TL;DR
- Per Google Ads Help, URL exclusions in AI Max for Search need both Text customization and Final URL expansion enabled, and can be set as exact URLs, page feed custom labels or URL contains rules.
- Google says hyphens, periods and slashes split a URL into tokens while underscores do not, so promo_page counts as one token and a rule for promo alone would not match it.
- Google's example: an exclusion for FAQ blocks /FAQ/page but not /faqs or /FAQ11. Spaces in a rule act as AND across the whole URL, and Google recommends not using them.
- After a Dynamic Search Ads campaign is upgraded to AI Max, page title and page content rules become read-only. Google's Ads Developer Blog (11 June 2026) sets DSA automigration for February 2027.
- The help page is undated (a trade-press report on it is dated 15 November 2026) and does not cover case sensitivity, query strings, rule limits or non-Latin URLs such as Thai slugs.
Google has documented how URL exclusions work in AI Max for Search campaigns. A Google Ads Help page titled "About URL exclusions in AI Max for Search campaigns", reported on 15 November 2026, says a "URL contains" exclusion matches whole tokens of a URL and not fragments of text. A rule for "auto" does not exclude a URL that contains "automation". Underscores do not split a URL into tokens, and a rule for "FAQ" does not block "/faqs" or "/FAQ11". The page carries no publication date, and it does not say whether this matching behaviour is new or has only now been written down.
What the Google Ads Help page says about URL exclusions in AI Max
According to the Google Ads Help page, read on 16 November 2026, URL exclusions let an advertiser stop specific web pages from serving when Final URL expansion is used in an AI Max for Search campaign. The page sets one precondition: both Text customization and Final URL expansion must be enabled before URL exclusions can be used.
Google lists three ways to exclude pages:
- provide exact URLs
- use custom labels from page feeds
- create custom "URL contains" rules
The article sits in the help section "Optimize Google Ads with AI Max for Search campaigns". It is short and covers custom labels, how "URL contains" reads the text of a rule, and what happens to old rules after an upgrade from Dynamic Search Ads (DSA). Most of the detail is in the matching section.
How a URL contains rule matches text, according to Google
Google Ads Help describes four behaviours of a URL contains rule. Each one can make an exclusion do less, or more, than the person who wrote it expected.
Rules match whole tokens, not substrings
Google's page says "URL contains" relies on consecutive tokens and does not do partial substring matching. Its example: an exclusion rule for "auto" will not exclude a URL that contains "automation". To block that page, the rule has to say "automation". A short stem typed in the hope of catching every longer word that starts with it will not work.
Hyphens, periods and slashes split a URL, underscores do not
The page says the system splits URLs into tokens at common separators such as hyphens, periods and slashes. Underscores are the stated exception. Google's example is the string promo_page, which is treated as one unbroken token.
A space in the rule works as AND
Google says spaces in the rule text split it into several phrases. They then act as an AND operator that matches tokens anywhere in the URL, in any order. The page adds that it is "highly recommended" not to use spaces in the rule text, to avoid confusion.
No plurals and no appended numbers
Because matching is on exact tokens, the page notes that a rule does not match plurals or words with numbers attached. Google's example: an exclusion for "FAQ" blocks "/FAQ/page" but does not block "/faqs" or "/FAQ11".
Worked examples: which URLs a rule would and would not block
The table below puts Google's own examples from the help page next to three examples derived strictly from the rules the page states. The derived rows are marked as such. They are readings of the documented rules and were not tested in a live account, so confirm them before relying on them.
| Rule text | URL | Result and basis |
|---|---|---|
| auto | a URL containing "automation" | Not excluded. Google's example. |
| FAQ | /FAQ/page | Blocked. Google's example. |
| FAQ | /faqs | Not blocked, plurals are not matched. Google's example. |
| FAQ | /FAQ11 | Not blocked, appended numbers are not matched. Google's example. |
| promo | /promo_page/ | Not blocked. Derived: Google says promo_page is a single token, and "promo" is not that token. |
| careers | /careers.html | Blocked. Derived: a period is a separator, so "careers" is its own token. |
| summer sale (with a space) | /sale/2026/summer/ | Blocked. Derived: a space acts as AND and tokens can sit anywhere, in any order. |
The same reasoning gives a few more cases. A rule for "blog" should block /blog/seo-tips/, because the slashes and the hyphen leave "blog" as a token of its own. By Google's plural and number examples it would not block /blogs/ or /blog2026/. By the underscore rule it would not block /blog_archive/ either, since blog_archive is one token.
The space behaviour cuts the other way. Someone who types "summer sale" may expect it to block only pages about a summer sale. Under the documented rule it applies to any URL that has both tokens somewhere, including a URL where "sale" comes first and "summer" sits three folders later. That is a wider net than the phrase suggests, which is probably why Google recommends leaving spaces out.
Google's wording about "consecutive tokens" implies that a rule made of several tokens, such as summer-sale, has to find those tokens next to each other in the URL. The page gives no example of a multi-token rule, so that reading is also one to test.
Where URL exclusions are set in Google Ads
The new page explains matching but not the clicks needed to add an exclusion. Those are on a separate Google Ads Help page, "About Final URL expansion in Search". That page says Final URL expansion is a campaign-level setting, available through AI Max for Search campaigns, and that it is enabled by default when an advertiser opts in to AI Max. It also says Text customization must be on for Final URL expansion to work, and that turning off Text customization turns off Final URL expansion as well.
Per the Final URL expansion help page, exclusions are added in the campaign's settings. Open the AI Max section, check the AI Max toggle is on, and under "Final URL expansion" select "Add URL exclusions". The panel has a URLs tab for listing exact URLs and a Rules tab for custom rules. URL inclusions are handled in a different place: the same page describes adding them from the Ad groups view.
Two more points on that page affect how exclusions behave:
- When Final URL expansion is on, pinned responsive search ad assets are not used if a more relevant URL is chosen. The same applies when URL inclusions are enabled. Google says that to always use pinned assets, both have to be off.
- In Search Campaigns for Travel, the feed stays the source of the landing page for travel promotion ads and booking links. If a feed item points to an excluded URL, the feed item keeps priority, and the exclusion only stops the AI from using that page to find new search terms.
Page feeds and custom labels as exclusions
For advertisers who upload a curated list of URLs through page feeds, the Google Ads Help page on URL exclusions says specific pages from that feed can be kept from serving by adding their custom labels as URL exclusions. The Final URL expansion help page says the same thing from the other direction: to target specific feed URLs with Final URL expansion enabled, use custom labels for campaign-level URL exclusion or for ad-group-level inclusions.
A label is attached to the URLs the advertiser chose in the feed, so it does not depend on how a URL happens to be spelled. Where a URL contains rule cannot express what is needed, for instance because the site mixes underscores and hyphens, a labelled feed or a list of exact URLs avoids the tokenisation question.
What happens to Dynamic Search Ads rules after an upgrade to AI Max
The Google Ads Help page has a short section for advertisers who upgraded an existing DSA campaign to AI Max. It says legacy rule types for URL inclusion and exclusion, such as "page title contains" or "page content contains", will become read-only. Advertisers can remove legacy rules. They cannot add new ones or edit existing ones after the upgrade.
The timing comes from a different Google source. A Google Ads Developer Blog post dated 11 June 2026 says the ability to create DSAs is removed in January 2027 and that automigration begins in February 2027, when remaining active DSA campaigns are upgraded automatically. The same post recommends managing the migration by hand before then.
Putting the two documents together: a rule based on page title or page content can be edited today in a DSA campaign and cannot be edited once that campaign has been upgraded. Reviewing those rules before the upgrade is the only point at which they can still be changed. The help page does not say whether automatically migrated campaigns are treated exactly like manually upgraded ones. It refers only to campaigns that were "upgraded".
What to check in your own Google Ads account
The work here is a careful read of rules that already exist, and it needs no new tool.
- List the Search campaigns that have AI Max on, and note whether Final URL expansion is on. It is on by default when AI Max is enabled, per Google's Final URL expansion help page.
- Open "Add URL exclusions" in each campaign and read the Rules tab line by line. Look for short stems such as "auto" or "promo" that were meant to catch longer words.
- Compare each rule with the real URLs on the site, from a sitemap or a crawl export. Check plural folders (/blogs/), numbered variants (/promo2026/) and slugs that use underscores.
- Remove spaces from rule text unless the AND behaviour is what was intended.
- Where a rule cannot be written cleanly, exclude exact URLs or use page feed custom labels.
- For DSA campaigns still running, review "page title contains" and "page content contains" rules before any upgrade, because they become read-only afterwards.
- After changing exclusions, watch where paid traffic lands. The help page describes no report that shows which exclusion blocked which URL, so landing page data in Google Ads or in a properly configured GA4 property is the practical check.
Teams that want a second pair of eyes on this can ask for it as part of Google Ads management.
What the page did not say
The Google Ads Help page leaves several questions open, and it is better to list them than to guess.
- When it was published. The page shows no date. A trade-press report on it is dated 15 November 2026.
- Whether the matching behaviour changed recently or was only documented now.
- Whether matching is case-sensitive. Google's example uses "FAQ" in the rule and in the blocked URL, and never states a rule about case.
- How query strings and parameters are handled. Hyphens, periods and slashes are given as examples of separators ("like"), so the list may not be complete. Question marks, equals signs and ampersands are not mentioned.
- Whether the domain or subdomain counts as part of the URL for matching.
- Whether there is a limit on the number of rules.
- Whether any report shows which exclusion blocked a page.
- Anything about the Google Ads API or Google Ads Editor.
- How URLs in non-Latin scripts, or their percent-encoded forms, are tokenised.
- Any country or language differences. The page names no markets.
How this connects to earlier AI Max coverage
Relevant Audience reported on 13 November 2026 that Google Ads Editor 2.14 turns on AI Max and Final URL expansion by default for new Search campaigns, and on 12 November 2026 that Google Ads had documented the {search_term_matchtype} parameter for separating AI Max expansion clicks. The help page on URL exclusions in AI Max is a third piece of the same picture. The earlier two were about where AI Max gets switched on and how its clicks are labelled. This one is about which pages it is allowed to send those clicks to.
If Final URL expansion is on by default, the exclusion list is the documented way to keep it away from pages that should not receive paid traffic while leaving expansion running. The other option Google documents is turning Final URL expansion off.
What this means for Thai marketers
The Google Ads Help page does not mention Thailand or any other market, so what follows is reasoning from the documented rules and not something Google said.
Take a site that separates languages with folders such as /th/ and /en/. Under the token rule, a rule for "th" should match the /th/ folder, because the slashes make "th" a token. For the same reason it should not match a slug like /thailand-office/, where the token is "thailand". Advertisers running separate campaigns per language should test this on their own URLs before trusting it to keep a campaign on one language.
Thai-script slugs are the open question. A Thai slug is percent-encoded when it travels as a URL, and the page says nothing about non-Latin characters, percent signs or encoded text. It is not safe to assume a Thai word typed into a URL contains rule will match, or that it will fail. For pages with Thai slugs, exact URLs or page feed custom labels are the options whose behaviour does not depend on tokenisation.
URL structure is normally decided by whoever looks after the site's SEO in Thailand. With token matching, choices such as hyphen or underscore, singular or plural folder, and Thai or Latin slug now decide whether a paid search exclusion works. Search and paid media teams have a reason to compare notes before the next site restructure.
Frequently asked questions
Does a "URL contains" exclusion in AI Max match part of a word?
No. Google Ads Help says "URL contains" matches consecutive tokens and does not do partial substring matching. A rule for "auto" will not exclude a URL containing "automation", and a rule for "FAQ" will not block "/faqs" or "/FAQ11".
Do I need Final URL expansion turned on to use URL exclusions?
Yes. The Google Ads Help page says both Text customization and Final URL expansion must be enabled to use URL exclusions in AI Max for Search campaigns. Google's separate page on Final URL expansion adds that AI Max itself must be enabled.
Are underscores treated as separators in a URL?
No. Google says URLs are split into tokens at hyphens, periods and slashes, and that underscores are an exception. A string such as promo_page is one token, so by that rule an exclusion for "promo" alone would not match it.
Can I still edit Dynamic Search Ads rules after upgrading to AI Max?
No, not the legacy types. Google says rules such as "page title contains" or "page content contains" become read-only after a DSA campaign is upgraded to AI Max. They can be removed, but not edited, and new ones cannot be added.
When did Google publish this page, and does it apply in Thailand?
The page does not say. It has no publication date and names no countries. A trade-press report on it is dated 15 November 2026, and nothing on it limits the rules to particular markets.
The short version
A URL exclusion in AI Max only works if its text lines up with whole tokens in the URL. Google's help page now spells out the rules, and its examples turn on small details: a plural folder, a number on the end of a slug, an underscore. For help auditing AI Max settings, Final URL expansion and exclusion rules, get in touch with the Relevant Audience team.







