What ChatGPT's GPT-6 Intelligent UI means for marketing teams

OpenAI began rolling out GPT-6 with Intelligent UI on 7 October 2026 to the more than 1.2 billion people who use ChatGPT each week. The model can now compose an answer as an interactive interface, with side-by-side comparisons, maps, forms and small tools built on request, so the page where buyers meet your brand has changed shape.

GPT-6 reaches every ChatGPT tier

OpenAI’s announcement says GPT-6 was trained to compose responses from text, visuals and interactive elements and to choose the mix for each question. A comparison can appear side by side, an explanation can become an interactive diagram, and the stops on a road trip can sit on a map with notes on which ones are worth a detour. Users can also ask for a tool, and OpenAI’s examples are a calculator for how savings could grow, a bill splitter and a game. OpenAI built a library of native components and a compiler that renders the interface as the model writes it, so the layout appears progressively.

Speed changed too. ChatGPT can now begin answering while it is still thinking, and for questions that need web search GPT-6 Instant starts answering 44% sooner on average than GPT-5.6 Instant. OpenAI says GPT-6 also makes better decisions about when to look something up.

According to the ChatGPT release notes, Plus, Pro, Business and Enterprise users get GPT-6 Sol in the Chat tab from 7 October, and Free and Go users get GPT-6 Luna from 8 October. The Pro reasoning option stays on GPT-6 Astra without Intelligent UI, and the models behind Work and Codex are unchanged.

Your website calculator now has a rival

Many brands built traffic and lead capture on simple interactive tools, such as mortgage and savings calculators, size and fit guides, ROI estimators and product comparison tables. A buyer who can type “build me a calculator for how much I’d save over five years” and get a working one in the chat has less reason to click through to yours. OpenAI chose a savings calculator as a launch example, which should get the attention of every finance brand that ranks for “savings calculator”.

Brands that sell against named rivals face a different risk from the comparison layout. When ChatGPT puts three products in columns, it fills each cell with whatever it found about price, specs and delivery. A brand whose facts are scattered across PDFs, images and inconsistent product pages risks a blank or wrong cell next to a competitor whose numbers were easy to read. OpenAI hasn’t said what data the comparisons draw on, so that is our inference from how ChatGPT search already works.

Our view at Manual Focus is that a generic calculator or comparison table is no longer a defensible traffic asset, and brands should rebuild those tools around what ChatGPT can’t make up, meaning live prices, stock, a customer’s own account data or a quote you will honour. That value can live on your site or inside a ChatGPT plugin. Radisson’s ChatGPT plugin, which shows hotels on a map with prices and amenities, converted visits to bookings at about 1.5 times Radisson’s organic search rate in July and August 2026, according to OpenAI’s case study. OpenAI picked that result to publish, so treat it as a signal, and our post on dots and ChatGPT plugins covers how to scope one.

Content teams have a smaller adjustment to make. An answer that starts while ChatGPT is still searching rewards pages that state the key fact early in plain language, which is the same discipline our SEO cluster generator playbook builds into briefs.

Five things to do before the end of October

1. Re-run your category prompts on both tiers. Once GPT-6 reaches your account, test your 20 most important buyer questions on a Free account (Luna) and a Plus account (Sol). Screenshot the format of each answer and note whether you appear in any comparison.
2. Audit your interactive tools. List every calculator, quiz, guide and comparison table on your site with its traffic and leads, then ask ChatGPT to build the same thing. Where its version is as good as yours, plan what you will add that only you can supply.
3. Put comparison facts in plain text. Make price, key specs, dimensions and delivery terms readable on each product page in the same format, rather than in images or downloads.
4. Set a new referral baseline. Mark 8 October in your analytics now, then after 5 November compare ChatGPT referral sessions, and conversions on tool pages, for the four weeks either side.
5. Test your plugin, if you have one. Check whether ChatGPT still surfaces it for the prompts it should win now that answers can carry their own maps and forms.

Gaps in what OpenAI published

OpenAI hasn’t said whether Intelligent UI changes how ChatGPT shows sources or links out to websites, whether ads can appear inside interactive answers, or how a plugin and a model-built interface compete for the same question. It gave no traffic or engagement figures, and it says there is “still work ahead” on the model’s design judgment, so the formats will keep changing. The 44% figure is OpenAI’s own measurement and its quality claims come from internal evaluations, and no outside group has checked either.

If you want help deciding which of your site tools still earn their place, talk to us.

I run marketing for [brand], [one line on what you sell] at [website], selling
mainly in [market].

ChatGPT can now build small tools inside a conversation on request, such as a
calculator for how savings could grow. I want to know which of our website
tools a buyer could now get in the chat instead.

Attached is an analytics export listing every calculator, quiz, guide and
comparison table on our site, with sessions and leads for each over the last
[period].

For each tool:
1. Build your own version of it from a one-line request, as a buyer would
   ask for it (for example "build me a calculator for how much I'd save over
   five years").
2. Compare your version with ours, using the URL in the export.
3. Say whether ours depends on anything you can't make up: live prices,
   stock, a customer's own account data, or a quote we will honour.

Give me a table with the columns: Tool | URL | Sessions | Leads | Could you
match it? (yes/partly/no) | What ours has that yours doesn't | What ours relies
on that only we can supply (or "nothing").

Sort by leads, highest first. Don't recommend what we should add; that's our
call. If you couldn't open a tool or build a fair version of it, say so in
that row rather than guessing.

We sell [product category] at [website] and compete with [named rivals].

When an AI assistant compares products in columns, it fills each cell with
whatever it can find about price, specs and delivery. I want our facts to be
easy to read on every product page.

Here are our product page URLs: [paste list].

For each page, check whether these are stated in plain, crawlable text (not
only in an image, PDF or download):
- price
- key specs: [list the specs that matter in our category]
- dimensions
- delivery terms

Give me:
1. A table with the columns: URL | Price in text? | Key specs in text? |
   Dimensions in text? | Delivery terms in text? | Where the missing facts
   currently live (image, PDF, download, not found) | Format matches other
   pages? (yes/no).
2. For each product, a draft fact block in one consistent format, using only
   facts you found on our pages. Leave a [blank] wherever a fact is missing
   rather than filling it in.

List any page you couldn't load or read in full.

Run this once, after 5 November 2026.

GPT-6 reached free ChatGPT users from 8 October 2026. I want to see whether
ChatGPT referrals to our site, and conversions on our tool pages, changed
after that date.

Attached is an analytics export covering 10 September to 5 November 2026, with
daily sessions and conversions by referral source and landing page. Our tool
pages are: [paste URLs of calculators, comparison tables and other tools].

Compare the four weeks before 8 October with the four weeks after it.

Give me:
1. A table with the columns: Measure | Four weeks before | Four weeks after |
   Change | Change (%), with rows for ChatGPT referral sessions, ChatGPT
   referral conversions, and sessions and conversions for each tool page.
2. Three to five sentences on what moved most.

Only report what the data shows; don't explain why something changed. If
ChatGPT referrals can't be separated from other sources in the export, or a
date range is incomplete, say so before giving any numbers.