding from the data, not a guarantee. More on underserved areas.
• Own stores: with access, your store list is uploaded through the API and stays private to your account. On the free search this criterion shows as locked, and the rest of the brief still runs. More on keeping away from your own stores.
• Not stated: the listing does not say. Visit a not-stated premises with a question; skip one marked not met.

Why every answer shows its evidence

Site-selection work in Spain usually follows one of three approaches. Property portals filter by fields such as price and size. Geomarketing tools colour areas on a map. Consultancies write a study.

Claro works from the premises on the market. Claro reads each listing's own words, adds measured location figures and gives a verdict per criterion with its source, so each answer can be checked.

Claro does not forecast sales. It checks the criteria you wrote. Each premises links only to the agency's own listing page, and private sellers are never named.

The guide to site selection tools in Spain compares what each approach answers and what to ask first.

For agents and AI assistants

Every use case on this page also runs without a person at the keyboard. With access, Claro is available as a REST API with an OpenAPI description and as an MCP server. AI assistants and internal tools run the same brief-to-shortlist search and read the same evidence.

An agent can run the one free search like a visitor: POST https://expansion.casafari.com/api/try with a brief in plain words (1,500 characters at most) or a file of up to 10 MB. The answer contains the brief as understood (chips) and up to six premises, each with a verdict per criterion and the agency's listing link.

The site's llms.txt file, at the root of expansion.casafari.com, describes both endpoints and how to request access for a person, with their consent.

• search_premises: run a brief
• refine_search: adjust a search in words
• get_premises: the dossier for one premises
• compare_premises: premises side by side
• explain_location: the location figures around a premises