OpenAI is retiring Max Bid and Max CPC in ChatGPT Ads on Jan. 1, 2027, replacing them with average cost targets that can bid above your number.
If you run ChatGPT Ads for a high-ticket Shopify store, the number you typed into the bid field stops being a ceiling. For readers of Ecommerce Paradise, this matters most when your store sees only a handful of conversions a month, because an average target needs data a store like that may not have.
OpenAI also changes how API partners send bids on Nov. 18, according to Search Engine Journal. Below you get the dates, how ChatGPT Ads got here, what an average target does to a store that closes on the phone, and five moves to make first. If the model itself is new to you, start with my guide to what high-ticket dropshipping is.
Better bids only pay off when the page after the click does the closing. Empire is a Shopify theme you can build those product and collection pages on. Look at the Empire theme for your store →
ChatGPT Ads Swaps Max Bid for Target CPA and Max CPC for Target CPC
OpenAI emailed advertisers about the change, and Search Engine Roundtable published the email’s contents on Nov. 5. The email said, “Starting Monday, November 9, you can optionally choose a cost target in Ads Manager.” New conversion campaigns get Target CPA in place of Max Bid. New click campaigns get Target CPC in place of Max CPC.
The difference is the word average. Per Relevant Audience, Target CPA is “the average amount you’d like to pay for a conversion, such as a purchase or lead.” OpenAI wrote that “Target CPC lets our system bid above or below your target in each auction.” Max CPC, by contrast, capped each bid.
The dates that matter
Nov. 9 is when new campaigns got the new options. Advertisers can also switch existing campaigns on their own, and existing campaigns keep their current bid setting through the end of December, per Relevant Audience.
Nov. 18 is the API date. Search Engine Journal reported on Nov. 10, citing an updated OpenAI help document, that a Max Bid or Max CPC value sent through the old API fields becomes an average target instead of a maximum. The legacy fields stay active until the end of December.
Jan. 1, 2027 is the cutoff. Max Bid and Max CPC retire, and OpenAI moves remaining campaigns to Target CPA or Target CPC, per Relevant Audience. Search Engine Journal adds that the legacy API fields are deprecated that day and that OpenAI told partners to update their integrations before then.
What OpenAI says about setting targets
OpenAI’s guidance is to base targets on recent actual CPA or CPC. For conversion campaigns, it also tells advertisers to confirm that the Pixel or Conversions API sends the selected event to the linked data source, and to watch Event Quality Score warnings for signal gaps. The warning is blunt: “Restrictive targets or weak conversion signals can reduce ad delivery and leave budget unspent.”
OpenAI’s worked example, as reported by Search Engine Journal, uses $400 in spend and 10 purchases. That is a $40 actual CPA, so a $20 target does not produce $20 purchases. The campaign may just underspend. The bidding goal and the billing event are separate settings, so a Target CPA campaign can optimize for conversions and still bill per click or per impression. OpenAI states, “The final winning auction bid sets the price you pay.”
Two quieter edits came with the change. Per Search Engine Journal, OpenAI’s advertiser basics page no longer describes a relevance-weighted, second-price auction, and it still suggests starting CPC campaigns with a max bid of $3 to $5 per click. The old Maximize results strategy is now called Intelligent bidding.
Gaps remain. Neither Relevant Audience nor Search Engine Journal found a stated target for campaigns still on Max Bid or Max CPC when they move on Jan. 1. OpenAI has not published a minimum conversion volume or a learning period, according to Relevant Audience. It has also not dated the new Target CPA and Target CPC API fields, per Search Engine Journal.
How ChatGPT Ads Went From a $60 CPM Pilot to Automated Bidding
The pilot launched Feb. 9 at a $60 CPM, the only pricing option at the time, according to PPC Land. Ads Manager opened to all U.S. businesses on May 5 with CPC bidding added and a recommended starting bid of $3 to $5 per click.
Automation crept in from there. PPC Land reported in August that Maximize results was preselected for eligible new ad groups. In the same report, OpenAI said the strategy does not guarantee delivery against a specific CPA, CPC or ROAS target “at this time,” and that a max bid was the only documented way to set a hard bid ceiling. Conversion-optimized campaigns, which need a pixel or Conversions API connection, began rolling out June 5.
I covered the earlier steps as they landed. Bulk product campaigns from a spreadsheet arrived first. The creative enhancements default came next, and I wrote up the warning in my opt-out post.
Microsoft made the same move on Oct. 1
Microsoft Advertising removed Max CPC from new non-portfolio campaigns on Oct. 1, per Search Engine Roundtable. The change covers new campaigns using Max Conversion, Max Conversion Value or Max Click. Portfolio strategies, target impression share and eCPC keep Max CPC, and so do existing campaigns that already use it.
Navah Hopkins, Microsoft’s Ads Liaison, wrote that “Max CPCs override stated goals,” which can cause spend pacing irregularities. I covered the change in my Microsoft Ads post.
The counterpoint
Hopkins’s argument carries over to OpenAI. A cap fights the goal you set, so removing it lets the system chase the goal. The argument on the other side is about control. PPC Land argued in August that a default volume-maximizing bid with no cost ceiling shifts control to the platform, especially for advertisers with no benchmark for what a ChatGPT conversion should cost. I found no OpenAI response to that point.
What Target CPA Means for a High-Ticket Store With Few Conversions
Everything from here is my read, not reported fact.
First, the ceiling. In August OpenAI’s help material called a max bid the only documented way to set a hard bid ceiling. If that wording still holds, Jan. 1 removes the last documented hard ceiling and leaves you with a target and a budget. The budget is softer than it sounds. PPC Land reported that OpenAI converted fixed daily budgets into seven-day averages in July with no opt-out. I would plan for variance, not for a clean cap.
Second, the math for a store that sells $2,000 products. Everything in this paragraph is hypothetical. Take a $3,000 monthly budget at a $3 CPC, the low end of OpenAI’s suggested starting bid. That buys about 1,000 clicks. A working paper I covered in my post on what paid ChatGPT clicks earn found a 0.37% conversion rate for paid ChatGPT visits against 1.17% for paid Google search. The paper is not peer reviewed. At 0.37%, 1,000 clicks produce about 3.7 purchases and a CPA near $811. At 1.17%, you get about 11.7 purchases and a CPA near $256.
Now put a $400 Target CPA on the 0.37% store. The system sees under four purchases a month and a typical cost double your target. OpenAI warns that restrictive targets or weak signals can leave budget unspent. My read is that the campaign slows or stops, and you conclude that ChatGPT does not work when the real problem is thin data.
Google’s help page recommends measuring Target CPA over at least 30 conversions in the last 30 days. OpenAI has published no such number and Google’s yardstick is not OpenAI’s, but 3.7 is a long way from 30. Meta’s finding that ad sets under 50 weekly conversions cost 28% more points the same way.
Third, the scenarios. These thresholds are my judgment, not OpenAI guidance. Under about 10 conversions a month, I would pick Target CPC over Target CPA, set it at or just above your actual recent CPC, and count a qualified lead as the conversion if you can track one. Between 10 and 30 a month, Target CPA works if you set it at your actual CPA and move in 10% steps. Above 30, tighten more freely, but only after you watch a week of delivery.
The lead event deserves a note. A buyer spending $2,000 often calls first. If the call is where the sale closes, a purchase-only signal starves the system. OpenAI’s own wording for Target CPA includes a “purchase or lead,” per Relevant Audience, so a lead event is a supported goal. The catch is quality. A cheap lead who never picks up is not a win, so tag every lead by campaign in a CRM such as HubSpot and compare close rates. My walkthrough on setting up HubSpot for a high-ticket store covers the setup.
A paid click only pays when a conversion system sits behind it, which is the argument in my post on high-ticket ad conversion systems. Bidding strategy cannot fix a product page with no phone number on it.
Timing matters too. Black Friday is Nov. 27 and Cyber Monday is Nov. 30, per PPC Land. I would not rebuild a working campaign that week. Existing campaigns keep their bid setting through December, so you can test Target CPC on one new campaign now and move the rest after Cyber Monday. API-managed campaigns are the exception. Nov. 18 changes how an API partner’s old Max Bid values behave, and that lands in the middle of the run-up.
ChatGPT is a test channel for most stores. Google Shopping is still the revenue driver I build high-ticket stores around, and I broke down how to turn Google Shopping clicks into sales. If you would rather have the store, the tracking and the ad accounts built by my team, that is the turnkey done-for-you build.
Cost targets reward stores that already convert. Want my team to scale the store you already have? See the scaling service →
ChatGPT Ads Bidding Checklist Before Nov. 18 and Jan. 1
Do these five in order:
- Open Ads Manager and list every campaign still on Max Bid or Max CPC. Export 30 days of spend, clicks and conversions for each, because every move below starts from your real numbers.
- Set targets from actuals. Divide spend by conversions for your CPA and spend by clicks for your CPC. Enter a target at or a little above that number, never below it, and tighten in 10% steps once delivery holds. That follows OpenAI’s advice to base targets on recent actual CPA or CPC.
- Check the signal. Confirm your Pixel or Conversions API sends the event you picked and read the Event Quality Score warnings. Give the landing page a business line on Quo so call leads reach a number you control.
- Ask whoever touches your account through the API what bid fields they send, and ask before Nov. 18. If you only use Ads Manager, write that down and move on.
- Cap the test. Pick a weekly amount you can afford to lose and log it as its own ad-spend category in QuickBooks. Then set a review date in the first week of December and build a follow-up flow for non-converting visitors in Klaviyo.
Still choosing what to sell before you spend on ads? Start with my free high-ticket niches list.
Frequently Asked Questions
When do Max Bid and Max CPC go away in ChatGPT Ads?
On Jan. 1, 2027, per Search Engine Roundtable’s report of OpenAI’s email. New campaigns got Target CPA and Target CPC on Nov. 9, and you can switch existing campaigns before the cutoff.
What changes on Nov. 18?
For API partners, a Max Bid or Max CPC value sent through the old fields becomes an average target instead of a maximum, according to Search Engine Journal. If an agency or tool manages your bids through the API, ask what it sends.
Is Target CPC a cap?
No. OpenAI says the system can bid above or below the target in each auction, and the final winning auction bid sets the price you pay. My post on measured ChatGPT click costs shows what clicks have run so far.
What target should I set?
Start at your recent actual CPA or CPC, per OpenAI’s guidance, and do not go below it. OpenAI’s own example shows a $20 target on a $40 actual CPA does not buy $20 purchases.
Does Google Ads still offer manual CPC?
Yes. I covered a new-customer-only option in my post on manual CPC in Google Ads.
Are ChatGPT Ads worth testing for a high-ticket store?
The evidence is mixed, so test small. The working paper found lower conversion than Google, and my post on multi-image ads and auto-rewrites covers the other settings to check first.
Where do I start if I do not have a store yet?
Take my free mini course and pick a niche before you buy a single click.
Want 1-on-1 coaching to decide which paid channels deserve your next dollar? Get the coaching details →
ChatGPT Ads will keep trading controls for automation. Your job is to feed it clean data and keep a number you can live with. Subscribe to the YouTube channel for daily breakdowns. More breaking news coming through the day.
Related Articles
If this was useful, these go deeper:
- OpenAI Tests Visual Ads Inside ChatGPT Image Generation
- Google Ads Management Process for High-Ticket Dropshipping: Daily and Weekly Checklist
- Black Friday Google Ads: How to Set Up Shopping and PMax for BFCM 2026
- Why High-Ticket Ecommerce Stores Need a Different Search Marketing Strategy
- Dropshipping Marketing Strategy in 2026: A Step-by-Step Playbook for High-Ticket Stores

Trevor Fenner is an ecommerce entrepreneur and world traveler from Seattle, Washington, and the founder of Ecommerce Paradise, a platform focused on helping entrepreneurs build and scale profitable high-ticket ecommerce and dropshipping businesses. With over a decade of hands-on experience, Trevor specializes in high-ticket dropshipping strategy, niche and product selection, supplier recruiting and onboarding, Google & Bing Shopping ads, ecommerce SEO, and systems-driven automation and scaling. Through Ecommerce Paradise, he provides free education via in-depth guides like How to Start High-Ticket Dropshipping, advanced training through the High-Ticket Dropshipping Masterclass, and fully done-for-you turnkey ecommerce services for entrepreneurs who want a faster, more hands-off path to growth. Trevor is known for emphasizing sustainable, real-world ecommerce models over hype-driven tactics, helping store owners build scalable, sellable, and location-independent brands.
He maintains homes in Seattle and Bali, spends most of the year overseas, and usually returns to Seattle for about a month each summer.

