OpenAI announced on August 31 that ChatGPT Ads hit a $1 billion annualized revenue run rate in under 200 days. That number is impressive on its face. It's also the kind of metric that deserves a CFO's scrutiny before anyone reallocates budget.
A run rate is a forward projection, not cash in the bank. Digiday's analysis estimates OpenAI has likely booked around $330 million in actual ad revenue through August, assuming a steady climb to the current $83 million monthly pace. The company told investors it expects $2.5 billion in ad revenue for 2026. To hit that target, the platform would need to average over $541 million per month through Q4, implying an exit run rate closer to $12 billion. That's a steep ramp, and it explains why OpenAI is pushing self-serve access into India, Europe, the Middle East, and North Africa simultaneously.
For marketing leaders evaluating this channel, the headline matters less than the operational reality underneath it.
The Measurement Gap Nobody Wants to Model
The most telling signal in the $1 billion announcement came days earlier. MediaPost reported that agencies were posting on LinkedIn about underwhelming performance from ChatGPT Ads campaigns. Peter Jaffray, managing director at Choice OMG, ran a $415 test across three campaigns in Canada. He got nearly 9,000 impressions against an expected 4,000, but his click-through rate landed at 0.6% compared to the 2% he'd expect from comparable search placements. Sixty clicks, roughly $7 per click, and zero conversions.
Jaffray's read: OpenAI's targeting hadn't matured enough to connect ads with buyers ready to act. That was June. The platform has since added conversion optimization, view-through conversions, and integrations with AppsFlyer and Adjust. But the core attribution challenge remains structural.
OpenAI's Ads Manager reports exactly seven metrics: impressions, clicks, spend, CTR, average CPC, average CPM, and a single rolled-up conversions number. No query-level data. No demographic breakdowns. No placement reporting. This isn't a beta limitation; it's a privacy-by-design boundary. Meanwhile, GA4 quietly buckets a meaningful share of ChatGPT ad clicks as Direct traffic because the in-app browser strips referrer data. One vendor's April sample showed 35.7% of AI-sourced traffic arriving without attribution.
If your CFO asks whether ChatGPT Ads worked, you need to answer that question before the test starts, not after.
What Advertisers Say They Need
The gap between OpenAI's revenue ambitions and advertiser satisfaction comes down to three capabilities that don't exist yet at the depth performance marketers require.
Granular targeting. OpenAI's current system matches ads by conversation context and, where personalization is enabled, signals from a user's broader ChatGPT experience. Advertisers can provide "context hints" describing relevant topics, but these aren't exact-match keywords and don't guarantee delivery. For B2B marketers accustomed to firmographic targeting or intent data layering, this is a significant step backward in precision.
Conversion path visibility. Conversational AI breaks traditional attribution models because the "destination" isn't always a webpage. A prospect might ask ChatGPT about project management software on Monday, revisit the conversation Wednesday to compare pricing, and convert Friday through a different channel. The attribution system sees one amorphous session. B2B cycles that typically require seven to thirteen touchpoints collapse into a single interaction that analytics tools can't parse.
Incrementality proof. With only seven native metrics and no holdout infrastructure, advertisers can't run the kind of geo-lift or matched-market tests that justify scaled spend. The channel is too young for media mix modeling. Until OpenAI provides incrementality tooling or third-party measurement partners fill the gap, budget decisions rest on faith rather than math.
The Pricing Evolution Worth Watching
EMARKETER documented the rapid democratization of ChatGPT Ads pricing. OpenAI launched in February at $60 CPM with a $200,000 minimum spend. By April, the minimum dropped to $50,000. Criteo pushed entry costs to $10,000. When self-serve opened to all U.S. businesses in May, the minimum spend requirement disappeared entirely. CPMs now clear as low as $25 in some categories, with CPC bidding available at $3 to $5 per click.
This trajectory tells you something about demand elasticity. OpenAI needed to lower barriers aggressively to hit volume targets. For advertisers, the falling floor creates an opportunity to test cheaply, but it also signals that the platform is still searching for product-market fit with performance buyers.
A Pilot Framework That Survives Board Review
If you're going to test ChatGPT Ads, structure the experiment so the results mean something regardless of outcome.
Start with a $5,000 to $15,000 budget over three weeks. Pick a single campaign objective, either clicks or conversions, and resist the temptation to optimize mid-flight. Document your assumptions: expected CTR, acceptable CPC, target cost per conversion. Set a kill threshold before you launch. If CTR drops below 0.3% or CPC exceeds $8 for your vertical, pause and reassess.
Solve the attribution problem upfront. Add static UTM parameters to every landing page URL. Deploy OpenAI's Conversions API with the oppref parameter to recover referrer data that GA4 would otherwise lose. Build a manual reconciliation process that compares Ads Manager conversions against your CRM within a 7-day and 28-day window.
Run a parallel holdout if your budget allows. Suppress ChatGPT Ads in one geographic region while running normally in a matched region. Compare conversion rates across both. This is the only way to approximate incrementality until OpenAI or a third party provides native tooling.
The Strategic Calculation
OpenAI is building a full-scale advertising platform around conversational AI. The $1 billion run rate, the 40-country expansion, the 50-plus technology and measurement partners: these are signals of commitment, not experiments. CFO Sarah Friar told employees the company will be public by 2027, and a diversified revenue model that includes advertising is part of the IPO narrative.
For B2B marketers, the question isn't whether ChatGPT Ads will matter. It's whether they matter now, for your specific pipeline, at your current CAC payback constraints. The honest answer: probably not yet, unless you're willing to treat this as a learning investment rather than a performance channel.
The advertisers who will spend more are the ones who get three things from OpenAI: query-level or topic-level reporting that lets them optimize creative and targeting, conversion path data that connects multi-turn conversations to downstream revenue, and incrementality measurement that proves lift against a holdout. Until those capabilities arrive, the $1 billion milestone is a story about OpenAI's trajectory, not yours.
Model the test. Document the assumptions. Set the kill criteria. That's how you turn a headline into a decision.
