1,676 restaurant and fast food websites scanned, 1,037 reachable. 34% of the reachable ones turn an AI crawler away at the edge, 42% take a booking or an order on their own domain, 304 score A or B. The month is still running; the page updates hourly.
sites scanned
turn an AI crawler away
book or order on their own domain
score A or B
Out of 424 restaurant and fast food websites scanned in Lisbon for October 2026, our scanner reached 295, while 129 were unreachable. Among the reachable sites, 83 turn at least one AI crawler away at the edge, which is 28% of the reachable sites, and 16 block an AI crawler in robots.txt. Looking at customer access, 143 sites take a booking or an order on their own domain, which is 48% of the reachable sites, with 120 offering booking or ordering on the business's own site, 23 through a vendor on the business's own site, 33 only through a marketplace, 21 through an unrecognised third party, and 98 having no booking or ordering path. Across all scanned websites, 117 earned an A or B together, which is 28% of the scanned sites, with individual grades standing at 39 for A, 78 for B, 64 for C, 59 for D, and 184 for F.
Em outubro de 2026, foram analisados 424 websites de restaurantes e fast food em Lisboa, dos quais 295 estavam acessíveis e 129 inacessíveis para o nosso scanner. Entre os sites acessíveis, 83, ou seja, 28%, bloqueiam pelo menos um crawler de IA no limite da rede, 16 bloqueiam no robots.txt e 143, o que corresponde a 48%, aceitam reservas ou pedidos no seu próprio domínio. Na avaliação global, 39 sites obtiveram A, 78 B, 64 C, 59 D e 184 F, com as notas A ou B juntas a somarem 117 sites, o que representa 28% do total verificado. Ao consultar assistentes de IA sobre onde fazer reservas ou pedidos nos negócios com melhor pontuação, a percentagem de respostas que direcionaram para o próprio site do negócio ou para um fornecedor direct foi de 78% de 150 respostas no ChatGPT, 82% de 150 respostas no Claude e 93% de 150 respostas no Gemini.
Written by Gemini from the numbers on this page only, 2026-10-14; every number in it is checked against the page before it is shown.
| F | 952 |
| C | 211 |
| D | 209 |
| B | 197 |
| A | 107 |
| none | 467 |
| HTML page | 406 |
| 105 | |
| on the page | 33 |
| elsewhere | 26 |
| none | 444 |
| own site | 379 |
| marketplace | 82 |
| third party | 78 |
| vendor on own site | 54 |
| Turn at least one away at the edge | 348 |
| Block one in robots.txt | 42 |
| Unreachable from our scanner | 639 |
Each site counted once, by its latest full scan before the month ended. Menu and channel counts are over the 1,037 reachable sites.
Each assistant was asked, with web search on, for the website, where to book and where to order, for the top-ranked businesses. Latest answer per business and question before the month ended.
| Assistant | Own site or direct vendor | Marketplace | Directory | Nowhere useful | Answers |
|---|---|---|---|---|---|
| ChatGPT | 81% | 15% | 1% | 3% | 150 |
| Claude | 85% | 13% | 1% | 1% | 150 |
| Gemini | 93% | 6% | 0% | 1% | 150 |
| Assistant | Own site or direct | Marketplace | Directory | No link | Places |
|---|---|---|---|---|---|
| ChatGPT | 100% | 0% | 0% | 0% | 41 |
| Claude | 82% | 0% | 5% | 13% | 38 |
| Gemini | 88% | 0% | 5% | 7% | 41 |
Every business of the index's kind with a website of its own is scanned the same way, about once a month, and every scan is kept. This page takes the latest scan of each site before October 2026 ended, so a site fixed in the following month shows up in the following month's page, not here. A past month does not change. The assistant tables use the latest answer per business and question before the month ended. Methodology · the live index · all cities.