Quick answer: what changed in ChatGPT Ads?

According to Search Engine Land, ChatGPT Ads added four practical upgrades:

  • Automated bidding: a new "Maximize results" bid strategy that automatically sets and adjusts bids based on your campaign goal.
  • Platform targeting: an Eligible platforms setting to choose where ads can run: iOS app, Android app, or Web.
  • View-through conversions: reporting for one-day view-through conversions at campaign, ad group, and ad level.
  • WorkMagic integration: connect ChatGPT Ads performance with other channels and send conversion signals back via the Conversions API.

They are also expanding availability into Brazil and Mexico as part of a wider Latin America push.

1. Automated bidding: less manual work, more responsibility upstream

"Maximize results" shifts bid decisions from you (or your media buyer) to the platform. That is great when you want speed, scale, and fewer knobs to turn. It is also risky if the platform's definition of "results" does not match your definition of "profit."

What this means for your ad performance

Automated bidding usually works best when:

  • Your conversion goal is clean and consistent (lead, purchase, booked call).
  • You have enough conversion volume to guide optimization.
  • Your funnel does not have massive delays or offline steps that never get reported back.

If those are not true, Maximize results can optimize toward what is easiest to generate, not what is best for the business. In lead gen, that often means more low-intent form fills because the algorithm cannot see what your sales team sees.

What you should do next

  • Audit your conversion goal before you touch bidding. If your "conversion" is a thank-you page that triggers for spam, duplicates, or unqualified inquiries, you are about to automate your way into garbage.
  • Start with a controlled test. Run Maximize results against your current approach with a fixed budget and a single, agreed success metric — not just "more conversions."
  • Build guardrails outside the platform. Lead validation, CRM stages, and offline feedback loops (even if imperfect) are what keep automation honest.

If you work with an agency, note the shift in where value gets created: less time on bid micromanagement, more value in tracking architecture, landing page strategy, and lead-quality operations.

2. Platform targeting: your first real lever for intent and CPA control

ChatGPT Ads now lets you choose iOS, Android, and/or Web placements at campaign creation. That sounds minor, but it is a real control lever because behavior differs by surface.

App users often browse and convert differently than web users. Conversion rates can swing due to friction (mobile form fill vs. desktop), and downstream value can differ by device ecosystem, user demographics, or how quickly someone can complete the next step in your funnel.

What this means for your budget

If one platform produces cheaper conversions but lower quality, automated bidding may chase it aggressively. Platform targeting gives you the ability to stop the algorithm from hiding that reality in blended averages — or at least to separate performance so you can see it clearly.

What you should do next

  • Split campaigns by platform when you have meaningful differences in conversion rate, qualification rate, or close rate. Do not average iOS + Android + Web into one KPI if your sales team can feel the difference.
  • Match landing pages to the surface. Web traffic should land on fast, form-friendly pages. App traffic may need shorter steps, a different CTA (click-to-call, chat-first, book-now), or more aggressive friction reduction.

This is one of those "one setting in the UI" changes that can prevent a lot of quiet waste.

3. View-through conversions: helpful visibility, but do not let it fool you

ChatGPT Ads Manager now reports one-day view-through conversions. That means someone can see an ad, not click it, then convert within a day and ChatGPT Ads will count it as view-through credit — assuming a qualifying click did not already get attribution.

What this means for reporting (especially for agencies)

Your reported conversions can rise overnight without revenue moving. That does not automatically mean the platform is lying — it means attribution just got broader, and the platform is now giving itself more ways to claim influence.

View-through impact can be real. It can also be over-credited, especially when:

  • You run multiple channels (Google, Meta, email, affiliates).
  • You already have strong brand demand.
  • Your audience is already in-market and converting quickly.

What you should do next

  • Separate click-through vs. view-through in your dashboards. Treat them as different categories of performance, not one blended number.
  • Ask incrementality questions, not attribution questions. If view-through is the main story, test what happens when you reduce spend, change targeting, or shift budget to another platform.
  • Align on definitions with stakeholders. Your client or CFO needs to know whether "conversion" means clicked-and-converted or "saw an ad at some point and later converted."

Plan your month-over-month reporting now. This is exactly the type of change that creates confusing comparisons if you do not annotate it.

4. WorkMagic + Conversions API: the real unlock for serious advertisers

The WorkMagic integration allows advertisers to view ChatGPT Ads performance alongside other channels and send conversion signals back to OpenAI through the Conversions API.

What this means for optimization

Automated bidding is only as smart as the conversion signals it receives. If you only send "form submit," you will get more form submits. If you can send "qualified lead," "booked appointment," or "closed-won," you give the system a chance to optimize toward outcomes that matter.

This is where ChatGPT Ads starts to look less like a test channel and more like something you can scale — because you can train it on your real definition of success.

What you should do next

  • Prioritize server-side or API-based tracking where possible. It is typically more durable as browsers, consent requirements, and privacy rules evolve.
  • Define a north-star conversion and supporting quality events. For example: submit → qualified → booked → closed. Even sending one deeper event back (like "qualified") can improve optimization meaningfully.
  • Make sure your landing page and CRM agree on identity. If you cannot connect ad-driven leads to outcomes, automation will optimize in the dark.

5. Expansion into Latin America: plan for multilingual creative and localized landing pages

ChatGPT Ads is expanding into Brazil and Mexico. If you serve those markets, you may see earlier access or increasing inventory.

What this means for your site and funnel

If you send Portuguese or Spanish-speaking users to an English-only landing page, you will pay a "confusion tax" in the form of lower conversion rates and weaker lead quality. Even when users can translate, trust and clarity drop fast when the offer, proof, and form fields do not match expectations.

What you should do next

  • Build localized landing pages (not just translated headlines).
  • Confirm tracking and CRM routing by region and language so leads do not get mishandled.
  • Adjust offers and proof points to match local context: pricing cues, testimonials, compliance language, contact preferences.

The practical outcome: more automation, less wasted spend — if you set it up right

ChatGPT Ads is moving toward the same reality we have seen in Google and Meta: automation increases speed, but it punishes messy measurement. The winners will not be the advertisers who "turn on Maximize results" first. They will be the ones who treat bidding as the final step, not the first.

If you want this to pay off, focus on:

  • Clean conversion definitions
  • Platform-separated testing when behavior differs
  • Landing pages built for the device and audience
  • Reporting that keeps view-through in its proper place
  • Strong conversion signals via API integrations

Do that, and you get the outcome that actually matters: clearer performance, better lead quality, and less budget wasted on "results" that do not close.

Source: ChatGPT Ads adds automated bidding, platform targeting — Search Engine Land