Every Observatory entry follows the same format, so two sectors or two dates can be compared.

The format

  • Twelve unbranded buying questions, written the way a customer would type them: four discovery, four comparison and four decision or pricing questions.
  • Two AI platforms, the same across the whole series, with Spain as the market and Spanish as the language. Each capture states whether web search was enabled in the session.
  • Two dates seven days apart. One round describes one day. Two rounds tell you whether what was observed was stable or changed.
  • One table per sector with the brands mentioned in each answer, their order when a list exists, the domains cited and a label: stable, variable or not assessable.
  • A three-paragraph reading: which brands appear, which sources repeat and which gaps seem to belong to the absent brands.

What an entry does not prove

  • It is not market share. "Appears in 9 of 24 answers" describes a sample of twenty-four, not what every buyer sees.
  • It does not explain why a brand appears. A source cited next to an answer does not prove it caused it.
  • It does not predict next month. That is why every sector is measured again.

Rules

  • Public questions and public answers only. No client accounts or private data are used.
  • Brand names are transcribed exactly as they appear in the answer, without adjectives of ours.
  • The sample, the dates and the platforms are always visible next to the table.
  • Correction channel: if you represent a mentioned brand and spot a transcription error, write to us through the contact form with the entry, the row and the correction. We review it and log the change with a date.

Planned sectors

Online accountants for freelancers · invoicing software and Verifactu · CRM for small businesses · dental clinics in Madrid · business schools and online master's degrees · health insurance · specialist travel agencies · hosting and domains in Spain.

Entries are published when the measurement exists. Until then, this page only shows the method: we do not fill the gap with numbers we have not observed.