Your quiz results and shared carts probably shouldn't be in Google. Google's new spam guidance explains why

Most CPG brand sites aren't platforms. They don't host other people's blogs or let strangers publish pages. So it's easy to read Google's latest spam guidance and decide it's someone else's problem.

But look at what a modern DTC food or beauty site generates on its own. A skincare quiz that produces a results page for every combination of answers. A "build your box" tool that saves each bundle at its own URL. Shared carts and wishlists. A gift-note preview. A custom-label page for a personalized bottle. Every one of those is a page made for one person and a link, not for search. And Google has just said, in writing, what it wants done with them.

What Google added on October 26

Google's guide to preventing and monitoring abuse now has a section called "Only index content meant for Search." It's two sentences long:

If your service lets users create shareable pages, generated outputs, or interactive tools, consider whether that content is meant to be found through search engines or only accessed directly by people with a link.

When pages are created for direct sharing, collaboration, or utility rather than public search discovery, apply the noindex rule (using a robots meta tag or X-Robots-Tag HTTP response header) or block crawling in robots.txt by default.

Search Engine Journal's Matt G. Southern reported that Google's changelog describes the change as being about "blocking indexing of link-shared content." The guide's opening also now names "tools that let users publish shared pages, generated content, or web apps" among the services open to abuse. The previous version was last updated December 10, 2025. Search Engine Roundtable covered the same change.

The context is AI chat share links. SEJ notes that Southern had confirmed in July that Claude's share path was blocked in robots.txt and returned an X-Robots-Tag of "none," and he couldn't determine what Google saw when it first found those links. That's a big platform. The same mechanics apply to a brand's quiz results.

The catch: robots.txt doesn't do what most people think

Google's guidance offers two options, noindex or a robots.txt block. They are not equivalent, and Google's other documentation says so plainly.

SEJ points out that the robots.txt introduction the new section links to says robots.txt "is not a mechanism for keeping a web page out of Google," and that a page "that's disallowed in robots.txt can still be indexed if linked to from other sites." Google won't crawl the page, but the bare URL can still show in results.

Meanwhile, Google's noindex documentation says that for noindex to work, "the page or resource must not be blocked by a robots.txt file." If it's blocked, "the crawler will never see the noindex rule, and the page can still appear in search results, for example if other pages link to it."

So the most common setup we'd expect on a brand site, blocking /quiz/results/ in robots.txt and also adding noindex to be safe, is the one combination that undercuts itself. The block stops Google from ever reading the noindex.

Method	Stops crawling	Keeps the URL out of results	Notes from Google's docs
robots.txt Disallow	Yes	Not reliably	A disallowed URL "can still be indexed if linked to from other sites"
noindex robots meta tag	No	Yes, once crawled	Works only if the page isn't blocked by robots.txt
X-Robots-Tag: noindex header	No	Yes, once crawled	Same effect as the meta tag; useful for non-HTML files and app-generated pages
Both robots.txt and noindex	Yes	Not reliably	The crawler never sees the noindex

Which CPG pages this covers

Here's how we'd sort the pages a typical brand site generates. The recommendations are ours, applying Google's guidance; Google didn't name these page types.

Page type	Made for search or for a link?	Our default
Quiz landing page ("Find your routine")	Search	Index
Individual quiz results URL	A link, one person's answers	noindex
Build-a-box or bundle builder results	A link	noindex
Pre-built bundles you sell as products	Search	Index, as normal product pages
Shared cart or wishlist	A link	noindex
Gift message or custom label preview	A link	noindex
Recipe generated from a pantry tool	A link, unless edited and published	noindex by default
Written guide summarizing quiz outcomes	Search	Index

The pattern: if a person made it, for themselves or to send to someone, it's a link page. If your team made it, for everyone, it's a search page. Where the two overlap, as with a quiz, index the part your team wrote and keep the generated parts out.

Why bother if nobody's spamming you?

Three reasons, none of them spam penalties.

First, these pages are thin. A results page that says "Your match: Hydrating Serum" and nothing else is a low-quality page with your brand on it, possibly thousands of times over.

Second, they can say things you didn't approve. A pantry recipe tool might combine your product with an ingredient your allergen page warns against. A custom-label preview shows whatever a customer typed. You don't want either surfacing in a search result, or being read by an AI crawler as something your brand published.

Third, they dilute what crawlers spend time on. We've written about how assistants read brand sites as reference material. A crawler working through every quiz permutation is spending less time on the product pages that matter. The crawler-visits tip shows how to see where bots actually go.

How to fix it, in order

1. List the generators. Ask whoever runs the site which tools create a new URL per user: quizzes, bundle builders, carts, wishlists, previews, referral pages. Don't forget third-party apps, which often add their own URL paths.
2. Check what's already indexed. Search Google with the site: operator plus the path, for example your domain followed by /quiz/results. Then check Search Console's Page Indexing report for those paths.
3. Add noindex. Use a robots meta tag in the template, or an X-Robots-Tag header if the pages come from an app you can't template. Google says the two have the same effect.
4. Remove any robots.txt block on the same paths. Otherwise Google never sees step 3. Once the pages have dropped out, you can decide whether to block crawling later, accepting that blocked URLs can resurface if linked.
5. Keep a search-friendly version. If you want the quiz to rank, make sure the quiz landing page and a written guide to the outcomes are indexable, with real content. That's the page to optimize.
6. Watch the reports. Google's updated guide tells site owners to verify in Search Console, check Page Indexing and Search performance "for unexpected spikes in indexed pages, impressions, or queries that don't match your platform's purpose," and alert on spikes in new public pages. SEJ adds that the Page Indexing report has a status, "Indexed, though blocked by robots.txt," and Google's help for it says to remove the block and use noindex.

Our tip on quiz and cart pages is the short version of this list.

Does this affect AI assistants too?

Partly. noindex is a search engine rule, and Google's covers Google Search. Other crawlers, including the ones that feed AI assistants, follow their own policies, and some read robots.txt rules aimed at them specifically. So the fix keeps generated pages out of Google's results, which is where Google's own AI features in Search look for pages to link. For other assistants, the best protection is the same as ever: make the pages you do want read clear and correct, so a stray quiz page isn't the most specific thing a crawler finds.

If you want to see what assistants are actually saying about your products, rather than guessing which pages they read, tools such as XENA Intelligence ask them shopper-style questions and score the answers. Its SearchPanel view lets you read the answers themselves, product by product.

The short version

Google's new section is two sentences, aimed at platforms, and easy to skip. For a brand site, it boils down to one test: was this page made for search, or for a link? If it's for a link, noindex it, and make sure robots.txt isn't standing in the way.