OpenAI's New ChatGPT Business Analytics Could Disrupt AI Visibility Tools

OpenAI's Businesses in ChatGPT is a beta platform that allows eligible companies to claim and manage their business profiles, track qualifying mentions and website clicks, and offer AI-powered Business Agents to customers when eligible.

OpenAI's new Businesses in ChatGPT beta lets eligible companies claim profiles, manage information, and track qualifying mentions, website clicks and discovery topics. Marketers must decide how to use this data to assess AI visibility investments.

As customers use AI assistants to research products and services, companies are investing in content, search optimization and third-party tools that test which prompts produce mentions in AI-generated answers. Those tools can show how a business appears in tested scenarios, but they cannot provide a complete record of actual ChatGPT activity. OpenAI's analytics offer a direct, though limited, view of what happens on its platform.

That could change how businesses and marketers evaluate third-party AI visibility services. Those third-party providers still offer prompt-level research and comparisons across AI systems, while OpenAI's dashboard does not reveal individual customer prompts or connect mentions and clicks to sales.

Businesses in ChatGPT gives companies a new way to measure parts of their presence in ChatGPT, but not the power to determine when the assistant recommends them. As consumers increasingly use AI assistants to discover businesses, how much visibility and control will companies have over their presence in those conversations?

OpenAI launched Businesses in ChatGPT in beta on November 10, giving eligible companies a central place to manage how they are represented on the platform. Called Business Hub, it combines three functions: maintaining business information, reviewing aggregated engagement data and creating optional Business Agents that answer customer questions for eligible companies.

For example, a retailer can update its business details and share a product feed containing current prices and availability. A service provider can clarify what it offers, while the Insights dashboard can show qualifying mentions, website clicks and discovery topics. These capabilities give businesses a way to provide more accurate information and better understand how people encounter them in ChatGPT.

Business Hub is separate from the paid ChatGPT Business subscription, and companies do not need an advertising account or ChatGPT Enterprise agreement to manage an eligible profile.

Some online businesses can create a new listing if they aren't already listed in ChatGPT. Local stores and restaurants, however, can currently claim only existing listings for U.S. locations. OpenAI can also reject businesses over safety or integrity concerns or delay access for certain business categories. The company has not published a complete list of eligible countries or industries, so the full extent of availability remains unclear at this time.

Business Agents have additional restrictions. They are currently limited to qualifying U.S. ecommerce and online service businesses with websites. OpenAI evaluates both a company's industry and how its agent will be used, and can delay access for categories it cannot yet support safely. The company plans to expand availability but has not announced a timeline.

To claim a business profile, an owner or authorized representative searches for the company in Business Hub and selects its listing. OpenAI offers four verification methods: a code sent to a business phone by text or automated call, a code sent to a business email, a DNS record added to the company's domain settings, or a connected Shopify account. OpenAI determines which methods are available for each listing. If someone else has already claimed the business, there is a process for disputing ownership.

Claim approval is only the first step. Businesses must also complete identity verification through Persona before they can publish profile changes, connect eligible product feeds or access their analytics. Depending on the business, Persona verification can require a government-issued ID, business registration information and proof that the person is authorized to represent the company. A blue verified badge appears only after Persona verification is approved, not simply when the initial business claim is accepted.

Once verified, companies can update their descriptions, websites, logos and information about their products or services. OpenAI reviews these changes before publishing them, and approved updates can take time to appear in ChatGPT listings or answers. ChatGPT can also continue drawing on other sources when describing a business.

But managing how a business appears in ChatGPT is only part of the picture. Companies also want to know whether potential customers are finding them, and what those interactions can tell them about their visibility.

Once verified, businesses can access ChatGPT's Insights dashboard to see activity that third-party AEO services have generally had to estimate. The dashboard reports qualifying mentions, website clicks and discovery topics, but understanding what those numbers represent is essential to interpreting their value.

Consider an online retailer and a neighborhood restaurant. For the retailer, ChatGPT counts a mention when it references the company's website in an answer. If the assistant names the brand without referencing its website, that appearance isn't counted. For the restaurant, a mention occurs when ChatGPT displays a result for its location, even if the customer never opens the listing. Those differences matter when businesses try to understand how often they're appearing in AI-generated answers.

Website clicks measure how often people select links to a business's website from supported ChatGPT experiences. They don't establish whether someone completed a website visit or made a purchase. Discovery topics provide another perspective by identifying common themes in ChatGPT answers associated with the business. For example, a retailer might discover that its website frequently appears in discussions about product comparisons or delivery options, without seeing the actual questions customers asked.

The dashboard summarizes activity over the previous 28 days, providing businesses with overall counts and themes rather than details about individual customer interactions. To protect user privacy, OpenAI only reports certain insights when enough conversations or interactions have occurred, although it has not disclosed the minimum reporting thresholds. As a result, businesses might see a blank metric when activity hasn't reached the required level. A blank result can also mean data is still being collected or reporting is unavailable, so it doesn't necessarily indicate zero activity. Businesses cannot access individual customer prompts or conversations, and OpenAI confirmed to Digiday that the beta does not include share-of-voice measurements comparing visibility with competitors.

These insights give businesses direct measurements of specific ChatGPT activity, rather than a complete picture of their brand visibility or why customers choose one company over another. Understanding those limitations becomes especially important when comparing OpenAI's data with the estimates provided by third-party AI visibility services.

For years, businesses have invested in search engine optimization (SEO) to appear near the top of Google's results, hoping potential customers would click through to their websites. Traditional search still matters, but consumers are increasingly using AI assistants to research products, compare companies and ask for recommendations without first scrolling through search results. That change has created a new challenge for businesses: understanding whether AI assistants recommend them at all.

A growing industry of third-party AI visibility services has emerged to help answer that question. Platforms such as Peec.AI, Otterly.AI and Yolando test selected customer questions across AI systems, including ChatGPT, Claude, Gemini and Perplexity. They measure how often a brand appears in those answers, calculate its share of voice relative to competitors and provide visibility scores that summarize performance. Businesses can also examine the individual prompts being tested, identify which competitors appear and track changes over time.

For example, an online retailer might discover that its brand appears in 40% of the ChatGPT answers its provider tested, but only 15% of the tested Claude responses. It can examine the questions behind those results, see which competitors appeared instead and identify opportunities to improve its content. However, those percentages reflect the provider's selected tests, not every question actual customers ask.

OpenAI's new first-party analytics introduce a challenge to that measurement model. Businesses can now compare third-party estimates with data drawn directly from qualifying ChatGPT activity, although the two approaches measure different things and aren't necessarily directly comparable.

Accuracast CEO Farhad Divecha told Digiday that his team has relied on services such as Peec.AI, Otterly.AI and Semrush to estimate brand visibility. He wants to know how closely their numbers match OpenAI's own measurements and whether the providers will incorporate the new data. Significant discrepancies, he suggested, could cause his company to reconsider paying for those services. Digiday did not report results from such a comparison.

Sonata Insights founder and chief analyst Debra Aho Williamson also told Digiday that OpenAI's analytics could weaken third-party providers' influence. However, she expects continued demand because ChatGPT is only one of several AI platforms, and OpenAI isn't sharing everything it knows.

That distinction leaves businesses with two different ways to evaluate AI discovery. Third-party services can test selected questions, compare competitors and measure visibility across multiple AI platforms, but they cannot access the questions real customers ask or establish why an AI assistant recommends one business over another. OpenAI measures qualifying activity from actual ChatGPT interactions, but doesn't reveal the prompts behind those mentions, provide competitive share-of-voice measurements or show activity across other AI platforms. Each approach provides information the other lacks, leaving businesses without a complete picture of how AI recommendations influence customer discovery.

For third-party providers, the challenge is demonstrating value beyond estimating ChatGPT mentions. For businesses, combining both approaches could make it easier to understand where they're being recommended, how competitors perform and whether their visibility efforts are making a difference.

But knowing whether customers are discovering a business is only part of the picture. What happens when someone finds that company through ChatGPT and wants to learn more? OpenAI's Business Agents give eligible businesses a way to continue that interaction by answering customer questions directly within ChatGPT.

Imagine someone using ChatGPT to compare two pairs of running shoes. After ChatGPT mentions a retailer, the customer can open its business profile and, if the company has published a Business Agent, start a separate conversation to ask about sizing, product differences or return policies.

Business Agents allow eligible companies to answer customer questions directly within ChatGPT, using information and instructions the business provides. Companies can review and test their agents before publishing them, including how they respond to common questions. The agent is clearly identified as representing the business, and its conversation is separate from ChatGPT's independent answers.

During the beta, Business Agents are available only to eligible U.S. ecommerce and online service businesses with websites. Companies must complete business and identity verification and meet OpenAI's industry and safety requirements. A verified business profile alone does not guarantee access to an agent.

For retailers, this creates an opportunity to help customers evaluate products without immediately leaving ChatGPT. However, Business Agents cannot currently complete purchases, make bookings or access customer orders. Those activities take place on the company's own website, where the agent can direct customers when they're ready to take the next step.

Businesses can publish eligible Business Agents without purchasing advertising. But, OpenAI also offers sponsored agent experiences through ChatGPT ads, although access is currently limited to selected advertisers. These paid placements are available to eligible Free and Go users, while Plus, Pro and business subscriptions remain ad-free. Unlike sponsored placements, ordinary Business Agents are accessed through business profiles and are not dependent on advertising. Neither experience gives businesses control over ChatGPT's independent recommendations.

There are privacy limitations as well. Businesses cannot access transcripts of individual conversations customers have with their agents. That protects customer privacy but also limits businesses' ability to see which questions customers struggle with, where they need more information or why they might abandon a purchase. OpenAI provides aggregated insights rather than individual conversations, leaving businesses with only a partial understanding of those interactions.

Business Agents give companies a way to continue a conversation after someone discovers them, but they don't determine whether ChatGPT recommends the business in the first place.

A retailer might update its ChatGPT profile with detailed product descriptions, current prices and availability, hoping more customers will discover its products. That information can help ChatGPT better understand what the company sells, but claiming a profile or providing more accurate information does not guarantee more recommendations.

OpenAI's vice president of monetization, Benji Shomair, told Digiday that claiming a business should not affect how frequently ChatGPT recommends it. However, more accurate information could help the assistant recognize when a company's offerings are relevant to a customer's question. Businesses can therefore improve the information ChatGPT has about them without knowing whether they'll appear ahead of competitors—or why one company is recommended over another.

OpenAI says its search systems consider factors such as user intent, relevance and recency. For shopping recommendations, the company also identifies product details, pricing, availability and reviews as information that can influence which products appear. But those factors don't explain every recommendation, and OpenAI has not published a comprehensive ranking formula that businesses can use to determine how prominently they'll appear.

Consider two service businesses competing for the same customers. One has operated for 20 years, accumulated hundreds of positive reviews and earned mentions across industry websites. The other is relatively new, has fewer reviews and receives little attention outside its own website. Would ChatGPT favor the more established company? Would the answer change depending on what the customer asks?

Those are reasonable questions, but OpenAI has not disclosed how business longevity, independent mentions or traditional SEO signals such as backlinks are weighted in its recommendations. ChatGPT can draw on outside sources, including information from search providers, but having more online references does not establish that a company will be recommended more frequently.

The uncertainty reflects how AI-generated answers work. ChatGPT uses language models to interpret questions and generate responses, sometimes incorporating retrieved information. Recommendations can therefore depend on the question, its context and the available sources. Although OpenAI has disclosed some factors behind ChatGPT's search and shopping results, businesses still cannot determine exactly why they were included in or excluded from a particular recommendation.

For consumers, a verified business profile or sponsored placement should not be mistaken for ChatGPT's independent endorsement. For marketers, the limitation is equally important: greater control over business information doesn't translate into control over AI recommendations. Even when ChatGPT recommends a company, another challenge remains: determining whether that visibility leads to website visits, customers and revenue.

A business might see its ChatGPT mentions and website clicks increase, but that doesn't necessarily mean more customers are buying its products or requesting its services. Visibility, website traffic and revenue are three different measurements, and OpenAI's new dashboard doesn't connect all three.

Consider a service business that appears more frequently in ChatGPT recommendations. Its Insights dashboard can show qualifying mentions and website clicks, but it cannot tell the company whether those visitors requested quotes, booked appointments or became paying customers. To evaluate those outcomes, the business would need to examine its own website analytics, customer inquiries and sales records.

That creates a challenge for marketers investing in AI visibility. More mentions might indicate greater exposure, while additional website clicks suggest people are taking an interest. But neither metric independently establishes whether the investment is generating enough revenue to justify its cost.

OpenAI's first-party analytics provide a useful starting point for evaluating AI discovery, particularly when combined with a company's existing marketing and sales data. Whether the platform will eventually offer more detailed reporting that helps businesses connect AI interactions to customer conversions remains unanswered. OpenAI has not announced those capabilities as part of its current roadmap.

OpenAI's Business Hub introduces a new source of evidence into a market where businesses have largely relied on third-party testing to understand their visibility in ChatGPT. Companies can now evaluate their AI visibility efforts using both OpenAI's measurements of actual platform activity and third-party estimates based on selected prompts. That creates an opportunity for more informed marketing decisions, while also putting pressure on analytics providers to demonstrate the value of their services.

For businesses and marketers, the immediate decision is how much to invest in each approach. OpenAI's data can help establish whether a company is appearing in qualifying ChatGPT interactions, while third-party tools can investigate specific questions, compare competitors and evaluate visibility across multiple AI platforms. Neither provides a complete picture on its own, and their results cannot necessarily be compared directly. Businesses will need to determine which measurements help them make decisions rather than simply accumulating more visibility reports.

The larger limitation is that measuring AI visibility is not the same as understanding why a business gets recommended or whether those recommendations generate revenue. OpenAI's dashboard does not resolve either question. Companies must still evaluate customer inquiries, website activity and sales outcomes to determine whether their AI visibility investments are producing meaningful results.

As AI assistants take on a larger role in consumer discovery, businesses that can connect AI visibility with measurable customer outcomes may gain a competitive advantage. OpenAI has made part of that discovery process more visible, but companies must still determine whether appearing in ChatGPT recommendations is translating into meaningful business growth.

Q: What is OpenAI's Business Hub, and is it free?
A: Business Hub is OpenAI's platform for eligible companies to claim and manage their ChatGPT business profiles, review aggregated visibility analytics and, where available, create Business Agents. Businesses do not need a paid ChatGPT Business or Enterprise subscription or an advertising account to use the free profile-management features.

Q: How do I claim and verify my business on ChatGPT?
A: Search for your company in Business Hub and select its listing. Depending on the business, OpenAI may offer verification by phone, business email, DNS record or connected Shopify account. After claiming the listing, an authorized representative must complete identity verification through Persona to publish profile changes and access analytics. Availability varies by business type and location.

Q: Can I see how often ChatGPT recommends my business?
A: You can see qualifying mentions in Business Hub's Insights dashboard, but these are not a complete count of every time ChatGPT names or recommends your company. Online business mentions are counted when ChatGPT references the company's website, while local business mentions are counted when its location appears in a result. The dashboard also reports website clicks and discovery topics over the previous 28 days, subject to reporting thresholds.

Q: Can I see the exact questions people ask ChatGPT about my business?
A: No. OpenAI provides aggregated discovery topics rather than individual customer prompts or conversations. Third-party AI visibility services can show the questions they test and which businesses appear in those responses, but those tests do not reveal the actual questions customers ask ChatGPT.

Q: Does verifying my business help it rank higher in ChatGPT?
A: OpenAI says claiming a business does not directly increase how often ChatGPT recommends it. Accurate business information can help ChatGPT understand a company's offerings, but verification does not guarantee placement or an advantage over competitors.

Q: Why does ChatGPT recommend some businesses instead of others?
A: OpenAI identifies factors such as relevance, user intent and recency in its search systems, while shopping recommendations can also consider product details, pricing, availability and reviews. However, OpenAI has not disclosed a comprehensive formula explaining why one business appears instead of another or how factors such as business age, backlinks and independent mentions are weighted.

Q: Will OpenAI's analytics replace third-party AI visibility tools?
A: Not necessarily. OpenAI measures qualifying activity from actual ChatGPT interactions, while third-party services test selected prompts, compare competitors and track visibility across multiple AI platforms. The two approaches provide different information, and neither currently offers a complete picture of AI recommendations or their effect on business revenue.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing support, AEO/GEO/SEO optimization, image concept development, and editorial structuring support from ChatGPT, an AI assistant. All final editorial decisions, perspectives, and publishing choices were made by Alicia Shapiro.