Advertising in AI: ChatGPT, Google AI Overviews, Perplexity & What Marketers Need to Know in 2026
Introduction
The advertising landscape just shifted dramatically. Within the past few months, every major AI platform has either launched ads or announced imminent plans to monetize through advertising. For performance marketers, this represents both a massive opportunity and a tracking nightmare.
On January 16, 2026, OpenAI announced it would begin testing ads in ChatGPT. Google has been expanding ads in AI Overviews globally. Perplexity launched sponsored questions in late 2024. Microsoft Copilot already shows ads within conversational search. The era of ad-free AI assistance is officially over.
But here’s what most marketers are missing: these platforms are launching with premium pricing, limited measurement, and CPM-based billing that fundamentally changes how you’ll need to think about campaign optimization. If you’re used to the robust attribution systems in Google Ads or Meta, you’re in for a rude awakening.
This guide breaks down exactly what’s happening across each platform, what it means for your paid media strategy, and how to prepare your tracking infrastructure before you waste budget on unmeasurable impressions.
Quick Decision Framework: Should You Advertise on AI Platforms?
Before diving deep, here’s a quick framework for whether AI advertising makes sense for your business right now:
| Your Situation | Recommended Action | Platform Priority | Budget Allocation |
|---|---|---|---|
| Enterprise brand with awareness goals | Test now (early mover advantage) | Google AI Overviews → ChatGPT | 5-10% of paid search budget |
| Performance marketer needing ROAS | Wait for better measurement | Google AI Overviews only | Monitor, don't invest yet |
| E-commerce with strong margins | Test Shopping in AI Overviews | Google AI Overviews → Copilot | Small test budget ($5K-10K) |
| SMB with limited budget | Focus on existing channels | None (no self-serve yet) | $0 until self-serve launches |
Think of it this way: AI Studio is for play, Vertex AI is for power, and Gemini is the brain that fuels both
ChatGPT Ads: Premium Pricing, Limited Data
OpenAI’s entrance into advertising marks a pivotal moment. With 800 million weekly users, ChatGPT represents one of the largest untapped advertising audiences. But the rollout comes with significant caveats for performance marketers.
What We Know
- Launch Date: Testing began late January 2026 for logged-in adult users in the U.S.
- Who Sees Ads: Free tier and ChatGPT Go ($8/month) users only
- Ad-Free Tiers: Plus ($20/month), Pro ($200/month), Business, Enterprise, and Edu
- Ad Placement: Below ChatGPT’s response, clearly labeled as “Sponsored”
- Pricing Model: CPM (cost per thousand impressions), NOT CPC
- Reported CPM: ~$60 per 1,000 impressions
The $60 CPM Problem
OpenAI is pricing ChatGPT ads at the premium end of the digital advertising spectrum. For context:
Platform | Typical CPM |
ChatGPT | ~$60 |
Premium CTV (Netflix, Hulu) | $40-65 |
Google Search Ads | ~$38 |
Meta (Facebook/Instagram) | $10-20 |
Google Display Network | ~$3 |
OpenAI justifies the premium pricing based on intent. When someone asks ChatGPT “what are the best running shoes for flat feet,” they’re expressing unfiltered purchase consideration before narrowing their choices. That’s the high-intent moment advertisers have always coveted.
But here’s the catch: the measurement is virtually nonexistent. At launch, advertisers receive only total impressions and total clicks. No granular conversion tracking. No demographic insights. No purchase attribution. No query-level data showing what prompt triggered the ad.
For performance marketers who’ve spent years optimizing based on detailed attribution data, this is a massive step backward. You’re essentially paying NFL-level rates for TV-style reporting.
OpenAI’s Ad Principles (Take With Grain of Salt)
OpenAI has outlined five principles governing ChatGPT ads:
- Answer Independence: Ads won’t influence ChatGPT’s responses
- Conversation Privacy: User conversations won’t be sold to advertisers
- Choice and Control: Users can turn off personalization and clear ad data
- Protected Categories: No ads near health, mental health, or political topics
- Age Restrictions: Users under 18 won’t see ads (AI estimates age from conversation patterns)
Whether these principles hold as revenue pressure increases remains to be seen. As one industry analyst noted, the question isn’t whether ads influence today’s responses it’s whether they’ll subtly shape recommendations months or years from now.
Google AI Overviews & AI Mode: The Familiar Platform With New Rules
Google’s approach to AI advertising is more mature but equally disruptive. Ads now appear within AI Overviews (the AI-generated summaries at the top of search results) and are being tested in AI Mode (Google’s conversational search experience).
Current Availability
- AI Overviews: Ads above/below available in 200+ markets; ads WITHIN AI Overviews available in 12 English-speaking countries (US, Canada, Australia, India, UK, and others)
- AI Mode: Currently testing ads in the U.S. only
- Eligible Campaigns: Search, Shopping, Performance Max, and App campaigns
- Requirements: Must use automated solutions like broad match or AI Max; ads must be relevant to both query AND AI-generated answer
Key Differences From Traditional Search
The fundamental change is that Google now considers both the user query and the content of the AI Overview when serving ads. This means your ad might appear for queries you’re not directly targeting if the AI-generated answer creates commercial intent.
For example, someone searching “why is my pool green and how do I clean it” might see ads for pool vacuum cleaners not because they searched for vacuums, but because the AI Overview mentioned vacuuming as a solution.
The Attribution Challenge
Here’s the critical issue: advertisers cannot currently see separate performance data for AI Overview placements. Your ads in AI Overviews are bundled into your existing Search campaign reporting. You can’t:
- Bid specifically for AI Overview placements
- Target conversational intent separately
- See granular performance metrics for AI Overview impressions
- Opt out of AI Overview placements
Google says dedicated controls, bidding strategies, and reporting for AI Overview and AI Mode placements will come “as these surfaces mature.” Until then, you’re flying partially blind.
AI Overviews Are Getting More Commercial
Data shows AI Overviews are expanding rapidly into transactional queries. According to Semrush, the percentage of keywords triggering AI Overviews has grown significantly:
- Commercial queries: 8.15% → 18.57% (in 12 months)
- Navigational queries: 0.84% → 10.33% (in 12 months)
For advertisers, this means AI Overviews are no longer just informational they’re becoming a core part of the commercial search journey. If you’re not prepared for this shift, you’re losing visibility on high-value queries.
Perplexity: Sponsored Questions (Currently Paused)
Perplexity was first to market with AI advertising in November 2024, launching “sponsored follow-up questions” that appear alongside AI-generated answers. However, as of October 2025, the platform has paused accepting new advertisers to “reassess its ambitions.”
How It Works (When Active)
- Format: Sponsored follow-up questions appear in the “Related Questions” section (accounts for 40% of platform queries)
- Answer Generation: Perplexity’s AI generates the answer to sponsored questions—advertisers don’t write the response
- Pricing: CPM-based (reported at $50+ CPM)
- User Base: ~22 million active users, 780 million monthly queries
- Demographics: 82% hold undergraduate degrees; 45% have graduate education
Why It’s On Hold
Early advertisers like Indeed and Whole Foods tested the platform, but feedback revealed significant challenges: limited scale, unclear ROI metrics, and inventory constraints. Jessica Chan, head of publisher partnerships, confirmed at Advertising Week NYC that ads “aren’t currently on the roadmap” for their new Comet browser.
The takeaway: Perplexity’s ad experiment shows the challenges of monetizing AI search at scale. Even with a differentiated approach (sponsored questions vs. traditional ads), the measurement and scale issues persist.
Microsoft Copilot: The Quiet Contender
Microsoft has been showing ads in Copilot (formerly Bing Chat) since 2023, making it the most established AI advertising platform. Ads appear below Copilot’s organic answers with “Microsoft Advertising” and “Sponsored” labels.
Key Features
- Conversational Targeting: Microsoft analyzes the entire conversation, not just the last query
- Performance Data: Copilot ads show 69% higher CTR and 76% higher conversion rates vs. traditional search (Microsoft’s data)
- Automatic Eligibility: Performance Max campaigns automatically qualify for Copilot placements
- Copilot Checkout: New feature enabling in-conversation purchases (partnerships with Etsy, Shopify, Walmart)
The Microsoft Advantage
Unlike ChatGPT and Perplexity, Microsoft already has 20+ years of advertising infrastructure. Your existing Microsoft Advertising campaigns can appear in Copilot without additional setup. This makes it the lowest-friction entry point for testing AI advertising.
The limitation: Bing’s market share (2.51% globally) means scale is limited compared to Google. But for B2B advertisers and reaching professional audiences through the Microsoft ecosystem, Copilot represents a serious opportunity.
Complete Platform Comparison
| Feature | ChatGPT | Google AI Overviews | Perplexity | MS Copilot |
|---|---|---|---|---|
| Status | Testing (Jan 2026) | Live globally | Paused | Live |
| Pricing Model | CPM (~$60) | CPC (auction) | CPM (~$50+) | CPC (auction) |
| Self-Serve | No | Yes | No | Yes |
| Conversion Tracking | None | Via Google Ads | None | Via MS Ads |
| AI-Specific Reporting | No | No (coming) | No | Limited |
| User Base | 800M weekly | 1.5B+ monthly | 22M active | 75M+ daily |
| Best For | Brand awareness | Performance + brand | N/A (paused) | B2B, professional |
What About Meta AI?
Meta isn’t launching ads inside its AI assistant (Meta AI), but it’s doing something potentially more impactful: using AI to fully automate ad creation and targeting by end of 2026.
Under Meta’s vision, advertisers will provide a product image and budget, and AI will generate the entire campaign creative, copy, targeting, and optimization without human intervention. The company’s Advantage+ campaigns already deliver 22-50% ROAS improvements and 12-30% CPA reductions.
This is a different kind of AI disruption: not ads inside AI, but AI running your ads. For Meta advertisers, the shift means less control but potentially better performance if you trust the algorithm.
The Tracking Problem: Why Your Attribution Will Break
Here’s the uncomfortable truth: the robust attribution systems that made digital advertising measurable don’t exist for most AI platforms yet.
ChatGPT? No pixel, no CAPI, no conversion tracking. You’re paying $60 CPM for impressions with zero visibility into downstream results.
Google AI Overviews? Bundled into your existing Search reporting with no way to isolate performance.
Perplexity? No confirmed integrations with Google Analytics, Adobe Analytics, or other measurement platforms.
Microsoft Copilot? The best of the bunch, leveraging existing Microsoft Advertising infrastructure, but still limited AI-specific insights.
For agencies and brands that have built sophisticated multi-touch attribution models, AI advertising represents a step back to the pre-digital era of “we know half our advertising works, we just don’t know which half.”
How to Prepare: Your AI Advertising Readiness Checklist
1. Audit Your Current Tracking Infrastructure
Before spending a dollar on AI advertising, ensure your foundation is solid:
- Server-side tracking implemented (GA4 Measurement Protocol, Meta CAPI)
- Enhanced conversions enabled in Google Ads
- First-party data collection strategy in place
- UTM parameters standardized across all campaigns
2. Set Up Baseline Metrics
Since AI platform reporting is limited, you’ll need to measure impact through:
- Brand search volume changes (monitor in Google Search Console)
- Direct traffic increases correlating with AI ad spend
- Post-purchase surveys asking “how did you hear about us?”
- Incrementality testing (geo-based holdout experiments)
3. Prioritize Google AI Overviews
If you’re going to test AI advertising, start with Google:
- Existing infrastructure means easier implementation
- CPC pricing (vs. CPM) aligns with performance goals
- Conversion tracking works (even if AI-specific reporting doesn’t)
- Enable AI Max for Search to qualify for AI Overview placements
4. Wait on ChatGPT (Unless You’re Enterprise)
ChatGPT advertising currently makes sense only for:
- Large brands with awareness-focused goals
- Companies with budget to test without ROI pressure
- Early movers seeking competitive positioning (“as seen on ChatGPT”)
Frequently Asked Questions
How to Prepare: Your AI Advertising Readiness Checklist
AI advertising is here, but it’s not ready for performance marketers who need measurable ROI. The platforms are launching with premium pricing (ChatGPT at $60 CPM), limited measurement (no conversion tracking), and CPM-based billing that shifts risk to advertisers.
For now, the smart play is:
- Prioritize Google AI Overviews — it’s the only AI ad surface with existing conversion infrastructure
- Fix your tracking foundation — server-side tracking and first-party data are prerequisites for any AI advertising success
- Test Microsoft Copilot — it’s the most mature platform with the lowest barrier to entry
- Watch ChatGPT — enormous potential, but wait for self-serve and better measurement
- Skip Perplexity — it’s paused accepting advertisers anyway
The brands that will win in AI advertising aren’t the ones rushing in first—they’re the ones building the measurement infrastructure to actually understand what’s working.
Need help building your AI-ready tracking infrastructure? At Hoerr Solutions, we fix tracking before we scale campaigns. That approach matters more than ever as advertising moves into AI platforms with limited native measurement. Let’s talk about getting your attribution foundation ready for what’s coming.
most guess, we engineer
20MM
Ad Spend Per Year
10+
Years of Experience
7x
Avg. ROAS
5
Industries Served
1
Martech Award
The Problem
47% of brands can’t accurately attribute revenue to ad spend. Broken tracking. Cookie loss. Siloed data. Most agencies optimize blind. Hoerr solutions mitigates these issues at the source before a dollar is spent.
How We Work
Complete analytics audit—GTM, client/server-side tagging, analytics, and consent management—before a single dollar is spent.
02 Then We Scale
Campaigns built on clean data. Every optimization backed by real attribution, traced back to the source of truth.
03 AI-Powered Insights
Proprietary tools we build for your business that surface what’s actually working and where you’re leaking.
04 No Black Boxes
Your ad account. Your data. Transparent reporting where you see exactly what we see and own every data point.
partners we've scaled
get in touch
Your Information
© All Rights Reserved.