ChatGPT Ads: Can Answers Really Drive Business?

For more than twenty years, paid search has been built around a simple concept. Users type in a query and advertisers bid to appear alongside the results. The keyword auction has shaped how budgets are planned, how campaigns are structured and how performance is measured.

That model may soon be challenged. Generative AI is changing how people access information – we are moving from a “search” economy to an “answer” economy.

Google has been heading in this direction for some time through featured snippets, knowledge panels and AI Overviews, all of which reduce the need to click through to websites. Conversational AI and chatbots take this a step further by delivering answers directly, without the user ever seeing a traditional search results page.

There is, however, a commercial reality behind this shift. Producing real-time, context-aware responses requires a lot more computing power than serving static links. OpenAI incurred losses of around $8 billion in the first half of 2025 according to the Financial Times [1], underlining how expensive large scale AI delivery can be.

AI platforms like ChatGPT now face a big challenge: how to fund the most power-hungry information engine ever. It is early days, but the idea of advertising within AI chatbots has become inevitable.

The ChatGPT Advertising Rollout

OpenAI has confirmed that advertising is being introduced to the “free” tier of ChatGPT. Alongside this, it has launched a lower-cost subscription called “ChatGPT Go” at £8 per month.

This plan provides additional features but still includes adverts, following a model already familiar from streaming platforms. Higher tiers such as Plus, Pro and Enterprise will remain ad free.

Advertising is currently being tested in beta in the United States, with a limited number of advertisers involved while formats and controls are refined. Broader international availability is expected later, but timelines remain unclear.

Unlike traditional search advertising, ads do not appear as promoted links above and below a list of organic results. Instead, ChatGPT displays adverts at the end of a response, triggered by the context of the conversation.

A user asking for advice on travelling to Mexico, for example, might see an advert for a hotel or travel provider that fits the topic.

Early indications suggest pricing may be impression-based, rather than relying on auction-driven cost-per-click models. If this proves accurate, it would represent a departure from Google’s approach, placing more emphasis on conversational relevance than on historic performance data.

For marketers, this points towards a form of conversational marketing built around intent rather than keywords. The challenge will be ensuring that any brand insertion feels genuinely relevant to the discussion taking place.

Key Differences from Traditional Search Ads

Feature	Traditional Search Ads	AI Conversational Ads
Primary Model	Keyword + AI intent matching (Smart Bidding, Performance Max)	Natural Language Intent/Context
Placement	Sponsored links at top and bottom of SERP	Contextual suggestions beneath the conversation
Targeting	First-party data, contextual signals, Privacy Sandbox	Potential session context
User Action	Click to website / click to call	Conversational discovery / click to website

What We Know About ChatGPT Ads So Far

Conversational Discovery

The first ad formats ChatGPT will go live with will be pretty standard ads, with external links to the brand website. However in the future ads within ChatGPT will be designed to be interactive.

Users will ask follow-up questions within the ad itself, as well as click through to a website. This shifts the experience from passive browsing to active dialogue, although brands will need to provide data to power the chat.

If done well, this could make ads feel like part of the solution rather than an interruption. If done badly, it risks breaking trust by intruding on what feels like a personal exchange.

Premium Pricing, Limited Data

Reports suggest OpenAI is offering ads in ChatGPT at a high price (around $60 for every 1,000 views [2]), which is about three times what Meta charges. But advertisers won’t get much detail in return.

The performance reporting is basic, covering things like total views or clicks, with no data on what happens after that, like whether someone made a purchase.

Limited Targeting

OpenAI has stated that conversation history and training data will not be sold to advertisers. Targeting is expected to rely on the immediate context of the conversation rather than long-term user profiles or cross-session tracking.

That approach makes it harder for advertisers to do the kind of targeting and tracking they’re used to on platforms like Google or Meta. Whether this level of privacy still allows campaigns to be planned and scaled effectively remains an open question.

Answer Independence

OpenAI has committed to keeping advertising separate from the core chat response. Advertisers cannot pay to alter the AI’s organic text, and ads will be clearly labelled.

This separation is important, although it will be interesting to see how distinct the ads will be. We know from experience that Google has constantly changed the appearance of search ads to try and ‘blend’ more with organic results – will we see a similar approach here?

Brand Safety and Controls

Generative AI introduces the risk of ads appearing alongside inaccurate or inappropriate content. OpenAI has indicated that adverts will not appear in sensitive areas such as health, politics or mental health.

Additional safeguards are expected to include exclusion lists, age gating for under-18s and contextual targeting that does not rely on persistent advertising profiles.

The Compounding Effect of Trust

While OpenAI has suggested that this format could help smaller brands compete, long-term effectiveness is likely to favour those already seen as credible. Ads are more likely to perform well when the brand already aligns with the chat advice being given.

In these cases, paid placement and organic mentions will reinforce one another. This is where investment in AI SEO and brand authority becomes a strategic long term play.

Where Do ChatGPT’s Competitors Stand on AI Advertising?

OpenAI has taken a bold first step, but their competitors are taking different approaches so far.

Google Gemini

Google benefits from established revenue streams through Search and Workspace, reducing pressure to monetise its chatbot directly.

Demis Hassabis of Google DeepMind has publicly stated there are no current plans to place ads within Gemini itself, however there have been noises about “AI mode” receiving ads. Google also continues to focus on monetising AI through AI Overviews embedded in traditional search results.

Claude

Anthropic positions Claude as a safety-first alternative, prioritising accuracy and trust. They have clearly stated that Claude will remain ad-free. They have even (ironically) created their own ads presenting the risk of ads within a chat model.

Perplexity

Perplexity has experimented with sponsored follow-up questions, but its commercial direction appears unsettled. Senior advertising leadership changes suggest that the long-term strategy is still being shaped.

Microsoft Copilot

Microsoft is integrating ads across its wider ecosystem, including Bing and Office products, using Copilot as part of a broader commercial framework rather than a standalone ad product.

Things for Businesses to Think About Now

Widespread access to ChatGPT advertising is still some way off, and the ad formats in the trial are unlikely to represent the final version. That said, there are sensible considerations businesses can start working through now, well before this is available as a marketing channel.

Build Credibility Before Considering Spend

Conversational ads will likely be more powerful if the brand is already present in the organic conversation. AI systems favour sources that already demonstrate experience and reliability.

Businesses should focus on producing clear, helpful content such as detailed guides, original insights and real case studies that show genuine knowledge of their sector. When advertising does arrive at scale, it will work best as reinforcement, not as a shortcut to trust

Start Thinking in Conversations, Not Slogans

Chatbot advertising is a blend of guidance and promotion. Businesses should begin documenting the real questions customers ask, the concerns they raise and the comparisons they make before choosing a supplier.

These insights will shape future conversational messaging far more effectively than traditional ad copy.

Expect a Different Path to Decision

The familiar journey of click, landing page and conversion is likely to compress. In AI-driven environments, discovery and evaluation can happen within a single interaction.

Businesses should consider how easily their product details, support information and service documentation can be accessed directly in chat. Reducing friction and answering questions quickly will become increasingly valuable.

Plan for Different Measures of Success

When conversational advertising becomes viable, success will not be defined by last-click performance alone. Influence within conversations, brand presence and lead quality will all matter. A broader way of measuring success needs to be considered.

Conclusion

Advertising within ChatGPT reflects a clear reality. Conversational AI is expensive to run, and platforms must generate revenue to sustain it.

For businesses, this introduces the possibility of a new channel that has the ability to reach high-intent users deep in conversation. It won’t have the reach of other ad platforms, but the ad format is novel.

That said, this remains an early-stage experiment. Targeting controls, pricing models and performance measurement are all still unclear. OpenAI’s US beta is for selected enterprise brands where ad budgets are in the region of $1 million [3], suggesting that widespread access to most businesses is some way off.

The key question is whether advertising enhances or undermines the experience. If ads feel helpful and relevant, they may become an accepted part of conversational tools. If they feel intrusive, users may quickly lose trust.

How this balance is struck will shape how ChatGPT ads develop. Advertising is likely to remain part of the platform out of necessity, but the format, controls and user experience will almost certainly be refined before they get rolled out to the masses.

Sources

[1] https://www.ft.com/content/a169703c-c4df-46d6-a2d3-4184c74bbaf7

[2] https://www.theinformation.com/articles/openai-seeks-premium-prices-early-ads-push

[3] https://www.mediapost.com/publications/article/412189/beta-advertisers-begin-testing-new-chatgpt-ads-wit.html