Key takeaways
- OpenAI is testing Sponsored Agents, allowing people to move from seeing an advert into a separate conversation with the advertiser.
- The new format could change advertising from a click-based experience into a longer, question-and-answer journey before someone visits a website.
- Advertisers will be able to create and manage more of their ChatGPT Ads activity using natural-language instructions rather than traditional campaign interfaces.
- OpenAI is introducing AI-assisted creative tools, including suggested ad copy and imagery and optional contextual text customisation.
- HubSpot and Shopify integrations bring ChatGPT Ads closer to CRM, ecommerce and first-party business data.
- For advertisers, the bigger opportunity is not simply another ad placement. It is the possibility of influencing consideration while a customer is actively researching and comparing options.
ChatGPT advertising is beginning to look less like traditional paid search and more like something entirely different.
OpenAI is testing Sponsored Agents, an advertising experience that allows someone to continue from an advert into a dedicated conversation with the business behind it.
Instead of an advert simply saying “click here”, a potential customer could ask further questions about a product, service, specifications, pricing or suitability before deciding whether they want to visit the advertiser’s website.
At the same time, OpenAI is adding more AI directly into the advertiser side of the platform, including campaign creation through ChatGPT Work, creative assistance inside Ads Manager and integrations with HubSpot and Shopify.
Taken together, these developments give us a much clearer picture of what OpenAI wants ChatGPT Ads to become.
This is not simply Google’s search advertising model moved into ChatGPT.
The advertising experience itself is becoming conversational.
What are Sponsored Agents in ChatGPT?
Sponsored Agents are being tested as a way for users to have a separate conversation with a business after interacting with an advert.
For example, imagine somebody is researching office furniture and sees an advert for an ergonomic chair.
Instead of immediately being pushed towards a product page, they might be able to ask:
- Does it work for somebody who is 6ft 3?
- Is the lumbar support adjustable?
- What is the difference between this model and the cheaper version?
- Does it come assembled?
- What is the returns policy?
The sponsored agent can answer those questions before the customer decides whether to continue to the advertiser’s website.
OpenAI says these sponsored conversations are clearly labelled and kept separate from both the user’s original ChatGPT conversation and ChatGPT’s independent answers. It also says advertising does not influence ChatGPT’s answers and advertisers do not receive access to private conversations.
For advertisers, however, the important change is the behaviour.
Traditional PPC normally asks an advert to achieve one immediate objective:
Win the click.
Conversational advertising could ask it to do something much broader:
Help the customer reach a decision.
Why conversational advertising could change the customer journey
Search advertising has traditionally relied on a relatively simple sequence.
Someone searches.
They see an advert.
They click.
They land on a website.
The advertiser then hopes the landing page answers enough questions to produce a conversion.
ChatGPT changes that sequence because much of the research can happen before the website visit ever takes place.
People already use conversational AI to compare products, understand services, investigate problems and narrow down choices. OpenAI describes ChatGPT Ads as reaching people while they are exploring options, comparing alternatives and making decisions.
Sponsored Agents potentially push advertising directly into that research process.
That could produce a journey that looks more like:
Question → recommendation → advert → conversation → qualification → website → conversion
rather than:
Keyword → advert → click → landing page → conversion
That difference matters.
An advertiser may eventually receive fewer completely uninformed clicks but more visitors who have already asked several questions and decided the product or service is relevant to them.
For ecommerce businesses, that could mean better-informed customers.
For B2B advertisers, it could potentially mean the beginning of lead qualification before the prospect even reaches the website.
ChatGPT Ads are moving beyond keywords
Another important distinction is how advertising opportunities are understood.
Google Ads has increasingly moved beyond exact keywords, but paid search remains rooted in queries and auctions.
ChatGPT conversations contain considerably more context.
Someone might not type:
Instead, they might explain:
We spend around £30,000 a month on Google Ads for a B2B company, but lead quality has dropped since moving more campaigns onto broad match and Performance Max. What should we look at?
That contains far more information than a conventional keyword.
There is potential intent.
There is an approximate advertising budget.
There is a business model.
There is a problem.
There are existing campaign types.
There is potentially a buying requirement.
OpenAI’s own advertiser guidance says ChatGPT Ads consider conversational intent and other relevance signals rather than simply matching against individual search queries.
For PPC teams, that requires a different approach to messaging.
The challenge becomes less about squeezing a keyword into a headline and more about having enough useful, specific advertising messages for the platform to identify where your business genuinely fits into a conversation.
OpenAI is also changing how campaigns are built
The conversational approach is not limited to users seeing adverts.
OpenAI is bringing the same model into campaign management.
Advertisers will be able to use prompts within ChatGPT Work to create, update and analyse advertising campaigns through its Ads Manager tools.
A business could potentially provide its website, marketing information or campaign brief and then describe what it wants to achieve conversationally rather than manually constructing every part of a campaign through an advertising interface.
OpenAI is also adding AI assistance to Ads Manager that can suggest copy and imagery based on the advertiser’s landing page and objective.
Advertisers remain able to review and edit those assets before using them.
There is also optional AI-powered text customisation, which can adapt existing headlines and descriptions to better suit the context of a conversation, including translating advertising into the user’s preferred language.
None of that eliminates the need for good PPC strategy.
If anything, it makes the quality of the advertiser’s inputs more important.
An AI system given a poorly defined proposition, weak conversion tracking and vague campaign objectives will simply be able to produce bad advertising more efficiently.
PPC Geeks’ View
The most interesting part of this announcement is not another AI creative tool.
We already have plenty of those.
It is the possibility that the advert itself becomes part of the sales conversation.
That changes what advertisers should optimise.
Historically, PPC teams have spent enormous amounts of time improving click-through rates, landing page conversion rates and cost per acquisition.
Those metrics will remain important.
But conversational advertising introduces another layer between impression and website visit.
The quality of the conversation could start determining whether the click happens at all.
A business therefore needs to think carefully about the information its sponsored experience can provide.
If customers consistently ask questions about delivery, pricing, compatibility, implementation, integrations or contract terms, those are not peripheral details anymore.
They may become advertising inputs.
For agencies and advertisers, this creates a significant new requirement: marketing information needs to match commercial reality.
There is little value in building an incredibly persuasive AI advertising experience that attracts people the sales team cannot convert.
The same principle we apply to Google Ads becomes even more important here.
Optimise towards the business outcome, not merely the easiest measurable interaction.
ChatGPT Ads and first-party data
OpenAI’s HubSpot and Shopify integrations are another important part of the announcement.
HubSpot becomes OpenAI’s first CRM partner for ChatGPT Ads, while Shopify is its first ecommerce partner.
For businesses using HubSpot, the integration is intended to bring advertising workflows closer to customer and CRM context.
For Shopify merchants in the US, OpenAI is offering a ChatGPT Ads app for creating and managing campaigns and monitoring results. Shopify product data is also already being incorporated through Shopify Catalog.
This direction should interest performance marketers.
The strongest advertising platforms increasingly depend on understanding what happens after the initial conversion.
A cheap lead is not necessarily a valuable lead.
A sale with poor margin is not necessarily a valuable sale.
A customer who cancels after a month is not equivalent to somebody who remains for three years.
If ChatGPT Ads develops stronger connections between advertising activity, ecommerce transactions and CRM outcomes, advertisers should eventually be able to judge the channel on commercial value rather than surface-level engagement.
That will be essential if ChatGPT Ads is going to compete seriously for performance marketing budgets.
What advertisers should do next
ChatGPT Ads is still developing, so advertisers should avoid treating every new feature as a reason to shift large budgets immediately.
But there are several things worth preparing now.
1. Start thinking beyond traditional keyword lists
Review the questions customers ask before buying.
Speak to your sales team.
Look through enquiry forms, CRM notes, live chat conversations and support requests.
Identify the problems customers are actually trying to solve.
Conversational advertising will reward businesses that understand the full buying conversation, not simply the final keyword entered before conversion.
2. Improve the information on your website
AI advertising systems need strong source material.
Make sure your landing pages clearly explain:
- who the product or service is designed for;
- what problems it solves;
- important features and limitations;
- pricing or pricing structure where appropriate;
- delivery or implementation information;
- relevant specifications;
- FAQs;
- evidence and proof points.
Generic landing pages will give AI systems generic material to work with.
3. Create more genuinely different advertising messages
OpenAI recommends producing varied advertising creative rather than multiple versions that essentially repeat the same message.
That makes sense in a conversational environment.
Think about different customer concerns.
One advert might focus on price.
Another might focus on speed.
Another might focus on expertise.
Another could focus on product compatibility or technical capability.
The system then has more meaningful options when deciding what is relevant to a particular conversation.
4. Get conversion measurement right
Do not judge a new advertising channel solely by clicks.
Make sure conversions can be traced through to meaningful business outcomes wherever possible.
For ecommerce, that could include revenue, margin, repeat purchases or customer lifetime value.
For lead generation, it could include qualified leads, opportunities and closed sales.
ChatGPT Ads becoming conversational does not change the fundamentals of good PPC measurement.
It makes them more important.
5. Treat ChatGPT Ads as an experiment before treating it as a strategy
There is understandable excitement around AI advertising.
That does not mean every business should immediately move substantial budget away from Google Ads, Microsoft Ads, Meta or other established channels.
Test it.
Measure it.
Compare lead quality and commercial outcomes.
Then scale when the evidence supports it.
Early advertiser results published by OpenAI show that ChatGPT Ads can already contribute measurable acquisition for some businesses, but individual case studies should not be treated as guarantees of performance for every advertiser.
What this means for the future of PPC
ChatGPT Ads is starting to answer one of the biggest questions surrounding advertising inside generative AI:
What does an advert look like when the interface is a conversation rather than a search results page?
Sponsored Agents give us one possible answer.
Instead of interrupting the experience and immediately pushing somebody elsewhere, an advertiser can potentially remain useful within the conversation.
That creates opportunities, but it also raises the standard.
Advertising cannot rely entirely on a clever headline if the customer can immediately ask five detailed follow-up questions.
Businesses will need accurate product information, strong propositions, useful content, clean measurement and a much deeper understanding of the questions customers ask while deciding what to buy.
That is ultimately where ChatGPT Ads could become particularly interesting.
Google Ads has spent decades becoming exceptionally good at capturing existing intent.
Conversational advertising could increasingly compete for the stage immediately before that: helping people define their requirements, compare possible solutions and decide what they actually want.
For PPC teams, that is worth watching closely.
If you’re unsure whether your existing paid advertising is making the most of your budget, a free PPC audit from PPC Geeks can identify tracking problems, wasted spend, weak campaign structure and opportunities to improve performance before adding another channel into the mix.
Frequently asked questions
What are Sponsored Agents in ChatGPT?
Sponsored Agents are a new advertising experience being tested by OpenAI. After interacting with an advert, users can start a separate, clearly labelled conversation with a business-sponsored agent and ask additional questions before deciding whether to visit the advertiser’s website.
Can advertisers see someone’s private ChatGPT conversation?
OpenAI says advertisers do not receive access to users’ private ChatGPT conversations. Sponsored Agent conversations are also kept separate from ChatGPT’s independent responses and the user’s original conversation.
Do ChatGPT Ads influence ChatGPT’s answers?
OpenAI says advertising remains separate from ChatGPT’s answers and does not influence the responses users receive.
What can advertisers do with ChatGPT Ads Manager?
OpenAI is developing tools for creating and managing campaigns, analysing performance and generating advertising creative. Advertisers can also use natural-language instructions through ChatGPT Work for parts of the campaign workflow.
Can ChatGPT Ads create advertising copy automatically?
OpenAI is adding AI tools that can suggest advertising copy and imagery using information such as an advertiser’s landing page and campaign objective. Advertisers can review and edit the suggested creative before using it.
What are the HubSpot and Shopify integrations?
HubSpot is OpenAI’s first CRM integration for ChatGPT Ads, while Shopify is its first ecommerce partner. The integrations are designed to bring advertising activity closer to existing customer, product and commercial data.
Should advertisers start using ChatGPT Ads?
ChatGPT Ads is worth testing where it is available and suitable for the business, but it should initially be treated as a measurable new channel rather than an automatic replacement for existing PPC activity. Track qualified leads, sales and revenue alongside traditional advertising metrics before deciding how much budget to allocate.
How are ChatGPT Ads different from traditional search advertising?
Traditional search advertising usually responds to a short query and aims to generate a website click. ChatGPT Ads can use the context of a wider conversation, and Sponsored Agents introduce the possibility of additional interaction with an advertiser before the user leaves ChatGPT.
