ChatGPT Ads for UK Businesses: The 2026 Setup and Measurement Guide

ChatGPT Ads for UK businesses are sponsored placements that appear below ChatGPT’s answers for users on the Free and Go plans. UK companies can now buy them self-serve through OpenAI’s Ads Manager, paying by impression, click or conversion-optimised bidding, and measure results with the OpenAI Pixel, the Conversions API and, since September 2026, HubSpot.

Your buyers already ask ChatGPT which supplier to choose, what a project should cost and how two options compare. ChatGPT Ads for UK businesses put your offer in front of them at that moment, and most of your competitors have not started yet. The channel is new, though, and new channels punish guesswork.

There are no keywords to bid on, no published benchmarks and a measurement setup that most marketing teams have never touched. Spend without a plan and you will learn very little; skip the tracking and you will not know whether a single lead came from it.

This guide walks through the whole process: how the auction and targeting work, how to set up Ads Manager as a UK business, how to write ads and landing pages for a conversation, how to track results through to HubSpot pipeline, and how to budget a first test you can actually learn from.

What ChatGPT Ads Are, and Who Actually Sees Them

ChatGPT Ads are paid placements that appear below a ChatGPT response when the system judges an advertiser’s offer to be relevant to the conversation. They are clearly labelled as sponsored and visually separated from the answer itself. OpenAI states plainly that ads do not influence ChatGPT’s answers and run on separate systems from the chat model.

Four facts about ChatGPT Ads: Free and Go audience, one chat card format, relevance selection and no search terms

That separation matters for how you think about the channel. You are buying a placement next to the answer, in the moment a person is working through a decision, and the answer itself stays independent of your spend.

Who Sees the Ads

According to OpenAI’s introduction to ads in ChatGPT, ads are shown only to users on the Free and Go plans. Plus, Pro, Business, Enterprise and Edu accounts do not see ads, and accounts identified as belonging to people under 18 are excluded.

This shapes the audience in ways worth thinking through before you spend anything. Free and Go users include a huge number of consumers and many small business owners who have not paid for a subscription. Senior decision-makers at larger companies who use ChatGPT through a Business or Enterprise licence will not see your ads at work.

What an Ad Looks Like

There is currently one main ad unit, which HubSpot calls the ChatGPT Chat Card. Each ad contains:

  • the advertiser name
  • a favicon, usually the advertiser logo
  • a title, which acts as the headline
  • copy, which acts as the description
  • an image
  • a landing page

There are no carousels, no video formats and no text-only search-style ads at this stage. OpenAI describes the platform as being in beta and says formats will evolve, so expect this list to grow.

How Ads Are Selected

OpenAI says its system chooses ads based on expected relevance and outcomes. It considers the context and intent of the current conversation, the ad’s landing page, title and copy, the advertiser’s context hints and, where a user has ads personalisation switched on, selected signals from their wider ChatGPT use.

Your landing page and your ad copy are targeting signals in their own right. On most platforms they only affect what happens after the click. In ChatGPT they also influence whether your ad is shown at all, which changes how much care they deserve.

What Advertisers Do Not Get

OpenAI states that conversations are kept private from advertisers and that customer data is never sold. You will not see the questions people asked, and there is no search-terms report showing the prompts that triggered your ads. Reporting covers impressions, clicks, spend, click-through rate, average CPC, average CPM and conversions.

For marketers used to Google Ads search term reports, this is the biggest adjustment. You optimise from outcomes and from the themes in your context hints.

Which UK Businesses Suit the Channel Best

Some businesses are naturally better placed to test ChatGPT Ads than others. The strongest fits share three traits: buyers research before they enquire, the decision involves comparing options, and the business has a landing page that answers the questions those buyers ask.

Considered consumer purchases are a particularly good match, because the Free and Go audience includes a large number of consumers. Home improvements, garden buildings, solar and insulation, private healthcare where policy allows, education and professional services for individuals all involve the kind of questions people bring to ChatGPT.

Small B2B services can also fit, especially where the buyer is an owner or manager of a small company who may use a free account. Enterprise B2B sellers whose buyers use ChatGPT through a company licence are the weakest fit, because those users never see ads. For them, budget is usually better spent on LinkedIn, search and content until the audience picture changes.

Where ChatGPT Ads Stand in the UK Today

OpenAI began testing ads in the United States on 9 February 2026. Over the following six months it expanded to eight further markets, then announced on 18 August 2026 that ChatGPT Ads would reach 31 European countries. An update on 31 August confirmed self-service access through Ads Manager across those markets.

Figures on ChatGPT Ads availability in Europe and GLO's own AI assistant referral traffic

UK Self-Service Is Live

OpenAI’s Ads Manager availability page lists the United Kingdom as available for self-service, alongside the United States, Ireland and most of Europe. To use self-service Ads Manager, the legal entity that advertises and pays must be based in an available country, so a UK limited company can sign up and run campaigns directly.

We have not found an OpenAI page giving a specific UK launch date, and secondary sources disagree on it. What matters for planning is that any UK business can open an account and launch a campaign today without going through a sales team or agency partner.

Still a Beta Platform

OpenAI’s own FAQ describes ChatGPT Ads as in beta and in the early stages of scaling. It warns that delivery systems, inventory, formats and buying options will keep changing, and that delivery and spend may fluctuate during the beta.

There are no published performance benchmarks yet. OpenAI says so directly: results depend on objective, creative, landing page and relevance, and it does not yet have benchmarks across advertisers, industries or campaign types. Anyone quoting a “typical” ChatGPT Ads cost per lead in the UK is guessing or extrapolating from a handful of accounts.

Scale of the Platform

OpenAI says that tens of thousands of marketers have now advertised on ChatGPT. It has also built the platform out quickly since February: conversion optimisation alongside CPM and CPC bidding, geo-targeting, custom audiences, the OpenAI Pixel, a Conversions API and third-party measurement integrations.

Why the Timing Matters for UK Firms

Early movers on any ad platform tend to face less competition in the auction while the rest of the market waits for proof. That advantage does not last, and it only pays off if you use the period to learn: which offers resonate, which context hints deliver, which landing pages convert.

There is a second reason to start now. GLO’s own analytics show AI assistants sending a growing share of referral traffic. Comparing the quarter to 30 September 2026 with the quarter before, sessions from chatgpt.com rose from 14 to 33, and claude.ai went from none to 27 sessions with six conversions. The numbers are small, but the direction is clear: buyers are using AI assistants to research suppliers, and some of that research ends in an enquiry.

How the Auction Works: CPM, CPC and Conversion Optimisation

ChatGPT Ads uses what OpenAI describes as a relevance-weighted, second-price auction. In plain terms, when several ads are eligible for a placement, the system considers both how relevant each ad is and how much each advertiser has bid, and the winner pays based on what was needed to win rather than its full bid.

The three ChatGPT Ads campaign objectives plus automated Maximize results bidding

Relevance is weighted heavily, so a well-matched ad can beat a higher bid from a poorly matched one. That rewards advertisers who invest in specific copy, accurate context hints and landing pages that genuinely answer the question.

The Three Campaign Objectives

Every campaign starts with an objective, and the objective decides how you pay. OpenAI offers three:

ObjectiveYou pay forBest for
ViewsImpressions, per 1,000 (CPM)Awareness and reach
ClicksValid clicks (CPC)Driving traffic to a page
ConversionsValid clicks (oCPC) or impressions (oCPM)Optimising towards leads or sales

With the Conversions objective, you are not billed per conversion. You choose click billing or impression billing, and delivery is steered towards people likely to complete your chosen conversion event.

Click Billing Against Impression Billing for Conversions

OpenAI explains the difference clearly. With click billing, you pay for valid clicks and delivery optimises towards conversions that follow a click. With impression billing, you pay for impressions and delivery optimises towards a broader set of conversions, including those influenced by someone seeing an ad and returning later.

OpenAI’s guidance notes that impression billing optimises for more conversions, because it can count view-influenced outcomes. For a lead generation business, click billing is usually the easier starting point because every pound spent maps to a visit you can see in GA4 and HubSpot. Impression billing may suit once you trust your measurement and want to capture the full effect.

Setting Bids

Advertisers set a maximum bid at ad group level: a maximum CPM for Views campaigns or a maximum CPC for Clicks campaigns. For CPC campaigns, OpenAI recommends starting with a maximum bid of $3 to $5 per click, quoted in US dollars. Ads Manager may also show bid-strength guidance indicating whether a bid is likely to limit delivery.

There is also an automated option called Maximize results, selected by default for eligible new ad groups. It sets and adjusts bids to get as many results as possible within your budget. OpenAI is clear that it prioritises volume and does not guarantee any particular cost per click, cost per acquisition or return.

A Simple Illustration of the Auction

The exact mechanics are not published in detail, so treat this as an illustration of the principle rather than a formula. Imagine two advertisers eligible for the same placement below an answer about choosing a garden office. Advertiser A bids a maximum of £4 per click with a generic ad and a homepage landing page. Advertiser B bids £3 with an ad about insulated garden offices and a landing page showing prices.

Because the auction weights relevance as well as bid, Advertiser B can win despite the lower bid. In a second-price model, B then pays roughly what was needed to beat A’s relevance-adjusted position, which may be less than its £3 maximum.

The lesson is that bids and relevance work together. Raising a bid on a poorly matched ad is an expensive way to buy delivery. Improving the ad and landing page is usually the cheaper route. The same principle runs through Google Ads Quality Score, so teams who already manage search campaigns well will recognise it.

What Cannot Be Changed Later

Several settings are locked once a campaign is created. The objective and, for conversion campaigns, the billing method and the chosen conversion event cannot be edited in place. To change them you clone or rebuild the campaign.

Decide the objective and conversion event before you build, because a wrong choice means starting again and losing whatever learning the first campaign had gathered. Country, currency and time zone on the account itself are also fixed at creation, so set them correctly for a UK business from the start.

Context Hints Against Keywords: Rethinking Targeting

The biggest mental shift for anyone used to Google Ads is that there are no keywords. Beyond geography, platform and custom audiences, the main lever for influencing where your ads appear is a feature called context hints, set at ad group level.

Five rules for writing ChatGPT Ads context hints: what, who, when, one idea and keeping them current

What Context Hints Are

OpenAI describes context hints as additional details about your business, product or service that help its systems understand when your ads may be relevant. OpenAI describes them as relevance signals rather than exact-match keywords, targeting rules or delivery instructions. They do not guarantee your ad will appear in any particular conversation.

OpenAI’s guidance suggests thinking about three angles: what you offer, who it helps and when it is useful. Each hint should be a clear, natural phrase focused on one idea rather than a string of keywords.

Good and Bad Examples

OpenAI’s own example contrasts “running shoes” with “cushioned everyday running shoes for beginners training for their first 5K”. The second tells the system far more about who the product is for and when it is relevant. Here is how that translates for typical UK lead generation businesses:

  • Weak: “garden rooms”. Stronger: “insulated garden offices for people working from home in the West Midlands who need planning advice”.
  • Weak: “HubSpot agency”. Stronger: “HubSpot onboarding for UK B2B companies moving from spreadsheets to a CRM for the first time”.
  • Weak: “solicitor”. Stronger: “fixed-fee commercial lease reviews for small business tenants in Suffolk”.

Each stronger hint describes a real need, a specific audience and a situation. That gives the matching system something to work with when a conversation drifts close to your offer without using your exact words.

What Hints Cannot Do

Hints cannot enforce geography, schedules or exclusions. OpenAI uses the example of a plumber: “Residential plumbing services available in Chicago” describes the business, but “Show this ad only to people in Chicago” is a delivery instruction the system will not follow. Location targeting has its own controls at campaign level.

Use campaign settings for restrictions and context hints for relevance. Mixing the two is one of the most common setup mistakes, and it produces ad groups that look targeted on paper but behave broadly in practice.

Context Hints for Multi-Location and Multi-Service Businesses

Businesses with several services or locations need to plan hints carefully. Because hints cannot restrict geography, location targeting should be set at campaign level, with hints describing the service and the customer situation. A firm with offices in Ipswich and London might run separate campaigns for each area, each with hints that reflect the local offer.

Multi-service businesses benefit from one ad group per service, each with a handful of hints covering what, who and when. A marketing agency, for example, might write “HubSpot onboarding for UK manufacturers replacing spreadsheets”, “Google Ads management for local trades businesses with small budgets” and “LinkedIn lead generation for B2B consultancies selling to finance directors”, each in its own ad group.

This mirrors how GLO structures lead generation campaigns on other platforms: one clearly defined need per ad group, so the data shows which need produces the best customers.

Keeping Hints Accurate

OpenAI asks that hints be accurate and updated when features, pricing, availability or service areas change. A hint promising a fixed price you no longer offer creates a mismatch between ad and landing page, which can hurt relevance and, in the worst case, breach advertising rules.

Review hints whenever your offer changes, just as you would review ad copy. Treat them as part of the creative rather than a one-off technical setting.

How Hints Relate to Your Ad Groups

Because hints sit at ad group level, the structure of your ad groups determines how precise they can be. An ad group covering every service you offer forces you into vague hints. Separate ad groups for distinct needs, each with its own hints, copy and landing page, let every element stay specific.

Before You Start: Account, Verification and Crawler Access

A few practical checks before launch will save days of delay. Most of the problems UK advertisers hit in their first week come from steps that happen before a campaign is even built.

Four checks that delay a first ChatGPT Ads campaign: verification, eligibility, crawler access and country limits

Business Verification

OpenAI runs a business verification check on new advertiser accounts. Its troubleshooting guide says the review can take up to seven business days. Verification can fail if OpenAI cannot confirm the accuracy of your business identity, ownership or affiliations.

Make sure your account, ads and landing pages all describe the same business accurately. If your trading name differs from your registered company name, set the Brand Name and Legal Name fields correctly rather than leaving the defaults from sign-up.

Industry Eligibility

Not every industry can advertise. OpenAI’s ad policies prohibit some categories outright and restrict others to approved advertisers or particular countries. Financial services, health and other regulated sectors should read the OpenAI ad policies and the eligibility table in the help centre before planning a campaign.

OpenAI notes that being a registered or licensed business does not automatically make you eligible for every category or market. If an ad is rejected, the fastest route is the appeal option in Ads Manager.

Let OpenAI’s Crawlers Reach Your Landing Pages

This step catches out a surprising number of advertisers. OpenAI uses crawlers to check that landing pages comply with its policies and to understand when ads are relevant. Its guidance on allowing OpenAI web crawlers says you must allow OAI-AdsBot and recommends allowing OAI-SearchBot too.

If your robots.txt, firewall, CDN or bot protection blocks those crawlers, your ads can be rejected or underdeliver. Cloudflare and similar services sometimes return 403 errors to automated visitors by default. Ask whoever manages your website to check each layer: robots.txt, the web application firewall, and any CAPTCHA or JavaScript challenge.

Home-Country Restrictions on New Accounts

OpenAI notes that some new self-serve accounts can initially advertise only in their home country. A UK account may therefore be limited to UK delivery at first, even if you select other countries.

The account becomes eligible for other supported countries once identity verification is approved and you have reached a required level of spend in your home country. For most UK businesses targeting UK buyers, this restriction will not matter, but it is worth knowing if you sell into Ireland or Europe.

Billing

Billing must be set up before campaigns can deliver. Self-serve accounts use a card and a postpay model, so spend accrues as ads run and is charged to the saved card. Your bank may show a temporary authorisation hold when you add the card.

Spend figures in Ads Manager can lag by seven to eight hours, according to OpenAI. A zero spend figure on launch day does not mean nothing has been charged, so avoid panicking and doubling budgets based on early numbers.

Setting Up Ads Manager as a UK Business, Step by Step

With the groundwork in place, the setup itself is straightforward. Here is the sequence GLO follows for a UK lead generation account.

Six steps to set up ChatGPT Ads Manager for a UK business, from account creation to launch

Step 1: Create the Account

Go to ads.openai.com and sign in with the OpenAI account you use for ChatGPT, or create one. OpenAI recommends that one person creates the advertiser account and then invites colleagues, rather than several people creating separate accounts for the same business.

You can either enter your website URL and let Ads Manager draft a first ad from your site, or skip straight to the dashboard. Skipping is usually better for a considered setup, because you will want to plan campaign structure before creating ads.

Step 2: Set Country, Currency and Time Zone

Choose United Kingdom, pounds sterling and UK time. OpenAI states that these three settings cannot be changed after account creation. Getting them wrong means opening a fresh account, so double-check before submitting.

Step 3: Confirm Names and Logo

In Settings, check the Account Name, Brand Name and Legal Name. The Brand Name is what appears beside your ads in ChatGPT; the Legal Name is used for billing and verification. Upload a clear logo, because it appears as the favicon on every ad.

Step 4: Set Up Billing

Add a billing profile with your business name, address and invoice email, then add a payment card. Campaigns will not deliver until billing is complete.

Step 5: Install Measurement Before the First Campaign

Install the OpenAI Pixel and, ideally, the Conversions API before launching anything. Conversion-optimised campaigns need at least one supported standard conversion event already in place, and even a Clicks campaign is far more useful when you can see which clicks became leads. The measurement section below covers this in detail.

Step 6: Build the First Campaign

Create a campaign, choose the objective, set the budget type and amount, select dates and target locations, and optionally choose platforms: Android app, Android web, desktop web, iOS app or iOS web. Within the campaign, create ad groups with context hints, then ads with titles, copy, images and landing pages.

Step 7: Submit for Review and Monitor

Ads must pass review before they serve. In Ads Manager, check that each ad shows as Serving. If something shows as Not serving, hovering over the status explains why. Impressions, clicks and click-through rate refresh roughly every 15 minutes; spend updates more slowly.

Who Should Own the Account

Ownership matters more than it first appears. The advertiser account should belong to the business, with agencies or freelancers invited as users rather than creating accounts in their own names. If you change agency later, you keep the account history, the Pixel data and the audiences.

Ads Manager supports multiple users with different roles, so give each person the access they need and no more. Keep at least two people from the business with administrator access, so a holiday or a departure never locks you out. GLO follows the same principle for every platform it manages, as explained on our about GLO page: the client owns the accounts and the data.

Using HubSpot Instead

If your business runs on HubSpot, you can also create ChatGPT Ads campaigns from inside HubSpot once the accounts are connected. That route suits teams who want ad creation, lead capture and reporting in one place, and it is covered in the HubSpot section below.

Structuring Campaigns and Ad Groups

Structure decides how well you can learn from a campaign. A flat, catch-all setup produces muddy data; a structure built around distinct customer needs produces clear answers.

Four structuring rules for ChatGPT Ads campaigns and ad groups

Campaigns Follow Business Goals

OpenAI recommends structuring campaigns around distinct business goals, with separate campaigns for meaningfully different objectives, regions or product categories. For most UK service businesses, that means one campaign per service line or per objective. A garden room company might run one campaign for garden offices and another for granny annexes, because the buyers, prices and landing pages differ. A B2B agency might separate HubSpot onboarding from paid media management for the same reason.

Ad Groups Follow Customer Needs

Within each campaign, ad groups should each focus on one product, theme or customer need. OpenAI’s example is a plumber with separate ad groups for emergency repairs and preventive maintenance: the same business, different needs, different messages. If two audiences need different landing pages, they need different ad groups. That keeps context hints specific and lets the reporting tell you which need produces the best leads.

Multiple Ads Per Ad Group

OpenAI advises building for coverage rather than one message. A high volume and diverse set of ads, with multiple title and copy variations per offering, gives the system more opportunities to find relevant placements. Each variation should introduce a different angle rather than repeating the same idea.

For a first test, three to five ads per ad group is a reasonable starting point. Vary the angle: one on price clarity, one on speed, one on a specific outcome, one on credentials.

A Worked Structure for a UK B2B Service

Here is an example structure for a HubSpot consultancy running its first ChatGPT Ads test:

  • Campaign 1: HubSpot onboarding (Clicks objective, UK)
  • Ad group: First-time CRM buyers moving from spreadsheets
  • Ad group: Companies migrating from Salesforce or Pipedrive
  • Campaign 2: HubSpot automation (Clicks objective, UK)
  • Ad group: Sales teams wanting automated follow-up sequences
  • Ad group: Marketing teams connecting ads to pipeline reporting

Each ad group has its own context hints, three to five ads and a landing page built for that specific need. Two campaigns with two ad groups each is enough to learn from without spreading a small budget too thin.

When to Split Into More Campaigns

Start lean, then split when the data justifies it. A campaign should be divided when one ad group is clearly outperforming the others and you want to give it its own budget, when you add a new region, or when you are ready to test a different objective such as conversion optimisation.

Splitting too early spreads a small budget across too many units, so none of them gathers enough data to judge. As a rough rule, wait until each ad group has produced a meaningful number of clicks and at least a few conversions before restructuring. Match each new campaign to a workflow in HubSpot, as described in our HubSpot sales automation setup guide, so new leads are routed correctly from day one.

Keep Naming Consistent

Consistent naming pays off when you report across platforms. Use the same service and audience labels you use in Google Ads, Meta and HubSpot campaigns, so comparisons are easy and nobody has to translate between naming systems when pulling numbers together.

Writing Ads for a Conversation Rather Than a Results Page

Writing for ChatGPT Ads is different from writing a Google search ad. The person seeing your ad has usually just read a detailed answer to a specific question. They are thinking, comparing and weighing options, so slogans and generic calls to action land badly.

A four-week routine for testing ChatGPT ad copy, from launching angles to repeating fortnightly

What OpenAI Recommends

OpenAI’s guidance on creating ads says that clear, specific, benefit-focused copy is more likely to be relevant. Benefit-focused copy explains the practical value of what you offer, who it is for and when it helps, rather than relying on a catchy line. The guidance breaks down into a few practical rules:

  • Spell out the value in the title, so it is immediately clear what is offered.
  • Keep titles concise and informative rather than broad marketing language.
  • Use the copy to add new information, not to repeat the title.
  • Represent the offer accurately, because misleading or inconsistent ads may be rejected.

Match the Mindset of the Conversation

Think about what the person was just discussing. Someone who asked ChatGPT how to choose a HubSpot partner has read a list of criteria. An ad that speaks to those criteria, with certification level, UK base and a clear onboarding scope, fits the moment. An ad saying “Grow Your Business Today” does not.

A useful exercise is to write down five questions a buyer might ask ChatGPT before choosing a business like yours, then draft one ad that would feel helpful directly below each answer.

Examples of Stronger and Weaker Ads

Weaker title: “The UK’s Best Garden Rooms”. Stronger title: “Insulated Garden Offices, Priced Online in Minutes”. The second tells the reader exactly what they get and why clicking helps. Weaker copy: “Contact us today for a free quote.” Stronger copy: “Design your garden office in 3D, see the full price including installation and get a PDF quote by email.” Specific, practical and aligned with a buyer who has just asked about costs.

Ad Examples for Three UK Sectors

These examples show how the principles translate into real copy. They are illustrations to adapt, not tested winners. Commercial law firm. Title: “Fixed-Fee Commercial Lease Reviews in Suffolk”. Copy: “A solicitor reviews your lease within five working days for a fixed fee, with a plain-English summary of the risks before you sign.”

B2B software consultancy. Title: “HubSpot Onboarding for Teams Leaving Spreadsheets”. Copy: “We migrate your contacts and deals, build your pipeline and train your team in four weeks, with a fixed scope agreed upfront.” Home improvement firm. Title: “Loft Insulation Quotes With Savings Shown Upfront”. Copy: “Enter your loft size and current heating bill to see an estimated annual saving and a fixed installation price.”

Each one names the service, the audience and the practical benefit. Any figure, timescale or price in copy like this must be accurate for your business, because ChatGPT Ads for UK businesses fall under the same advertising rules as every other channel.

Images

OpenAI recommends simple, relevant images that support the message rather than abstract or cluttered visuals. A clear photo of the product or a real project usually outperforms stock imagery of people pointing at laptops. Make sure the image matches the title and copy, so the ad feels coherent at a glance.

Avoid Claims You Cannot Back Up

Claims such as “best”, “cheapest” or “number one” need evidence under UK advertising rules, and OpenAI’s own policies reject misleading ads. Specific, verifiable facts persuade better than superlatives anyway, especially for an audience that has just read a balanced comparison from ChatGPT.

A Copy Testing Routine for the First Month

Without search term reports, copy testing is one of the few levers you control directly, so run it deliberately. In week one, launch three to five ads per ad group with clearly different angles. In week two, compare click-through rate and conversion rate by ad, then pause the weakest one or two.

In week three, write replacements that build on what worked. If the price-focused ad won, try two new price angles. If the credentials ad won, test different proof points. Repeat each fortnight.

Judge ads on conversions where you can, not just clicks, because a catchy title that attracts curious clicks without enquiries costs money and teaches the system the wrong lesson. GLO’s SEO copywriting team applies the same discipline to ad copy, and our guide to A/B testing for CRO explains how to judge results fairly.

Landing Pages That Match a Researched Question

Landing pages carry more weight on ChatGPT Ads than on most platforms, because OpenAI uses them as a relevance signal as well as a conversion tool. A strong page helps your ad get shown and helps the visitor take the next step.

Four requirements for a ChatGPT Ads landing page that matches a researched question

Send People to the Most Relevant Page

OpenAI advises linking to product pages, collections or content pages rather than defaulting to a homepage. The visitor has just asked a specific question, so the landing page should pick up that conversation where it left off rather than restarting with a company introduction.

If the ad promises a price, the page should show a price or a way to get one. If the ad speaks to companies migrating from another CRM, the page should talk about migration from the first screen.

Answer the Question Before Asking for Anything

ChatGPT users arrive informed. They have read a summary and probably seen several options. Pages that front-load substantive information, such as prices or price ranges, timescales, what is included and who it suits, work better than pages that hide everything behind a form.

This is where lead generation tools come into their own. A calculator, quote tool or product configurator gives the researched buyer exactly what they came for and captures them at the same time. Pairing ChatGPT Ads with an interactive tool is one of the most promising combinations available to UK lead generation businesses.

Make the Next Step Obvious

One clear call to action, repeated where it makes sense, beats a page with six competing options. A short form, a booking link or a quote request should be visible without hunting. GLO’s guide to conversion rate optimisation in the UK covers the layout principles in depth.

Technical Requirements

Landing pages must be valid, reachable and must not block OpenAI’s user agents. Test every URL before launch, on mobile as well as desktop, and confirm it loads quickly. Add UTM parameters directly to each destination URL; OpenAI confirms that static tracking parameters persist on ad clicks and show up in your analytics tools.

A consistent UTM pattern such as utm_source=chatgpt and utm_medium=paid, with the campaign and ad group names in utm_campaign and utm_content, makes ChatGPT traffic easy to isolate in GA4 and HubSpot.

Mobile Matters

Many Free plan users access ChatGPT on their phones. Platform targeting lets you choose Android app, Android web, iOS app, iOS web or desktop web, and the reporting can be segmented by platform. Check performance by platform after the first fortnight, because a page that converts well on desktop may struggle on a small screen.

Measurement: OpenAI Pixel, Conversions API and Consent

Without measurement, a ChatGPT Ads test produces clicks and opinions. With it, you can see which ads, ad groups and landing pages produce leads, and the platform can optimise towards them. Set measurement up before you launch, not after the first invoice arrives.

Five stages of ChatGPT Ads measurement, from consent through Pixel and Conversions API to deduplication

The Two Ways to Send Conversions

OpenAI’s conversion measurement works through a data source in Ads Manager, fed by one or both of two methods:

  • The OpenAI Pixel, a JavaScript snippet on your website that records events such as form submissions and stores OpenAI’s click reference in a first-party cookie.
  • The Conversions API, which sends conversion events from your server or CRM directly to OpenAI, so they are not lost to ad blockers or browser restrictions.

OpenAI recommends using both together for more resilient measurement. When the same conversion is sent through both, include the same event ID so OpenAI can remove the duplicate.

How Attribution Works

When someone clicks a ChatGPT ad, OpenAI appends a click reference called oppref to the landing page URL. The Pixel captures it and stores it so later conversions can be linked back to the ad. If you use the Conversions API, you can include oppref in the event.

A conversion is reported when OpenAI receives an event that matches a conversion configured for the campaign, within the attribution window, and can connect it to an eligible click. OpenAI also supports advanced matching, using hashed contact details to link conversions when the click reference is missing, and modelled measurement for some otherwise unattributed conversions.

Automatic Advanced Matching

The Pixel can detect customer information in recognisable forms, normalise it and hash it with SHA-256 in the browser before sending it with a conversion. OpenAI states that raw customer information is not sent through this feature. Hashing is not the same as consent. Hashed personal data is still personal data under UK GDPR, so automatic advanced matching needs to be covered by your privacy notice and your consent setup.

Consent Under PECR and UK GDPR

The OpenAI Pixel sets a cookie and reads data from the visitor’s device. Under PECR, non-essential cookies and similar technologies need the user’s consent, as the ICO explains in its guidance on cookies and similar technologies. Advertising pixels are not essential. OpenAI’s own documentation makes the same point: only send conversion data after giving users clear information and obtaining any consents required by law. In practice this means:

  1. Adding the OpenAI Pixel to your consent management platform as a marketing or advertising tag.
  2. Firing it only after the visitor has accepted that category.
  3. Updating your cookie policy and privacy notice to name OpenAI as a recipient.
  4. Considering server-side events through the Conversions API for leads who have given consent on the form itself.

The ICO notes that its cookie guidance is being revised following the Data (Use and Access) Act, so check the current position before launch.

A Consent-Friendly Tracking Setup in Practice

For most UK websites, the cleanest setup runs through Google Tag Manager and an existing consent banner. The OpenAI Pixel is added as a tag that fires only when the visitor has accepted marketing cookies. A lead submission event fires on the thank-you page or the form success state, again subject to consent.

Server-side, the Conversions API can send lead events from your CRM or a server container for visitors who gave consent, including the oppref value captured from the landing page URL and a shared event ID. That combination recovers conversions lost to browser restrictions without bypassing anyone’s choices.

Test the whole chain with consent accepted and with consent declined, and document what happens in each case. That record is useful for your own compliance and for explaining the numbers when someone asks why Ads Manager and HubSpot differ.

Choosing the Right Conversion Event

Conversion-optimised campaigns use one supported standard conversion event per campaign, and custom events are not currently supported for that objective. For lead generation, the natural choice is the lead submission event fired when a form is completed. Pick an event that reflects a genuine enquiry. Optimising towards newsletter sign-ups or page views will teach the system to find people who do those things, which is rarely what a sales team wants.

What to Report Each Week

A short weekly report keeps a test honest. Include spend, impressions, clicks, click-through rate and cost per click from Ads Manager; conversions recorded by the Pixel or API; and, most importantly, leads, qualified leads and opportunities from HubSpot.

Add one line of sales feedback on lead quality and one line on what changed that week. The changes log is what lets you explain movements later, especially on a beta platform where delivery can shift for reasons outside your control.

Event Quality

Ads Manager shows event quality warnings when the data coming from your Pixel or API has problems, such as missing parameters or duplicate events. Check these after installation and again after any website change, because a broken tag silently undermines every optimisation decision the platform makes.

Connecting ChatGPT Ads to HubSpot for Pipeline Attribution

On 16 September 2026, in its Fall 2026 Spotlight announcement, HubSpot described itself as the first CRM to integrate with ChatGPT Ads. For UK businesses already running HubSpot, this is the most practical way to see whether ChatGPT Ads produce pipeline rather than just clicks.

Four things the HubSpot integration adds to ChatGPT Ads, from campaign creation to cost reporting

What the Integration Does

According to HubSpot and OpenAI’s setup guide for ChatGPT Ads in HubSpot, the integration lets you:

  • create ChatGPT Ads campaigns from inside HubSpot
  • send traffic to HubSpot landing pages and add new leads directly to the CRM
  • trigger follow-up workflows for those leads
  • view performance by campaign, ad group or ad, including impressions, clicks, contacts, customers and cost
  • compare ChatGPT Ads with your other ad channels side by side in HubSpot’s ads reporting

The ability to see contacts and customers alongside cost is what turns a novelty channel into a measurable one. It answers the question every business owner eventually asks: did this actually bring in revenue?

Who Can Use It

The integration is available to HubSpot customers wherever ChatGPT Ads are available, including the UK. You need a ChatGPT Ads account, a Marketing Hub Free, Starter, Professional or Enterprise account, and permission to connect ad accounts in HubSpot.

One wrinkle catches some businesses out. If your company uses ChatGPT Business or Enterprise, a global admin must approve HubSpot’s Ads access in the OpenAI Admin Console before the connection will work. That access is off by default, and being an Ads account admin or HubSpot admin does not automatically give you that permission.

Connecting the Accounts

The connection runs from HubSpot’s side:

  1. In HubSpot, open Settings, then Marketing, then Ads.
  2. Select Connect account and choose ChatGPT Ads.
  3. Sign in to your ChatGPT Ads account and follow the authorisation flow.
  4. Review the requested permissions and confirm.

Once connected, you can create a campaign from Marketing, then Ads, by choosing ChatGPT Chat Card and completing the ad, targeting, and budget and schedule tabs.

Building the Follow-Up

A lead from ChatGPT has usually done more research than a typical paid search lead. The follow-up should acknowledge that and move quickly. A HubSpot workflow can create a sales task immediately, send a confirmation email with something genuinely useful, and set the lead source so reporting stays clean.

Our guide to HubSpot sales automation setup walks through building these workflows, and GLO’s HubSpot service can configure the full loop from ad click to closed deal.

Reporting That Means Something

Within a few weeks, the integration should let you answer three questions: how many contacts ChatGPT Ads produced, how many became opportunities, and how many became customers, each against cost. That is the same standard you should hold every channel to. GLO’s article on measuring ad ROI with HubSpot explains how to build those reports.

Custom Audiences and First-Party Data

Custom audiences let you upload your own customer or prospect lists to ChatGPT Ads. They are useful, but the thresholds mean they suit some businesses far better than others.

Custom audience thresholds for ChatGPT Ads: twenty-five thousand matched users minimum and processing time

How Custom Audiences Work

Audiences are built from email addresses, phone numbers, SHA-256 hashed emails or phone numbers, or Google Advertising IDs. Once processed, which OpenAI says usually takes 20 to 30 minutes, they can be used in three ways:

  • Inclusion: limit a campaign to people in the audience.
  • Exclusion: stop a campaign reaching people in the audience.
  • Bid adjustments: raise or lower bids for an ad group when the viewer matches the audience.

The Size Thresholds

Inclusion audiences and bid adjustments need at least 25,000 matched users, and OpenAI recommends 100,000 or more. Exclusions can use smaller audiences. For most UK SMEs, exclusion is the realistic use case. Few small businesses have 25,000 matched contacts, but almost any business can upload its existing customer list to stop paying for clicks from people who already buy from it.

Preparing a Clean Upload From HubSpot

Most UK businesses will build their audience lists from HubSpot. Create an active list of current customers, using lifecycle stage or a customer property, and another of open deals. Export only the identifier columns you need, usually email address and phone number, and remove anyone who has objected to marketing.

Normalise the data before upload: lower-case email addresses, strip spaces, and format phone numbers consistently with the country code. Clean data raises the match rate, which matters when exclusion lists are already small. Schedule a monthly refresh, or automate the export if your team has the capability. GLO’s HubSpot team sets up these lists as part of channel launches, so audiences stay current without manual effort.

Data Protection Considerations

Uploading customer data to an advertising platform is processing personal data, and it needs a lawful basis under UK GDPR. Your privacy notice should explain that customer data may be shared with advertising platforms for audience matching, and people who have objected to marketing should be removed before upload.

Note that OpenAI advises avoiding custom audiences for campaigns targeting the EEA or Switzerland while ads personalisation is unavailable there. That restriction does not apply to the UK, but it matters if you plan to extend campaigns into Ireland or Europe.

Practical Uses for UK Businesses

  • Exclude existing customers from acquisition campaigns to save budget.
  • Exclude open deals in your pipeline, so sales conversations are not interrupted by generic ads.
  • For larger B2C brands with big lists, include lapsed customers in a win-back campaign.
  • Use bid adjustments to compete harder for high-value segments once lists are large enough.

Keep exported lists up to date. OpenAI lets you add, remove or replace audience members without rebuilding campaigns, so a monthly refresh from HubSpot is easy to maintain.

Budgeting a First Test, With a Worked Example

The most common budgeting mistake with a new channel is spending too little to learn anything, then concluding the channel does not work. The second most common is spending heavily before tracking is in place. A well-planned test avoids both.

Funnel from sixteen hundred pounds of ChatGPT Ads spend through to three new clients

How Budgets Work in Ads Manager

You can choose a campaign-total budget or a daily budget. OpenAI explains that a daily budget is an average over a seven-day period: spend can reach twice the daily figure on an individual day but will not exceed seven times the daily figure across the week. The budget week runs from Sunday to Saturday in your account’s time zone.

A campaign-total budget is the stricter control, and you can switch from campaign-total to daily but not back again. OpenAI suggests that new advertisers start with a daily budget so they can watch delivery before committing to a total. The minimum daily budget varies by billing currency and is shown in Ads Manager.

A Worked Example

The figures below are illustrative assumptions to show the arithmetic. They are not benchmarks, because OpenAI says none exist yet. Use your own numbers once data arrives. Suppose a UK B2B service business sets a daily budget of £40 for one campaign with two ad groups, running for six weeks. Total planned spend is £40 multiplied by 42 days, or £1,680.

Assume an average cost per click of £2.50. That buys roughly 670 clicks. If the landing page converts 4 percent of visitors into enquiries, the test produces about 27 leads, at roughly £62 per lead.

Turning Leads Into a Decision

Leads alone do not decide anything. Suppose sales qualifies half of those 27 leads, giving 13 opportunities, and wins one in four, giving three new clients. If each client is worth £6,000 in first-year revenue, the test generated £18,000 against £1,680 of media spend.

Change the assumptions and the answer changes. At a 2 percent conversion rate, the same budget produces about 13 leads and roughly one or two clients. The point of the worked example is to set your decision rules before you see the results, so the test ends in a clear yes or no rather than a vague feeling.

Setting Decision Rules in Advance

Before launch, write down:

  1. The maximum cost per qualified lead you can accept, based on your close rate and client value.
  2. The minimum number of clicks or leads needed before you judge results, so one good or bad week does not decide everything.
  3. What you will change if results fall short: copy, landing page, context hints or objective.
  4. What scaling looks like if results succeed, such as doubling budget or adding a conversion-optimised campaign.

What Not to Do With a Test Budget

A few habits reliably waste test budgets. Changing everything at once makes it impossible to know what worked. Judging results after three days ignores the fluctuation OpenAI warns about during the beta. Sending traffic to the homepage throws away the relevance advantage the platform rewards.

Running the test without sales team involvement is another common error. The sales team knows which leads were genuine, and their view of lead quality matters more than any platform metric. Agree before launch how lead quality will be scored, even if it is just a simple good, average or poor rating in HubSpot.

Finally, avoid comparing a six-week ChatGPT test with a mature Google Ads account that has years of optimisation behind it. Compare against what Google Ads looked like in its first weeks, or against your own decision rules. Our guide to online lead generation in the UK sets out how to compare channels fairly.

Moving to Conversion Optimisation

Once conversions are flowing reliably, consider a Conversions objective campaign optimising towards lead submissions. OpenAI says the goal for conversion-optimised campaigns is to improve conversion outcomes rather than simply use the full budget, so spend may be lower than the budget while the system learns.

Where ChatGPT Ads Sit Against Google Ads, Meta and LinkedIn

ChatGPT Ads do not replace the channels you already run. They reach people at a particular moment, and understanding that moment helps you decide what share of budget and attention they deserve.

Bar chart of an illustrative monthly paid media split with a ChatGPT Ads test alongside Google and Meta

Against Google Ads

Google Ads search campaigns capture demand expressed as a query: someone types what they want and sees ads matched to it. You get keyword control, search term reports, mature bidding and years of benchmark data.

ChatGPT Ads reach people while they work through a decision in conversation. There is less control and less data, but the person has often already defined their needs in detail. For most UK lead generation businesses, Google Ads remains the core paid search channel and ChatGPT Ads is the test budget alongside it. GLO’s guide to setting up Google Ads AI Max covers how Google is moving towards similar intent-based matching.

Against Meta Ads

Meta reaches people while they scroll, before they are actively looking. It is excellent for creating demand, retargeting and visually led products, and its lead forms make capture easy. ChatGPT Ads reach people who are already researching, which is a later and usually more qualified moment.

The two can work together: Meta builds awareness and interest, and ChatGPT Ads catch the same type of buyer when they start asking detailed questions.

Against LinkedIn Ads

LinkedIn offers precise B2B targeting by job title, company size and industry, which ChatGPT Ads cannot match. For enterprise B2B, LinkedIn Ads remain far more controllable. Remember too that ChatGPT Business and Enterprise users do not see ads. If your buyers mainly use ChatGPT through a company licence, they are outside the ChatGPT Ads audience entirely, which strengthens the case for LinkedIn in enterprise B2B.

Against Microsoft Advertising

HubSpot added a Microsoft Advertising integration in the same Fall 2026 release, highlighting reach across Bing, Copilot, MSN and Outlook. For businesses interested in AI assistant audiences, Microsoft’s Copilot placements are another option to test alongside ChatGPT.

An Illustrative Budget Split

Here is how a test might sit within an existing budget, using illustrative numbers rather than a recommendation for any specific business. Take a UK SME spending £3,000 a month on paid media: £2,000 on Google Ads search, £700 on Meta and £300 on retargeting.

A sensible test would carve out around £600 to £1,200 a month for ChatGPT Ads for six weeks, funded partly by trimming the weakest-performing Google Ads campaigns rather than cutting what already works. That keeps total spend steady and makes the comparison fair, because the money moves from a known weak performer to a new channel.

At the end of the test, the HubSpot data decides whether the budget stays, grows or returns to search. The point is to make the decision on evidence, using the same qualified lead and revenue measures for every channel, rather than on how exciting the new platform feels.

Choosing the Mix

A sensible starting position for a UK SME already running Google Ads is to allocate a small, fixed test budget to ChatGPT Ads for six to eight weeks, measured through HubSpot alongside every other channel. If cost per qualified lead compares well, expand. If not, refine once and reassess. GLO’s lead generation service builds these multi-channel plans around each client’s sales cycle.

Paid Placement and Organic AI Visibility: Running Both

Paid ads and organic visibility in AI answers are separate things, and it helps to keep them apart in your planning. OpenAI states that ads do not influence ChatGPT’s answers and that advertisers cannot shape, rank or alter responses. Buying ads will not make ChatGPT recommend you in its answers.

Four shared foundations that support both ChatGPT Ads relevance and organic AI visibility

What Organic AI Visibility Means

Organic visibility is whether ChatGPT and other assistants mention your business when people ask relevant questions. It depends on how well your website explains what you do, how often reputable sources mention you, the reviews you have and how clearly your content answers buyer questions.

GLO’s guide on how to appear in ChatGPT covers the organic side, and our article on Google AI Overviews for UK businesses explains the equivalent changes in Google search.

Why You Need Both

The two work on different timescales. Organic visibility builds slowly through content, reputation and technical work, and once established it costs nothing per click. Paid placement can start today but stops when the budget stops.

Running both gives you coverage while the organic work matures, and the paid data can inform it. If an ad group about a particular service performs well, that tells you buyers are asking about it, which is a strong signal to strengthen your organic content on the same topic.

Shared Foundations

Several things help both at once:

  • Clear, specific service pages help OpenAI judge ad relevance and help AI systems understand your business for organic answers.
  • Allowing OAI-SearchBot supports organic discovery as well as ad landing page checks.
  • Accurate, consistent business information across your site, directories and reviews supports verification and organic credibility.
  • Content that answers buyer questions gives both your ads and your organic presence something substantive to point to.

Using Ad Data to Plan Content

Paid results can guide organic priorities. When an ad group about a particular service or buyer need performs well, it suggests real demand for that topic in AI conversations. That is a cue to strengthen the matching service page, add an in-depth guide or improve the FAQ content AI systems draw on.

Equally, an ad group that struggles despite good copy may point to a landing page that does not answer the question clearly enough. Fixing that page helps paid relevance, organic visibility and conversion rate together, which makes it one of the highest-value changes available.

Measuring the Two Separately

Keep reporting distinct. Paid ChatGPT traffic will carry your UTM parameters and the oppref click reference. Organic referrals from chatgpt.com arrive without them. In GA4, separate the two so you can see whether paid activity lifts organic referrals over time, or whether they simply sit side by side. GLO’s SEO service can set up this kind of AI referral tracking alongside the organic work.

Brand Safety, Disclosure and UK Advertising Rules

Advertising inside a conversational product raises reasonable questions about where your brand appears and what rules apply. The good news is that both OpenAI’s policies and UK rules point in the same direction: be accurate, be clear and stay out of sensitive contexts.

Four CAP Code requirements that apply to ChatGPT Ads shown to UK users

OpenAI’s Brand Safety Approach

OpenAI says its policy is to place ads only near conversations that are safe, appropriate and aligned with user trust and brand safety, with safeguards designed to keep ads away from sensitive or brand-unsafe contexts. Ads shown in logged-out experiences are limited to content appropriate for all audiences.

You do not get placement-level reporting showing exactly which conversations your ads appeared beside. That is a trade-off of the privacy model, and it means you rely on OpenAI’s safeguards rather than your own placement exclusions.

Clear Labelling

Every ChatGPT ad is labelled as sponsored and visually separated from the answer. This matters for UK compliance, because the CAP Code requires marketing communications to be obviously identifiable as such. The platform handles the label; your job is to make sure the ad content itself is not misleading.

The CAP Code Still Applies

ChatGPT Ads shown to UK users are non-broadcast advertising, so the UK Code of Non-broadcast Advertising, known as the CAP Code, applies just as it does to Google, Meta or a printed brochure. In practice the rules that matter most are:

  • Substantiation: you must hold evidence for objective claims, including “best”, “fastest” or price comparisons.
  • Pricing: quoted prices should be accurate and include any compulsory costs, such as VAT for consumer audiences.
  • No misleading omissions: important conditions should not be hidden behind the click.
  • Sector rules: financial, health and other regulated sectors have extra requirements.

Consistency Between Ad and Landing Page

OpenAI’s ad review checks that the ad and its landing page describe the same offering. The CAP Code expects the same honesty. A headline promising a fixed price that the landing page qualifies heavily risks both rejection and a complaint.

Keeping Records

Keep a simple record of every ad version, its launch date and the evidence behind any objective claim it makes. If a competitor or customer complains to the ASA, you will need to show substantiation quickly, and if OpenAI rejects an ad, the record makes an appeal straightforward.

This is good practice on every platform, but it matters more on a new one where review standards are still settling. Running ChatGPT Ads for UK businesses in regulated or competitive sectors without that paper trail is an avoidable risk.

Regulated Sectors

If you work in financial services, healthcare, legal services or another regulated field, check three things before launching: OpenAI’s eligibility for your category in the UK, your own regulator’s rules on financial promotions or professional advertising, and whether your claims meet both. Some categories are currently restricted to a small group of approved advertisers.

Troubleshooting Delivery and Measurement Problems

New platforms have teething problems, and ChatGPT Ads is still in beta. These are the issues UK advertisers are most likely to hit, in roughly the order they appear.

Four common ChatGPT Ads problems: ads not serving, low delivery, missing conversions and mismatched numbers

Ads Not Serving at All

Work through OpenAI’s own checklist:

  1. Review status: ads must pass review; check they show as Serving.
  2. Account verification: a banner appears if verification is incomplete.
  3. Billing: a payment method problem stops delivery.
  4. Budget: a campaign stops when it reaches its budget limit, and a daily budget campaign that uses its weekly allowance early can pause until the next budget week.
  5. Dates: campaigns do not serve outside their start and end dates.

Ads Rejected

Read the rejection reason in Ads Manager and identify whether the issue is the text, the image or the landing page. The most common causes are inconsistency between ad and page, claims the reviewer considers misleading, and landing pages the crawler cannot reach. Use the appeal option if you believe the decision is wrong; OpenAI describes it as the fastest route.

Low Delivery

If ads serve but impressions are low, the system may not be finding enough relevant conversations or your bids may be uncompetitive. Check bid-strength guidance in Ads Manager, add more ad variations with genuinely different angles, broaden context hints that are too narrow, and confirm your landing page clearly describes the offer.

Remember that early delivery fluctuates. OpenAI says changes to its delivery systems during the beta can make delivery and spend vary, so give changes a few days before judging them.

Conversions Not Recording

If clicks arrive but no conversions appear, test the full journey yourself. Click through a live ad, submit the form and check whether the event appears in Ads Manager. Common causes include the Pixel firing only after consent that the tester declined, the form submitting through a method the Pixel does not detect, or a missing event ID causing deduplication problems between Pixel and API.

Leads Arriving but Not Converting to Sales

Sometimes the channel produces leads, but sales finds them weak. Start by reading a sample of the enquiries. If buyers misunderstood the offer, the ad or landing page is promising something different from what you sell, so tighten the copy and make price or scope clearer.

If buyers understood the offer but are the wrong size or budget, add qualifying questions to the form or show indicative prices earlier, so unsuitable buyers rule themselves out. If leads are good but go cold, the problem is follow-up speed rather than the channel, which a HubSpot workflow and a same-day call usually fix.

GLO’s CRO service works through exactly this kind of diagnosis, combining landing page changes with feedback from the sales team.

Numbers Not Matching Across Tools

Ads Manager, GA4 and HubSpot will rarely agree exactly. They use different attribution windows, count different things and lose data to consent choices and ad blockers at different points. Pick one source of truth for business decisions, usually HubSpot for leads and revenue, and use the others to diagnose problems rather than to argue about totals.

A 30-Day Launch Plan for UK Businesses

Here is a practical plan for a UK business launching its first ChatGPT Ads test, assuming HubSpot is the CRM and Google Ads is already running.

A thirty day ChatGPT Ads launch plan in four stages from foundations to the decision point

Week 1: Foundations

  • Open the Ads Manager account with UK country, GBP currency and UK time zone.
  • Complete business verification and billing.
  • Check that OAI-AdsBot and OAI-SearchBot can reach your landing pages.
  • Install the OpenAI Pixel through your consent management platform and set up the Conversions API if possible.
  • Connect ChatGPT Ads to HubSpot.

Week 2: Build

  • Choose one or two services with clear landing pages and a strong reason for a researched buyer to enquire.
  • Build one campaign per service with two ad groups each, based on distinct customer needs.
  • Write specific context hints and three to five ads per ad group.
  • Set up UTM parameters and HubSpot workflows for the new lead source.
  • Write down your decision rules: target cost per qualified lead, minimum data before judging, and next steps either way.

Weeks 3 and 4: Launch and Learn

  • Launch with a daily budget and a Clicks objective.
  • Check delivery, rejections and event quality daily for the first few days.
  • After two weeks, review click-through rate, cost per click and conversion rate by ad group and platform.
  • Pause the weakest ads, write new variations and refine context hints.
  • Feed sales team feedback on lead quality back into copy and landing pages.

After 30 Days

Compare cost per qualified lead against your decision rules and against your other channels in HubSpot. If ChatGPT Ads compare well, increase budget and test a conversion-optimised campaign. If they do not, make one round of meaningful changes to landing pages or offers before deciding.

How GLO Approaches ChatGPT Ads

GLO does not start with the platform. Every engagement begins with a Growth Strategy: understanding the business, its commercial objectives and the audiences most likely to buy. For each audience, GLO builds vertical profiles covering the job titles that sign off on purchases, the pain points driving decisions, the objections that slow conversion and the lead magnets that attract enquiries.

That groundwork is what makes context hints, ad copy and landing pages specific enough to work on a platform without keywords. Only once it is in place does GLO configure the HubSpot architecture, set up measurement and launch campaigns, so every new channel is judged on the same pipeline data as the rest.

As a HubSpot partner running Google Ads, Meta Ads and LinkedIn Ads for UK businesses, GLO can test ChatGPT Ads alongside your existing channels and show you, in the same reports, whether it earns a permanent place in your budget. Our AI sales automation service then makes sure every new lead is followed up quickly.

Questions to Ask Before Handing a Test to an Agency

If an agency will run your test, ask how they will measure it before asking how they will build it. Useful questions include: who owns the advertiser account and Pixel data, how conversions will reach HubSpot, what decision rules they propose, how they will handle consent, and how they will report lead quality as well as volume.

An agency that can answer those clearly will give you a test you can trust. Our guide on how to choose a lead generation agency in the UK covers the wider questions worth asking before you sign anything.

Frequently Asked Questions

Can UK Businesses Advertise on ChatGPT?

Yes. OpenAI lists the United Kingdom as available for self-service Ads Manager, so a business whose advertising and billing entity is based in the UK can sign up at ads.openai.com and run campaigns directly.

Five frequently asked questions about ChatGPT Ads for UK businesses with short answers

New accounts go through business verification, which OpenAI says can take up to seven business days, and must set up card billing before ads deliver. Some industries are restricted or prohibited under OpenAI’s ad policies, so regulated businesses should check eligibility first.

How Much Do ChatGPT Ads Cost?

There is no fixed price. ChatGPT Ads run on an auction where you bid per thousand impressions or per click, or use conversion-optimised bidding billed by click or impression. OpenAI recommends starting CPC campaigns with a maximum bid of $3 to $5 per click and shows the minimum daily budget for your currency in Ads Manager.

OpenAI has not published performance benchmarks, so any quoted UK cost per lead is an estimate. The reliable approach is a small, fixed test budget with clear decision rules, measured through to qualified leads in your CRM.

Do ChatGPT Ads Influence ChatGPT’s Answers?

No. OpenAI states that ads run on separate systems from the chat model and that advertisers cannot shape, rank or alter ChatGPT’s responses. Ads appear below the answer, labelled as sponsored and visually separated from it.

That means paying for ads will not get your business recommended in ChatGPT’s answers. Organic visibility depends on your website content, reputation and the sources that mention you, which is why many businesses run organic AI visibility work, covered in our guide on how to appear in ChatGPT, and paid ChatGPT Ads side by side.

How Do ChatGPT Ads Compare With Google Ads for Lead Generation?

Google Ads offers keyword control, search term reports and years of benchmark data, so it remains the core paid search channel for most UK lead generation businesses. ChatGPT Ads use context hints rather than keywords and reach people while they work through a decision in conversation.

The audience also differs: only Free and Go users see ChatGPT ads, so people using ChatGPT through a company Business or Enterprise licence are excluded. Most businesses should treat ChatGPT Ads as a measured test alongside Google Ads rather than a replacement.

Can I Track ChatGPT Ads Leads in HubSpot?

Yes. HubSpot describes itself as the first CRM to integrate with ChatGPT Ads. Once connected, you can create campaigns in HubSpot, capture leads on HubSpot landing pages, enrol them in workflows and see contacts, customers and cost by campaign, ad group or ad.

You need a HubSpot Marketing Hub account at any tier, including Free, and permission to connect ad accounts. If your company uses ChatGPT Business or Enterprise, a global admin may need to approve HubSpot’s Ads access in the OpenAI Admin Console first.

Start Small, Measure Properly, Then Decide

ChatGPT Ads give UK businesses a new way to reach buyers while they research, and the companies that learn the channel early will have a head start when competition grows.

Four takeaways for UK businesses testing ChatGPT Ads in 2026

  • UK self-service is live, but the platform is in beta and has no published benchmarks.
  • Targeting relies on context hints, ad copy and landing pages rather than keywords.
  • Install the OpenAI Pixel through your consent platform, and add the Conversions API where you can.
  • Connect ChatGPT Ads to HubSpot so you judge the channel on pipeline, not clicks.
  • Run a fixed test budget with decision rules written down before launch.

If you would like help planning a ChatGPT Ads test that fits your wider lead generation strategy, book a Growth Strategy session with GLO and we will map it out with you.