Should Local Businesses Buy ChatGPT Ads Yet?

Abstract

This report reviews what the first nine months of ChatGPT advertising, from February through November 2026, show about cost, performance and fit for local service businesses. ChatGPT ads went from a 200,000-dollar-minimum pilot in February to a self-serve platform open to any U.S. business by May, and more than 100,000 advertisers across 89 countries had run campaigns by late October. The early results are not encouraging for local businesses. The largest independent study, covering 1,898 online stores, found that a visit from a ChatGPT ad was worth about a third as much as a visit from a Google search ad, 0.89 dollars versus 2.45 dollars in revenue, and converted at 0.37 percent versus 1.17 percent. A local field test by an Atlanta-area EV charger installer found the same pattern: only 0.83 percent of ChatGPT ad visits included a tap on the phone number, versus 3.30 percent of Google Ads visits, a gap of roughly 4 to 1. Reported click prices range from 1.72 to 40 dollars, targeting is no finer than ZIP code, and ads are blocked near health conversations. For most local businesses, ChatGPT ads are a small test at best in late 2026, while earning unpaid recommendations in ChatGPT's answers remains the better investment. Paid ads and unpaid AI recommendations are separate: OpenAI says buying ads does not make ChatGPT recommend a business.

Methodology

This report was prepared by Make It Loud Digital Marketing and reviews publicly available information on ChatGPT advertising as of November 3, 2026: OpenAI's announcements and documentation, peer-reviewed and working-paper research, third-party advertiser tracking and individual advertiser reports, compared with published Google Ads benchmarks. It adds one local field test from Car Charger Specialists' Google Analytics 4 data, January 1 through November 2, 2026, in which a phone call tap is the website's tracked tap or click on the phone number, not a connected call or a booked job. Figures from external sources are reported as published by those sources and were checked against their original references in November 2026. This is a review of public evidence plus one client field test, not an original survey or a peer-reviewed study, and it is not legal, financial or purchasing advice.

Findings

ChatGPT opened advertising to any U.S. business in under three months, from a test launch on February 9, 2026, to a self-serve Ads Manager with no minimum spend on May 5, and more than 100,000 advertisers across 89 countries had run campaigns by October 21. Early adopters are mostly software and productivity brands, not local businesses: Similarweb counted 3,770 active advertisers in the week ending June 8, 2026, and eight of the ten largest were software, productivity or developer brands. The largest independent study, by Kaiser and Schulze, found a ChatGPT ad visit generated 0.89 dollars in U.S. revenue versus 2.45 dollars for a Google search ad visit, converting 0.37 percent of the time versus 1.17 percent. Reported ChatGPT click prices range from 1.72 to 40 dollars against a Google median of 8.33 dollars for home services. In a local field test, Car Charger Specialists saw 7 of 839 ChatGPT ad visits (0.83 percent) include a phone-number tap versus 75 of 2,271 Google Ads visits (3.30 percent), about 4 to 1. Platform rules limit local use: no ads near health or political conversations, location targeting no finer than ZIP code, topic matching rather than keywords, and ads shown only to logged-in adults on the Free and Go plans. Ads do not influence ChatGPT's unpaid answers.

Overview

What the first nine months of ChatGPT advertising, from February through November 2026, show about cost, performance and fit for local service businesses.

On February 9, 2026, OpenAI began testing ads in ChatGPT, which now has more than 1.2 billion weekly users. By May, any U.S. business could buy ads through a self-serve Ads Manager with no minimum spend, and agencies have been fielding the same question from clients ever since: should we be advertising on ChatGPT?

This report pulls together what is publicly known as of November 2026: how the program works, who is buying, what independent researchers have measured and what advertisers report paying, plus a local field test from one of Make It Loud’s North Georgia clients. Most published results come from online retailers and software companies, so we focus on what the evidence means for local service businesses such as contractors, home services, medical practices and law firms.

The short answer is that ChatGPT ads opened to everyone remarkably fast, but the early evidence on results is not encouraging. The largest independent study found that visitors from ChatGPT ads were worth about a third as much as visitors from Google search ads, and a local field test found the same roughly 4 to 1 gap on phone calls. For most local businesses, ChatGPT ads are a small test at best in late 2026, while earning unpaid recommendations in ChatGPT’s answers remains the better investment.

Download the full report

You can download the full research report as a PDF. It is a review of public evidence plus one client field test, prepared by Make It Loud Digital Marketing and dated November 2026, and it reproduces the tables, methodology, limitations and sources below.

Key findings at a glance

Sources are listed in full at the end of this report.

Finding 1: ChatGPT Ads Opened to Everyone in Under Three Months

OpenAI launched ChatGPT ads as a U.S. test on February 9, 2026. Early pilot advertisers reportedly committed around 200,000 dollars; the minimum fell to 50,000 dollars in April and disappeared on May 5, when OpenAI opened a self-serve Ads Manager with cost-per-click bidding. Daily budgets start at 25 dollars. Growth followed quickly: more than 100 million dollars in annualized ad revenue by late March, a 1 billion dollar run rate by the end of August, and more than 100,000 advertisers across 89 countries by October 21, according to OpenAI and news reports.

OpenAI has projected about 2.5 billion dollars in ad revenue for 2026 and roughly 100 billion dollars by 2030, according to Axios. Those targets explain the speed of the rollout. They also mean the platform, its rules and its prices are changing month to month.

Source: OpenAI announcements; CNBC; Axios; Digiday and eMarketer; PPC Land. Revenue figures are as reported by news organizations.

Finding 2: Software Companies, Not Local Businesses, Are Buying

Similarweb, which tracks ChatGPT ad impressions, counted 3,770 active advertisers in the week ending June 8, 2026, up 46 percent from 2,585 the week before. Eight of the ten largest were business software, productivity or developer brands, led by Monday.com with about 5 percent of all ad impressions. No major retailer made the top ten, and local service businesses did not appear among the leaders.

That mix matters. ChatGPT’s audience skews toward professionals, and its ads work by matching an ad to the topic of a conversation rather than to exact keywords. That suits a software product people ask ChatGPT about at work. It is a less natural fit for a plumber or roofer whose customers often need someone today, nearby.

Finding 3: The Largest Study Found ChatGPT Ad Visitors Worth About a Third as Much

In October 2026, Maximilian Kaiser of WU Vienna and Christian Schulze of the Frankfurt School of Finance and Management released the first large-scale study of ChatGPT ad performance. Using 60 weeks of data covering 1,898 online stores with 16 billion dollars in annual sales, including 387 that advertised on ChatGPT, they compared traffic from ChatGPT ads with other channels. In the U.S., a visit from a ChatGPT ad generated 0.89 dollars in revenue on average, versus 2.45 dollars for a visit from a paid Google search ad. ChatGPT ad visitors bought 0.37 percent of the time; Google search ad visitors, 1.17 percent.

Paid ChatGPT traffic was the only channel that fell short of direct traffic on all six measures the researchers tracked, from conversion rate to time on site. At the return advertisers get from paid search, they estimate a ChatGPT ad click is worth about 0.26 dollars; one of the largest known accounts paid 1.72 dollars. The paper is a working paper under review, and its data came from an analytics firm one author advises. The same researchers’ earlier peer-reviewed study in Marketing Science found that even unpaid visitors from ChatGPT converted less often than visitors from most other channels.

Source: Kaiser and Schulze, “Can Advertising Fund Free AI?” working paper, October 2026. U.S. traffic, June 2025 to August 2026.

Finding 4: Reported Costs Are All Over the Map

OpenAI has not published cost or performance benchmarks. It recommends starting bids of 3 to 5 dollars per click. The handful of advertisers who have shared numbers paint a scattered picture: one e-commerce account reported 1.72 dollars per click over 15 days; a Canadian agency test paid about 7 dollars per click with a 0.6 percent click-through rate and no purchases; and an agency testing legal leads reported 20 to 40 dollars per click and roughly 700 dollars per lead. These are single accounts, mostly self-reported, and not representative.

For comparison, WordStream’s 2026 benchmarks put the median Google search ad click at 8.33 dollars for home and home improvement businesses and 9.87 dollars for attorneys, with click-through rates around 6 percent. ChatGPT clicks are not obviously cheaper, and early click-through rates appear far lower.

ChatGPT figures are individual advertiser reports, not benchmarks. Google figures: WordStream 2026 Google Ads Benchmarks, 13,474 U.S. search campaigns, April 2025 to March 2026 (medians).

Finding 5: A Local Test: Google Ads Visitors Called 4 Times as Often

Published data from local service businesses is almost nonexistent, so we looked at one of our own clients. Car Charger Specialists, an EV charger installer serving Metro Atlanta and North Georgia, ran ChatGPT ads from July through early November 2026 and is publishing its results. The campaign drove 839 website visits. Those visitors engaged with the site about as often as Google Ads visitors (28 percent versus 29 percent engaged visits), but only 7 of 839 visits (0.83 percent) included a tap on the phone number, compared with 75 of 2,271 Google Ads visits (3.30 percent).

That gap of roughly 4 to 1 is statistically significant and lines up with the one-third conversion rate the Kaiser and Schulze study found for online stores. At that rate, a ChatGPT click would need to cost about a quarter of a Google Ads click to produce a phone lead at the same price, roughly 2.10 dollars against the 8.33 dollar home services median.

Source: Car Charger Specialists, Google Analytics 4, January to November 2, 2026. 839 ChatGPT ad visits; 2,271 Google Ads visits. A phone call tap is a tracked tap or click on the phone number, not a connected call, a qualified lead or a booked job.

Finding 6: Rules That Limit Local Businesses

Several platform rules matter more for local businesses than for national brands. Ads are not shown near conversations about health, mental health or politics, which limits medical, dental and behavioral health practices. In the U.S., advertisers can target by state, metro area or ZIP code, but not by radius around a business. Ads are placed by matching the topic of a conversation rather than exact keywords, which gives less control over the searches an ad appears for. And ads reach only logged-in adults on ChatGPT’s Free and Go plans; paying subscribers on Plus, Pro, Business, Enterprise and Education plans never see them.

Most important: OpenAI says ads do not influence ChatGPT’s answers and are shown separately from them. Buying ads does not make ChatGPT recommend a business. As Make It Loud’s research on whether AI assistants recommend the same local businesses as Google Maps found, ChatGPT often names different local businesses than Google does, and earning those recommendations depends on reviews, listings and website content, not ad spend. Paid ChatGPT ads and unpaid AI recommendations are two separate things, and spending on ads does not improve the unpaid answers.

What this means for local businesses

Here is how we are advising North Atlanta clients who ask whether to advertise on ChatGPT.

First, do not move Google budget to ChatGPT yet. Independent research and our local test both found ChatGPT ad visitors converting at roughly a third to a quarter of the rate of Google search ad visitors, so keep proven channels such as Google Ads funded while the new platform matures.

Second, if you test, test small and measure leads. A capped test budget, ZIP code targeting limited to your service area, and tracking of calls, forms and booked jobs by source will tell you more than clicks or engagement.

Third, healthcare practices should expect limited reach. Ads are not shown near health and mental health conversations, so medical, dental and therapy practices may find few places for their ads to run.

Fourth, invest in being recommended for free. ChatGPT’s unpaid answers name local businesses every day, and strong reviews across platforms, accurate listings and clear service pages help a business earn those mentions, which is the heart of AI search optimization and also helps in Google. If you are weighing where ChatGPT ads fit against the rest of your marketing, our AI consulting and digital marketing teams can help you decide, and you can read more in our original research library.

Limitations

The Kaiser and Schulze ad study is a working paper under review; its data came from an analytics firm that one author advises, and it covers online stores, not service businesses. Its estimate of lost organic visitors (about 6 percent) was not statistically significant, and its per-user revenue figures are the authors’ calculations from OpenAI’s figures. Advertiser cost reports are single accounts, often self-reported by agencies, and are not representative. Revenue milestones are as reported by news organizations, and some OpenAI pages could not be accessed directly. The Car Charger Specialists test is one business, its ChatGPT and Google campaigns ran at partly different times, and it measures call taps rather than booked jobs. ChatGPT’s ad platform is changing quickly, so confirm current details before relying on them. This report offers general information only. It is not legal, financial or purchasing advice.

Sources

Kaiser, M., and Schulze, C. “Can Advertising Fund Free AI? Evidence from ChatGPT’s International Advertising Rollout.” Working paper, Version 1, October 2026. ecommerceresearch.org/chatgpt-ads

Kaiser, M., and Schulze, C. “ChatGPT Referrals to E-Commerce Websites: How Do LLMs Compare Against Traditional Channels?” Marketing Science 45(4): 699 to 715, 2026. doi.org/10.1287/mksc.2025.0489

OpenAI. “Testing ads in ChatGPT” (Feb. 9, 2026, with updates); “New ways to buy ChatGPT ads” (May 5, 2026); ChatGPT ads expansion announcement (Oct. 21, 2026). openai.com

Similarweb. “ChatGPT Ad Stats,” June 15, 2026. similarweb.com

CNBC; Axios; Digiday; eMarketer. Reporting on ChatGPT ad revenue milestones, projections and minimum spend, March to August 2026.

PPC Land; Search Engine Roundtable. Reporting on ChatGPT Ads Manager features, location targeting and bidding options, May to October 2026. ppc.land; seroundtable.com

Opascope. “ChatGPT Ads Benchmarks,” June 15, 2026. opascope.com

MediaPost; Tech Times. Reporting on Choice OMG’s ChatGPT ad test, August to September 2026.

WEBRIS. “ChatGPT Ads Results,” October 5, 2026. webris.org

WordStream. 2026 Google Ads Benchmarks. wordstream.com

Car Charger Specialists. “We Tried ChatGPT Ads. Here’s How They Compared With Google,” November 2026; Google Analytics 4 data, January to November 2026.

Make It Loud. “Do AI Assistants Recommend the Same Local Businesses as Google Maps?” November 2026.