# Why Google can't read your PDF menu > A PDF menu is a picture as far as search is concerned — no dishes, no prices, no allergens. Here is what that costs you and what a readable menu looks like. > > Published: 2026-08-04 > by The Menugram team · Menugram · https://menugram.bg/blog/en/why-google-cannot-read-your-pdf-menu > Topics: seo, qr-menu ## Key takeaways - Google indexes a PDF as a document, not as a menu — not one dish, price or allergen inside it becomes structured data a search engine can surface. - A QR code that opens a PDF makes the guest pinch and drag; many give up before they reach the prices at the bottom of the page. - A menu on its own URL, with real HTML text and Schema.org markup, can rank for a specific dish rather than only for the restaurant's name. - One menu at one address updates once and instantly, instead of exporting a new file and reprinting every QR code on every table. --- You export the menu as a PDF, generate a QR code that points at it, and stick the code on the tables. It works — a guest scans and sees the menu. So far, so good. The problem is what happens **away** from the table. Someone who has not chosen where to eat yet searches, or asks an assistant. In both cases your menu is not part of the conversation. ## What a search engine actually sees Google does index PDFs — that much is true, and it gets quoted as though it settles the matter. But it indexes them as *documents*: it pulls out a title, some text, and stops. :::stat 0 The amount of structured data about a dish, a price or an allergen a crawler can extract from a typical PDF menu. The format has nowhere to put it. ::: The difference is between "this file exists" and "this restaurant serves carbonara at £14, no nuts, until 11pm daily". The second is information an engine can answer a question with. A PDF does not carry it in a machine-readable form — even when the letters are right there. And if your menu was scanned or exported as an image, even the letters are gone. What is left is a several-megabyte file that means nothing to a machine. ## What the guest loses Search aside, a PDF is a poor thing to hand a phone. It opens in a separate viewer, the page is A4 on a screen that is not, and every guest's first gesture is to pinch. :::callout Try it yourself Open your current menu on a phone and time how long it takes to read a price without zooming. That is the tax every guest pays at every table. ::: A menu that is readable on arrival is not a cosmetic improvement. It is the difference between a guest who browses the whole list and a guest who orders the first thing they managed to read. ## What a readable menu looks like Three things, in this order: 1. **Its own URL.** The menu is a page — `yourrestaurant.com/menu` — not a file. A page can be linked, shared, indexed and updated. 2. **Real HTML text.** Dish names, descriptions, prices and allergens are text in the page, not pixels in a picture. 3. **Schema.org markup.** `Restaurant`, `Menu`, `MenuItem` and `Offer` tell a crawler which part is a dish and which part is a price, instead of leaving it to guess. The third is what opens up results for a specific dish rather than only for the name of the place. :::quote Menugram A menu is not a file you hand out. It is the page people find you on. ::: ## And then there is updating it There is a more mundane reason too. You run out of something. With a PDF: open the document, edit, export, upload, wait for the cache, and the old file carries on living in somebody's phone. With a page: change the item's status, and the next person to open the menu sees it. The QR code on the table never changes, because it points at an address rather than at a file. :::cta / Menugram does exactly this — a branded menu on its own URL, readable by Google and by your guests, that you update yourself. ::: If you are working out how to lay yours out, [look at a live menu](/tasteplace/menu) and then [how the platform works](/). ## Frequently asked ### Does Google index PDF files at all? Yes. Google indexes PDFs and can show them in results. But it treats them as documents — it extracts a title and some text, and it does not recognise a dish, a price, an allergen or an opening time. For a search like "carbonara near me" the contents of your PDF are effectively invisible. ### What if I put the menu on my site as an image? That is worse than a PDF. An image has no text to extract at all, so a crawler and an AI assistant both learn nothing about what you serve. Alt text helps a little, but it is a caption, not a menu. ### How quickly does this show up in search? Indexing a new page usually takes days to a few weeks. The effect on guests at the table is immediate — the menu opens readable on the first scan, with no zooming. ## Sources - [Google Search Central — JavaScript SEO basics](https://developers.google.com/search/docs/crawling-indexing/javascript/javascript-seo-basics) - [Google Search Central — Restaurant structured data](https://developers.google.com/search/docs/appearance/structured-data/restaurant) - [Schema.org — Menu](https://schema.org/Menu)