# PDF Menus Brochures and Price Lists: When to Add HTML By AI Visibility editorial team AI-assisted article. A Dubai restaurant publishes a menu as a printable PDF. On paper, each price sits beside its dish. When someone copies the text, words from the next column land on the same line. The restaurant’s booking page still works, but a customer comparing two dishes on a phone has to open the file, zoom and work out which price belongs to which item. A brochure or a price list can produce the same problem when its layout carries meaning that plain text loses. Choose the facts customers need to make their decision, then see whether they can find and read those facts in the current format. Keep a useful download for printing, sharing or formal documentation. Add HTML where a stable, maintained page gives the reader a clearer route to key information. The AI Visibility access checklist covers general readability. This walkthrough compares a published Dubai menu PDF with a separate HTML menu and shows how to decide what belongs on a web page. AI Visibility site access checklist (/article/can-ai-search-access-your-website-a-practical-checklist/) Decide what the customer needs from the document List the questions that lead to a booking or enquiry. A menu visitor may need the branch, the service period, a dish name and the relevant price or option. A brochure visitor may need to distinguish what is included in two packages. A price-list visitor may need the unit, currency, effective date and conditions that change the amount. Pick a few facts and note what belongs with each one. An amount means little if its unit or qualifying condition appears several pages away. Then identify why the business offers the download. A printable menu, a formal specification or a document that a buyer shares internally may remain useful. The web page can give the customer a faster route to the decisive facts and link to the detailed file. Where the full list is too large to maintain by hand in two formats, agree on a common data source or a narrower HTML summary with a clear scope. Customer question Useful HTML Useful role for a PDF Can I order this at my branch? Branch, service period and relevant menu choices Printable branch menu What does this price include? Amount, unit and conditions together Detailed rate card or formal schedule Which package fits my needs? Comparison of relevant inclusions and limits Full brochure or specification Is this still current? Effective period and current download link Dated document for reference What should I do next? Booking, enquiry or ordering route Copy to print or share Compare the visual page with the extracted text Start from the operator’s public page and follow the exact download link. Save the HTTP status, final URL and content type. A successful response is the beginning of the check. Open the file and locate the chosen fact. Check the heading or column that gives it context, then select or extract the text and see whether those relationships survive. Test a representative item on each relevant layout rather than assuming every page uses the same export settings. Google lists PDF as a format it can index. This describes a supported format, while the actual access and clarity of a particular file still need checking. W3C’s technique for scanned PDFs explains how OCR can add actual text to an image-based scan and why the result should be checked. A scan with no text layer, a document with unreliable reading order and a properly structured document need different remedies. A successful text extraction is also narrower than a full accessibility review. Google indexable file types (https://developers.google.com/search/docs/crawling-indexing/indexable-file-types) · W3C scanned PDF technique (https://www.w3.org/WAI/WCAG22/Techniques/pdf/PDF7) For the worked example, the public Wagamama Dubai Mall branch page linked to an English menu PDF. The download returned HTTP 200 with an application/pdf content type on 6 November 2026. Both pages were rendered and inspected, and their text was extracted locally. The file had extractable text on both pages. The first page visually separated dessert and drink columns. The plain extraction placed a dessert line and nearby drink information together in one line, and also brought the dessert and juice headings together. That is an observed reading-order issue for the extraction used here. It gives no reason to infer the price of an item from neighbouring text. Wagamama Dubai Mall branch page (https://www.wagamama.ae/restaurants/dubai/dubai-the-dubai-mall) · Published English menu PDF (https://www.datocms-assets.com/144406/1731582013-1728979889-english-menu.pdf) To repeat the check with the actual file your business links, you can use these commands with curl and the Poppler utilities installed. Compare both text outputs with the visible page. If a page contains little or no extractable text, inspect the document’s authoring or scanning process before interpreting the result. url='https://example.com/replace-with-the-current-menu.pdf' curl --silent --show-error --location --max-redirs 8 --max-time 20 \ --dump-header menu.headers --output menu.pdf \ --write-out 'status=%{http_code} final=%{url_effective}\n' "$url" pdfinfo menu.pdf pdftotext -f 1 -l 1 menu.pdf menu-reading-order.txt pdftotext -f 1 -l 1 -layout menu.pdf menu-layout.txt The worked comparison used pdfplumber for the extraction and a rendered image for visual review. Different extractors can order a page differently. Check meaningful relationships rather than comparing only character counts. For a menu, see whether the dish name, portion, price and qualifying note stay together. For a brochure, check whether a heading follows the relevant package and whether a footnote is associated with the right row. If you manage the origin and need to investigate a failing file request, correlate the exact file URL and time in permitted logs. The public sample did not include server logs. Test the web page as a reader would use it An HTML menu makes it possible to publish structured headings and readable text on a normal page, but the implementation still matters. GAIA Dubai’s menu page provides dish names and menu sections in HTML. A public GET and a clean Chrome session at 390 pixels wide were used to check one item. The response source contained the item name. Neither of two matching elements was initially visible in the mobile-width test. Selecting the visible à la carte control made one match visible. That separate check matters because a browser can contain a dish name while the selected tab shows a different menu. GAIA Dubai menu (https://gaia-restaurants.com/dubai-restaurant/gaia-dubai-menus) For your page, choose the intended starting state and follow the ordinary reader route. Confirm the important fact is visible, or that the menu choice clearly indicates where it is. Check the page after switching service periods. Keep the branch and menu name visible so someone can tell which prices or dishes they are reading. For HTML intended to convey a decisive condition immediately, put a short summary before the control and test that it stays accurate when the detailed list changes. A social post can introduce a seasonal menu. Emirates Woman embeds an official GAIA Dubai post in its coverage of a summer menu. The post and article provide context for a dated campaign, while the restaurant’s current menu page gives readers a place to check the available sections. Use public social links to point readers towards the maintained page where practical, and keep effective dates or seasonal context clear when preserving older posts. Emirates Woman coverage with the official GAIA post (https://emirateswoman.com/gaia-dubais-mediterranean-summer-menu-is-a-sun-drenched-culinary-escape/) · GAIA Dubai menu (https://gaia-restaurants.com/dubai-restaurant/gaia-dubai-menus) Check the difference between a useful summary and a fragment. A catering page that says “packages from AED 100” leaves the reader without the unit, minimum order and scope. If those details apply to the starting price, show them together on the page. If the business needs to quote each event, describe the inputs that affect the quote and link to the enquiry route. The downloadable schedule can still give the buyer the complete comparison. The HTML should help the reader understand what they can decide before contacting the business. View the selected official Instagram post (https://www.instagram.com/p/DEuRmlGo6kH/) Official GAIA Dubai Instagram post embedded in seasonal menu coverage. Keep the HTML and download aligned Give each menu or rate card an owner and a publication rule. For example, a business can manage item names, units, prices and conditions in one record that feeds the web page and the download. Where the formats are maintained manually, record which one carries the authoritative information and require the two to be reviewed together before release. Keep the effective date and branch visible on each. If one format intentionally covers a smaller set of information, explain its scope. Use appropriate web elements for the content. Headings can identify sections, lists can group related items, and a table can compare prices or inclusions when the relationships are tabular. Keep units and terms close to the values they qualify. Where a price changes by location or timing, attach the relevant condition to that price. Check the result on a narrow screen and through the download link from the same page. W3C tables tutorial (https://www.w3.org/WAI/tutorials/tables/) Before replacing or moving a downloadable file, check which public pages still link to it. Update links to the intended current version, and decide with the owner whether the previous issue should remain available as a clearly dated archive. Google documents that a canonical for a non-HTML file can be supplied through an HTTP Link header. Use that mechanism only when the relationship and target make sense, such as genuinely duplicate versions of the same resource. A brief HTML summary and a full brochure can serve distinct purposes. Google canonicalization guidance (https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls) Record the decision and retest after publishing Use a Document format record with the file address, page address, owner, effective date, sampled facts, extraction result, browser result and decision about what belongs in HTML. When there is a discrepancy, record the exact field that differs. For example, a rate can appear in both places but have a condition in only one. A follow-up test can confirm the amount and condition together and check that the download link points to the reviewed edition. When prices change frequently, pick one routine update and walk it through both formats. Record who changes the source, who approves the new figures and which public links are checked. Check at least one changed entry and one unchanged entry in the rendered page and downloaded file. If the formats publish at different times, agree how the website indicates which edition applies in the gap. This keeps a clear route for the reader who arrives through a previously shared link. Test the new or updated page as a public reader. Confirm the intended fact is readable, the relevant file opens and the business can update both when its offer changes. If you need search-platform evidence, use authorised inspection for the exact page or file where available. Treat a text extraction or browser result as evidence about that test. The availability of a file in a particular answer or index is a separate outcome. AI Visibility is published by Lunasol, which offers GEO and AEO services. The examples document format and delivery observations from public restaurant material. About AI Visibility (https://aivisibility.ae/about/)