On February 9, 2026, OpenAI quietly started showing ads to ChatGPT users in the US — Free tier and the new Go tier subscribers, with Plus, Pro, Business, Enterprise, and Education tiers remaining ad-free. The format was deliberately understated: clearly labeled "Sponsored" results below the model's organic answer, with technical safeguards that OpenAI claims prevent the ads from influencing the response itself.
By mid-May, the pilot was generating roughly $100 million in annualized revenue. That's a hundred-day path from launch to a nine-figure run rate inside the product most marketers have spent the last two years trying to win organically.
Around the same time, Perplexity did the opposite. After running a small ads program through 2025, Perplexity removed every ad from the platform — for free users, paid users, every tier. The company's stated rationale: long-term trust is worth more than short-term ad revenue, and the differentiation from OpenAI is the bet. Brands can still appear in Perplexity's organic citations through good content, but they cannot buy placement at any price.
Google, meanwhile, is testing sponsored placements inside AI Mode answers — explicit "Sponsored" markers above or below the generative response — with rollout signals pointing to general availability in late 2026 or early 2027.
Three engines. Three different monetization models. Three different things your landing page now has to do to surface in each one. The single-playbook era of AI search optimization — "do AEO, get cited everywhere" — is over. What's replacing it is a per-engine strategy that most marketing teams haven't built yet.
This piece is the framework I've been working through with founders in May. It's not pretty. The new model takes more thought than the old one. But the teams that build engine-specific strategies in mid-2026 will compound advantages that single-playbook competitors can't catch up to without rebuilding from scratch.
What Each Engine's Monetization Actually Looks Like
Before getting into strategy, it's worth being precise about how each engine is currently structured — because the differences shape everything downstream.
ChatGPT: organic answers + labeled sponsored results below. When a user asks a question, the model generates an organic response with its usual citation behavior. Below that response, in a clearly delineated section, one to three sponsored placements appear. These are auction-priced, similar to Google's Sponsored results, but with significant restrictions: sponsored content cannot influence the organic answer, advertisers cannot target users by sensitive characteristics, and the model is instructed to maintain the same response whether ads are present or not.
The performance is striking: $100M annualized revenue in 100 days, with the bulk of advertiser spend concentrated in the most commercially valuable verticals — software, ecommerce, financial services, professional services. Sponsored CPMs are reportedly running 5-10x what equivalent Google Search Ads command, because the audience intent is more concentrated and the placement is functionally below a complete answer the user already trusts.
Perplexity: organic citations only. The company stripped ads entirely. There is no path to paid placement. The only way your brand surfaces is through Perplexity's organic citation behavior, which preferentially weights Reddit, technical community sources, comparison content, and reputable independent reviews. For categories where Perplexity is your audience's primary AI search tool (technical buyers, researchers, comparison-stage shoppers), this is a binary outcome: be cited organically or be invisible.
Perplexity's trust positioning is the strategic asset. Without ads, the company's pitch is that its answers are not influenced by advertiser money — and for the subset of users who prioritize that property, Perplexity has consolidated as the default. Brands cannot buy out of this dynamic. They can only work within it.
Google AI Mode: organic + sponsored testing. AI Mode currently surfaces organic responses powered by Gemini 3.5 Flash, with the same citation behavior described in our earlier coverage of the I/O 2026 redesign. Sponsored placements are being tested in select queries and verticals, with industry signals suggesting general availability in late 2026 or early 2027. When that lands, AI Mode will look structurally similar to ChatGPT: organic answer above, sponsored placements below or beside.
The implication: AI Mode's organic surface is wider and more contested now than it will be in nine months. Brands that establish strong organic citation positions before paid bidding compresses the space will hold those positions when ads launch. Brands that wait until the paid market clarifies will be competing for organic citations against both peers and advertisers' organic teams who built the position earlier.
The Per-Engine Strategy
Here's the strategy framework I've been recommending. It maps each engine to a primary investment, a secondary investment, and a defensive move.
| Engine | Primary investment | Secondary investment | Defensive move |
|---|---|---|---|
| ChatGPT | Organic citations on category queries | Selective sponsored placements on high-intent terms | Paid presence on your branded queries (competitors will bid) |
| Perplexity | Reddit + technical community citations | Comparison content + independent reviews | N/A — no paid surface, organic is the only game |
| Google AI Mode | Organic position before paid bidding launches | Long-form structured content the engine extracts cleanly | Reserve paid budget for AI Mode ad launch (Q4 2026 / Q1 2027) |
Three implications worth pulling out.
Perplexity is now an organic-only category for your strategy. If your buyers use Perplexity, the only way to win is by getting cited in the sources Perplexity preferentially pulls from. That means Reddit (still 31% of Perplexity's citation share), technical comparison content, and reputable independent reviews. There is no shortcut, no paid bid, no boost. The teams winning Perplexity citations are the teams investing systematically in the upstream surfaces — and reaping the compounding benefit of a competitor that can never simply outspend them.
ChatGPT is now a hybrid surface. Organic citations still matter and still drive the bulk of conversion — but a brand can no longer afford to ignore the sponsored layer, particularly for branded queries where a competitor's ad can intercept the user. We're already seeing competitors bidding on each other's brand names in ChatGPT's ad inventory. If you're a leader in your category and not running defensive paid presence on your own brand terms, you're letting cheap acquisition slip to whoever is.
Google AI Mode is in its pre-paid window. Right now, organic citations on Google AI Mode are uncontested by paid bidding. That ends when AI Mode ads ship to general availability. Every month between now and then is a month of accumulating organic position that competitors will have to displace later, against ad budgets, with content investments that take quarters to compound. This is the largest unclaimed surface in AI search right now.
Brand-Term Defense Is the First Move
If you only do one thing in response to ChatGPT's ad launch, it's this: defend your branded queries with paid presence inside ChatGPT.
Search your own brand name in ChatGPT. See what comes up in the Sponsored section below the answer. For a meaningful percentage of B2B brands we've audited, the sponsored slot is being bought by a direct competitor — sometimes a small competitor most users have never heard of, who's discovered that ChatGPT brand-term inventory is cheap because the brand owner isn't bidding on it.
The mechanics are the same as Google Search Ads circa 2010. Brand-term bidding by competitors works because the user's intent is your brand specifically, but the ad placement offers an alternative. Conversion rates are unusually high because the alternative is encountered at the exact moment of evaluation. The defensive bid by the brand owner is cheap (you have quality-score advantages on your own name) and converts at top rate because the user wanted you in the first place.
ChatGPT's ad inventory is exhibiting the same dynamics, with one important twist: the sponsored placement appears after the organic answer that already named your brand. The user has been told what your product is, may have read a brief description, and is now being offered a competitor's clickable alternative. Defending your own name here is the highest-ROI ad spend in AI search right now. If you have any paid AEO budget at all, this is where it goes first.
Where Organic Investment Pays Off Most
The trickier question is where your organic investment goes — because the citation behavior across engines is fragmenting, and the right allocation depends on which engine your buyers actually use.
The general pattern, based on our tracking across roast.page customers:
B2B SaaS buyers skew heavily Perplexity and ChatGPT. Technical evaluators, product managers, engineering leads, and procurement teams over-index on Perplexity for comparison-stage research and ChatGPT for general category exploration. Google AI Mode shows up later in the funnel, often for branded queries or final-stage validation. If your audience is here, your organic investment should be Perplexity-heavy (Reddit, technical community sources, comparison content) supplemented by ChatGPT-friendly long-form structured content.
Consumer ecommerce buyers skew Google AI Mode and ChatGPT. Google retains the bulk of consumer query traffic, and AI Mode is rapidly becoming the default search interface for those users. ChatGPT is increasingly used for product discovery — particularly in higher-consideration categories (furniture, electronics, fashion). Perplexity has lower consumer share. Allocation here weights Google AI Mode organic citations heavily, ChatGPT secondarily, Perplexity as a marginal investment.
Local and service businesses skew Google AI Mode. Local intent queries — restaurants, services, "near me" — are dominated by Google. AI Mode now generates conversational responses with embedded local results. ChatGPT and Perplexity are not yet meaningful local-search surfaces. Allocate organic investment toward Google AI Mode and traditional local SEO; the AI-search investment in ChatGPT and Perplexity has lower ROI for this category.
Creative and developer tools skew ChatGPT and Perplexity. Audiences for design tools, dev tools, technical productivity software use ChatGPT and Perplexity at unusually high rates. Google AI Mode is secondary. Allocate aggressively toward ChatGPT (Reddit-adjacent technical citations, YouTube transcripts, structured comparison content) and Perplexity (technical communities, independent reviews).
What "Selective Paid Placement" Actually Looks Like
Beyond brand defense, the question of whether to invest in ChatGPT ads for non-branded category queries is harder. Sponsored CPMs are high. Conversion rates from sponsored placements are above Google Search average but below organic citation conversion rates. The unit economics work for some categories and not others.
The rough heuristic that's working as of May 2026:
Worth paying for: high-intent, late-funnel category queries where the organic answer already names competitors and the sponsored slot offers a clear alternative pitch. ("Best CRM for small business" — your competitors are in the organic answer, your sponsored placement says "the one that doesn't require a 30-day setup.") Categories with high LTV and short consideration windows.
Not worth paying for: exploratory, top-of-funnel category queries where users are still defining the problem. ("What's a CRM" — the user isn't ready to convert.) Categories with low LTV. Categories where your organic citation position is already strong.
The math: a typical ChatGPT sponsored placement converts at roughly 1.5-3% to landing-page click, with CPMs in the $40-100 range for software categories. That's an effective CPC of roughly $2-6 — competitive with Google Search Ads but with meaningfully more concentrated intent. If your blended CAC math works at those numbers, the test is worth running on a small budget. If your LTV doesn't support it, stay organic.
What to Stop Doing
Three patterns from the 2024-25 era that no longer fit the bifurcated 2026 landscape:
Stop running a single AEO playbook across all engines. "We do AEO" as a single line item in your marketing plan is now too coarse. The right framing is "we do ChatGPT AEO, Perplexity AEO, and Google AI Mode AEO, with different investment ratios and different content surfaces for each." A unified strategy under-invests where the leverage is highest and over-invests where it's lowest.
Stop ignoring competitor ads on your branded queries. If you haven't audited your branded queries in ChatGPT recently, do it this week. The sponsored placements below your organic answer are a leak. Defending them is cheap. The conversion you protect compounds across every quarter the leak stays unfixed.
Stop assuming Google AI Mode will be open organic territory forever. The unclaimed organic territory in Google AI Mode is a 9-12 month window before paid bidding launches. Treat it the way you would have treated Google Search organic in 2003 — invest aggressively in position now, because the cost of acquiring the same position post-paid-launch is meaningfully higher and competitive density rises sharply when bidding opens.
What This Means for Your Pricing and Comparison Pages
One of the under-discussed consequences of the AI ad-market split is the increased pressure on your pricing and comparison pages. Here's why.
When a user asks ChatGPT "what's [your product]'s pricing?" — a query that's now incredibly common — the organic answer surfaces your pricing page. The sponsored placement below frequently shows a competitor's "cheaper alternative" pitch. The user is now evaluating your pricing in direct comparison to a competitor's claim, in the same screen, in the same moment.
Your pricing page is now an AI-cited entity that needs to defend itself in direct competitive context. Vague pricing pages ("Contact us") leak the engagement to whoever the sponsored alternative is. Specific pricing pages with clear plan structure, dated pricing, comparison rationale, and value justification retain the user.
Same for comparison pages. The "X vs Y" pages on your site — which we've written about extensively as the highest-leverage organic AEO investment — now matter doubly because they're competing not just for organic citations but against sponsored placements that explicitly position alternatives. The teams treating comparison pages as a category-strategic surface (not a one-off content asset) are winning the highest-intent traffic in the AI search ecosystem.
What's Coming Next
A few predictions for the rest of 2026 with reasonable confidence:
Google AI Mode ads ship to general availability before end of 2026. The testing has been visible since Q1. The revenue pressure on Google to monetize the surface is high. Expect general availability between October 2026 and January 2027, with format closer to ChatGPT's labeled sponsored placements than to legacy Search Ads' integration.
ChatGPT sponsored CPMs will rise. The current $40-100 software CPMs reflect a thin advertiser base. As marketing teams figure out the placement, demand will compress supply. Expect 2-3x CPM growth through 2026 in commercially valuable categories. The arbitrage window for being early is real but short.
Perplexity's no-ads positioning will face revenue pressure. Pure organic is a great differentiator and a difficult business model. Perplexity will either (a) hold the line and accept slower revenue growth, (b) introduce some form of paid-but-non-intrusive surface like sponsored deep dives, or (c) lean further into subscription as the monetization model. The current pure-organic posture may not hold for 24 months.
Anthropic's Claude will join the conversation at some point. Claude is currently ad-free across all surfaces. As Anthropic's revenue mix matures (post-$30B raise at $900B+ valuation), expect monetization choices to surface. The current organic-only behavior on Claude is a window worth using.
The Three-Quarter Plan
If you want to operationalize the bifurcated landscape, here's the rough quarterly plan we've been building with marketing leads:
Q3 2026. Defend branded queries with paid placement in ChatGPT. Audit Perplexity citation rate and start the upstream investment (Reddit presence, technical community contributions, comparison content). Accelerate Google AI Mode organic position-building — every month is a month before paid bidding launches.
Q4 2026. Layer selective ChatGPT sponsored placements on highest-intent category queries where unit economics support it. Continue Perplexity organic compounding. Prepare for Google AI Mode ad launch — define what your sponsored placement strategy looks like once that surface opens.
Q1 2027. Execute on Google AI Mode paid surface from a position of organic strength. Re-evaluate Perplexity strategy based on whether the company has shifted monetization. Begin testing emergent AI search surfaces (Claude paid surface if it appears, agentic shopping placements, multimodal sponsored).
The teams that build this plan deliberately, with engine-specific investment ratios and a clear understanding of which surfaces support paid vs. require organic, will outperform single-playbook competitors by a wide margin through 2026. The unit economics will reward the precision.
If you want to baseline where your brand currently surfaces across ChatGPT, Perplexity, and Google AI Mode — both organically and against competitor sponsored placements — our free landing page analysis includes a cross-engine visibility audit that flags exactly which engines you're winning, which you're invisible in, and which competitors are intercepting your branded queries with paid placement. The picture is rarely uniform. The work to fix the uneven spots is meaningfully different per engine. And the teams that start now, while the bifurcation is still cleanly visible, will compound advantages the single-playbook teams won't see coming.