Psst… We received a tip…

You spent three weeks arguing over your website’s headline.

You changed “Get Started” to “Find My Fit”, cut the text in half, picked the button color, rewrote the product blurb, and moved the testimonial higher because someone in the team said it “felt warmer.”

Then a customer searches on Google.

But before they reach the page you built, Google shows them something else first that you didn’t write at all. Google thinks it's doing you a favor! It thinks it's cleaner, faster, and shaped around what the searcher wants. 

This week’s patent may cause website designers some panic. It describes a system where Google can score a company’s landing page, then surface an AI-generated page instead of yours, meaning the customer may see Google’s pitch before they ever see yours.

This week’s filing comes from Google and it points to a future where Google may generate a new landing page for sites, before the viewer can get to the one the owner actually created.

Start with a normal search.

You type something into Google. Maybe “best laptop for video editing,” “family SUV deal,” or “skin clinic near me.”

Usual search results show up.

Google then builds the usual search results page. One of those results is an ad which points to a company’s landing page, the page that company built, designed, tested, and probably argued about in seven different marketing meetings.

Then Google scores that page.

The patent calls this a “landing page score.” That score can look at whether the page is useful for the user, whether it has the right content, whether it performs well, whether people usually bounce, and whether it has helpful things like filters, product options, or clear next steps.

Here is where it gets spicy…

If Google decides the company’s landing page could be better for that search, it can update the search results page with a new link. But that link sends the user, not to the company’s landing page, but to an AI-generated page for that company.

That AI page is built around the searcher’s actual intent. Google can use the current query, past search context, the user account, product information, and other signals to shape the page around what this person seems to want.

So instead of one static landing page for everyone, the user might see a custom page with product feeds, filters, offers, a chatbot, call-to-action buttons, and links into the company’s real product pages.

To be clear, Google doesn’t delete the original company website or edit the company’s actual page. The site still exists. But the first pitch (aka the first landing page that shows up) may no longer belong to the company.

The company’s landing page used to be where the sale started. In this patent, Google may decide the sale starts one step earlier, on a page Google generates itself.

Google can show you the right ad or result. But once you click, the whole experience often falls into someone else’s website, and that website might be slow, messy, too broad, or missing the exact thing you came for.

Google already treats the page after the click as part of ad quality. In Google Ads, Quality Score looks at expected clickthrough rate, ad relevance, and landing page experience. So Google cares about both sides of the moment, being whether the ad earns the click, and whether the page after that click actually matches what the searcher wanted. (Google Help)

The customer (or the searcher) has a similar problem. They search with a job in mind. But instead of landing on the answer, they hit a generic page and have to scroll, filter, guess, and translate marketing copy into a decision.

And users are brutal when that happens. Google has previously said 53% of mobile visitors leave a page that takes longer than three seconds to load, and bounce probability rises 32% when load time goes from one second to three seconds. Google’s patent is a way to stop losing them after the initial click.

Nobody is publicly doing the exact scary version yet.

The closest public products still mostly work inside the company’s own website. Webflow’s Optimize can use AI to test page changes and personalize the on-page experience. Mutiny connects to a company’s data so the website can change logos, case studies, messaging, and branding for different accounts. Adobe Target does the enterprise version, using AI, testing, and Adobe Analytics data to decide which experience a visitor should see.

Google is already walking in this direction with AI Max for Search campaigns. Its public ad tools can generate customized ad text from landing pages, ads, and keywords, and Final URL expansion can send users to the most relevant page on the advertiser’s site based on search intent. Google is already trying to match the searcher to a better post-click path. (Google Help)

There is also a builder-side version of this. Claude Code, Cursor, Webflow, HubSpot, Unbounce, and Instapage can help teams write, test, and build better pages faster.

Google’s patent asks a more uncomfortable question. What if Google decides it can build the better page itself?

In the U.S. alone, internet advertising hit $294.6 billion in 2025, and search was still the biggest slice at 38.8% of the market. For Google, better post-click pages can mean more clicks, more conversions, and more spend.

Google’s own numbers make the stakes higher. In 2025, Google Search brought in $224.5 billion, up from $198.1 billion the year before. Paid clicks rose 6%, while cost-per-click rose 7%. (SEC)

And customers are expecting more personalization. McKinsey found that 71% of consumers expect personalized interactions, and 76% get frustrated when they do not get them. It also found personalization often drives a 10% to 15% revenue lift. (McKinsey & Company)

The AI market is pushing the same way. Adobe’s 2026 report says 80% of organizations think breakthrough customer experience will be highly personalized and able to anticipate needs in real time. It also says 58% expect agentic AI to handle product or brand discovery and search within the next 18 months. The catch? Only 18% have broadly deployed agentic AI for personalization and recommendation. (Adobe Business)

Is this the market Google is going to try and capture? Do you think they will released this feature?

Mutiny, a startup that plugs into a company’s data and website to personalize copy for different visitors, raised US$50 million at a reported US$600 million valuation in 2022. Mutiny works inside the brand’s own site, but it shows the thesis investors bought early that one static page for every customer is starting to look old. (TechCrunch)

Webflow, a website-building platform used by marketing teams to design and manage sites, acquired Intellimize, an AI website personalization and conversion-rate optimization company, in 2024. Braze, a customer engagement platform for messaging across channels like email, push, and SMS, agreed to buy OfferFit, an AI decisioning company, for US$325 million in 2025. OfferFit’s whole pitch is that AI agents can choose the best marketing action for each person, instead of relying on old-school A/B tests.

The data layer is consolidating too. Contentsquare, an experience analytics company that shows how users behave on websites and apps, acquired Heap, a product analytics company that captures user behavior data, in its largest acquisition to date in 2023. If you can see what users do, unify that data, and change the experience in real time, you own a more valuable piece of the funnel. (Contentsquare)

Private equity and strategic buyers are chasing the same thing from another angle of turning website traffic into pipeline. Drift, a conversational marketing company that uses chat and buyer signals to convert website visitors, took a strategic investment from Vista Equity Partners in 2021. Then Salesloft, a sales engagement platform, acquired Drift in 2024 to connect web engagement with revenue teams. Salesforce, the CRM giant, signed a deal in 2025 to acquire Qualified, an agentic AI marketing company that converts website visitors into a sales pipeline. (PR Newswire)

Who is responsible for the words the customer sees?

A company can spend weeks making sure its landing page says exactly the right thing. Not too much. Not too little. No promises the product cannot keep. No claims the legal team has not approved, especially in highly regulated markets. But if Google generates a new page for that company, the customer may see a version of the pitch the company never wrote, never checked, and never knew was live.

That creates a nasty misrepresentation problem.

Imagine a clinic carefully avoids saying “guaranteed results.” A travel company avoids promising a visa will be approved. A software company avoids claiming its tool is compliant in every country. Then an AI-generated page summarizes the offer a little too confidently, trims the caveats, or turns careful wording into a cleaner sales pitch. Helpful? Maybe. Dangerous? Absolutely.

AI can hallucinate and it can sound more certain than the original business ever intended to be.

If Google owns the first pitch, brands may need a new kind of defensive SEO (Search Engine Optimisation) to ensure AI doesn’t misrepresent their claims.

Landing pages are a crucial part of the sales pipeline. If Google Search can score your page, generate a better version, and show that to the customer first, the first impression may move from your website to Google’s AI layer.

So read the filing for yourself: US 12,536,233 B1, filed by Google, titled “AI-Generated Content Page Tailored to a Specific User.” 

Reply with your take. We read everything. Find us on Instagram and LinkedIn for more on the deals, patents, and market signals we're watching.