|
Listen to this article
Getting your Trinity Audio player ready...
|
BLUF: AI ad spend is up 63% in 2026. Meta’s agents are running campaigns without us. Google is about to put ads inside Gemini. LinkedIn is the most-cited professional domain in AI. And 85% of pages ChatGPT reads never get cited. Here’s what’s changed – and the specific moves to make right now.
$57B
AI ad spend in 2026
63%
YoY growth in AI ads
15%
pages ChatGPT cites
AI marketing news round-up. Every story sourced from reports and platform announcements published 18–21 March 2026. Each update is structured the same way: what the problem is, what’s changed, and what you can do about it.
1. Why Is LinkedIn the Most-Cited Domain in AI Chatbot Responses [and how can marketers use that to build visibility]
➡️ The Problem: Most B2B LinkedIn Strategies Are Built for Impressions, Not AI Visibility
The dominant B2B LinkedIn strategy in 2025 was built around feed engagement :: post frequently, optimize for hooks, chase impressions + follower growth. That strategy produces vanity metrics but doesn’t produce AI citation visibility, which is the emerging channel that determines whether our brands appear when a potential buyer asks ChatGPT, Gemini, or Claude a question relevant to our expertise.
The gap between a social content strategy and an AI visibility strategy is not a minor tactical difference. It’s a different goal, a different content format, a different publishing frequency, and a different measure of success. Most marketing teams have not made that distinction yet.
✓ Why LinkedIn Has Become the Most-Cited Professional Domain in AI Responses
According to new research reported by MarketingProfs in March 2026, LinkedIn has become the most frequently cited professional domain in responses generated by AI chatbots including ChatGPT, Claude, and Google Gemini. LinkedIn citation frequency in AI responses has doubled in recent months, making it the top domain referenced when professionals ask AI systems questions about business strategy, marketing, leadership, and industry trends.
The mechanism is straightforward. AI language models are trained on and retrieve content from publicly accessible sources. LinkedIn posts, long-form articles, newsletters, and company pages are all publicly indexed + are high-signal sources of professional expertise. When an AI system is asked a business question, it preferentially cites sources that demonstrate topical depth, named authorship, and consistent publishing history. LinkedIn content that meets these criteria gets cited. Content that doesn’t, doesn’t.
This makes LinkedIn publishing a form of AEO infrastructure [Answer Engine Optimisation], not social media management. The distinction matters enormously for how we resource it, brief it, and measure it.
→ How to Build a LinkedIn AEO Strategy That Gets Your Brand Cited by AI Systems
The foundation is a thought leadership publishing system built around your brand’s most credible senior voices.
Target 4 posts per month per senior voice across a mix of formats: short-form declarative takes on industry developments, long-form analysis articles that go deeper than any competitor would bother to, and structured case studies with specific outcomes and named methodologies.
The goal is topical depth and consistency over a 90-day minimum – not follower volume or impression counts.
For AI citability, each piece of content should do 3 things: answer a specific question a professional in your target market would ask, include an author with visible credentials, and be part of a topic cluster that signals genuine expertise rather than scattered posting.
A CMO publishing 4 insightful articles on B2C demand generation over 90 days builds a citable body of work. 14 disconnected posts about company news do not. Prioritise depth. AI systems cite sources that know things – not sources that post things.
2. Why ChatGPT Doesn’t Cite 85% of the Pages It Reads [and how to make sure your content gets through]
➡️ The Problem: Most Content Strategies Are Optimized for Search Rankings, Not AI Citations – and Those Are Different Things
Traditional SEO success is measured by ranking position and organic traffic. Answer Engine Optimization (AEO) success is measured by whether our content is cited in AI-generated responses. These two outcomes require different content strategies, different structural choices, and different signals – and the majority of marketing teams are currently optimizing only for the first.
Research published in March 2026 quantifies the gap directly: ChatGPT retrieves content from many pages when formulating a response, but only 15% of those pages are ultimately cited in the final answer. The remaining 85% receive what might be called ghost traffic — the content is read by the AI system but never surfaced to the human user. Appearing in an AI retrieval set is not the same as being cited. Most brands cannot currently distinguish between the two, which means they cannot measure or improve their AI citation performance.
✓ What Makes Content Get Cited by ChatGPT, Gemini, and Other AI Systems
AI language models cite content that demonstrates four properties: specificity (it answers a precise question, not a vague topic), authority (it comes from a named source with visible expertise and is published on a domain with earned trust signals), originality (it contains data, analysis, or perspective that is not generic or widely replicated), and extractability (it is structured in a way that makes the answer easy to isolate – clear headers, concise paragraphs, direct declarative sentences).
Content that fails on these criteria, brand-led copy, generic industry overviews, thin blog posts without named authors or original insight, is retrieved but not cited. The practical implication: the same content investment that improves AI citation rates also tends to improve Google’s helpful content ranking signals, E-E-A-T assessment, and traditional SEO authority. Optimizing for AI citation and optimizing for search quality are now largely the same discipline.
→ How to Audit and Improve Your Content for AI Citation
Start with a citation test :: open ChatGPT, Perplexity or Gemini. Type the 5 questions your target audience commonly asks. Note which sources get cited.
If your brand doesn’t appear, your content lacks authority or the structure needed for the answer to be extracted. Both are fixable.
Check each question against 4 criteria.
- Does it have an author with visible credentials?
- Does it contain original data, proprietary framework, or a unique point of view?
- Do H2/H3 headers answer questions a user would ask?
- Does it link to credible sources?
Rewrite your 3 lowest-performing pages as direct expert answers to specific questions. Remove brand-centric framing. Lead every section with the answer, then support it.
This is the structural pattern AI cite – and it is also the pattern that earns Google’s highest helpful content signals in 2026.
3. Why AI Ad Platforms Are Outperforming Human-Managed Campaigns in 2026 [and what that means for your strategy]
➡️ The Problem: Marketers Are Still Optimizing the Wrong Thing
AI-powered ad platforms, including Google Performance Max, Meta Advantage+, and automated bidding systems, are now control targeting, budget allocation, and bid strategy with minimal human input. According to Madison & Wall’s 2026 US ad spend forecast, $57 billion will flow through AI-driven advertising platforms in 2026, representing a 63% year-on-year increase. The portion of ad spend still fully managed by human marketers is growing at just 5% by comparison.
The core problem is not that AI is taking over, it’s that most marketers are still investing time in the settings the algorithm has already made irrelevant. Manual audience targeting, bid adjustments, and placement exclusions now have diminishing returns. Meanwhile, the inputs that actually determine AI ad performance – creative quality, first-party data, audience signal strength, are being neglected by us.
When given the choice between campaign transparency and campaign performance, research shows advertisers chose performance every time. The algorithm has earned the trust. The question is whether marketers have updated their skill set to match.
✓ What’s Changed: The Role of the Marketer Has Shifted from Operator to Strategist
In 2024, the dominant marketing skill was campaign configuration. In 2026, it’s creative and data strategy. Google Performance Max and Meta Advantage+ have effectively automated the middle layer of digital advertising. The targeting and bidding decisions that used to define a media buyer’s day. What they can’t automate is the quality of the creative assets fed into them, the richness of first-party audience signals, and the brand distinctiveness that makes one ad worth clicking over another.
The marketers outperforming benchmarks on AI-driven platforms in 2026 share a common pattern … they treat creative development and audience data as the primary performance levers, not the campaign structure itself. This is a meaningful shift in where senior marketing talent should be focused.
→ What Marketers Should Do About AI Ad Platform Dominance in 2026
Run a creative quality audit on your top 5 active campaigns this week.
For each one, ask 3 questions:
- Does the headline lead with customer benefit, or with brand messaging?
- Are you testing at least 3 distinct creative angles – not three color variations of the same concept?
- Is your brand voice differentiated enough to earn attention in a feed where every competitor is also using AI to generate content?
Fix the weakest asset before running any further optimization.
AI ad platforms amplify performance signals. This means poor creative under-performs faster in 2026 than it did in 2023. Better inputs produce compounding returns. This is now the highest-leverage activity in paid media.
The right question is no longer “how do I optimize this campaign?” It’s “what does this AI system need from me to perform?”
The answer: high-quality creative, clean first-party data, and a brand signal strong enough for the algorithm to amplify.
4. What Do Meta’s New AI Agents Actually Do in Ads Manager [and how should marketers be using them]
➡️ The Problem: Manual Marketing Workflows Are Now a Competitive Liability
The average marketing team still runs fragmented manual workflows. Reports are pulled from one platform, pasted into another, briefed to a creator over email, and tracked in a spreadsheet. This approach was inefficient in 2022. In 2026, it’s a competitive disadvantage.
AI agents become capable of performing these tasks autonomously – reading data, making recommendations, drafting communications. Yet teams are still relying on manual processes are operating at a structural speed deficit.
The problem isn’t awareness. Most marketers know AI tools exist. The gap is integration. Only 14% of small business owners have embedded AI into daily operations, according to a Goldman Sachs survey of 1,256 owners (March 2026), despite 93% reporting positive results from AI tools they’ve tried.
✓ What Meta’s AI Agents Now Do Inside Ads Manager, Instagram, and WhatsApp
In March 2026, Meta deployed AI agents across its entire marketing platform – only two months after acquiring Manus. Inside Meta Ads Manager, the agent now reads campaign performance data and finds optimisation recommendations, eliminating the need to manually cross-reference reports. On Instagram’s Creator Marketplace, a separate agent evaluates audience-creator compatibility and flags match quality. On WhatsApp Business, an agent drafts client replies and manages project communications.
This is not a feature update. Meta has built an autonomous marketing layer that operates inside the tools marketers already use daily. The practical result is routine analytical and communication tasks that previously took hours per week, can now be delegated to the agent layer entirely.
→ How to Start Using Meta AI Agents Effectively as a Marketer in 2026
Open Meta Ads Manager and read what the agent is flagging – not that you need to act on every recommendation, but to understand its analytical frame.
Knowing what the system prioritizes trains our own strategic judgement and surfaces patterns we may have missed. Spend 20 minutes reviewing its current output before making any campaign changes.
Then identify one manual workflow our team runs every week – a performance report, or creator brief, or a client update, and assess whether the agent can inform or replace it.
Teams that integrate AI agents into one core workflow first build the organizational muscle to scale it. The efficiency gap between agent-integrated and manual-workflow teams will compound significantly over the next 12 months.
“Discovery is shifting from search engines to AI agents. Brands will need to optimise for agent-driven recommendations rather than traditional search rankings.”
— MarketingProfs AI Update, March 2026
5. Will Google Put Ads Inside Gemini? What the Latest Signals Mean for Marketers [and how to prepare now]
➡️ The Problem: Every Paid Media Playbook Was Built for Interruption – Not Conversation
Paid search and display advertising are interruption formats. They work by placing a brand message in front of a user who is browsing, scrolling, or searching, and capturing attention before it moves elsewhere. Conversational AI interfaces operate on a completely different psychological model. A user inside Google Gemini is in deep answer-seeking mode. They asked a specific question and are waiting for a specific answer. An intrusive or off-context ad format in that environment will be ignored at best, and will damage brand perception at worst.
Most marketing teams have no conversational ad creative framework, no measurement model for AI interface placements, and no briefing process for native AI ad formats, because those formats don’t formally exist yet. That’s the problem. The window to prepare is now.
✓ What Google Has Signalled About Advertising Inside Gemini
Google has not confirmed ads inside Gemini. But at the IAB NewFronts in March 2026, Google’s VP and General Manager of Ads & Commerce, Vidhya Srinivasan, explicitly declined to rule it out, leaving the door open to future advertising integration within AI-generated responses. Industry analysts interpret this as a deliberate signal rather than an omission.
The commercial logic is clear. As AI Overviews and Gemini responses reduce traditional organic search click-through rates, Google must develop new monetisation surfaces within conversational interfaces. The most likely format is a contextually integrated sponsored response, a brand answer that appears native to the AI conversation rather than as a separate ad unit. This represents the next major disruption to the $200+ billion global paid search market.
→ How to Prepare Your Brand for Google Gemini Advertising Before It Launches
The marketers best positioned when Gemini ads launch will be those who have already developed and tested a conversational creative framework.
Start building yours now, using this brief … what does a genuinely useful brand message look like inside an AI-generated answer?
The answer is not a headline and a CTA. It’s a piece of content that earns its place in the conversation – that the user would find valuable even if they knew it was sponsored.
Begin testing conversational copy formats in our existing paid channels, particularly in Performance Max asset groups and Meta Advantage+ creative.
Measure engagement on copy that leads with helpfulness over brand messaging. Document what works.
When Google Gemini ad formats are confirmed, we’ll have a tested playbook. Everyone else will be writing their first brief under a two-week deadline.
Take away for this week …
01
LinkedIN Solution
LinkedIn is the most-cited professional domain in AI responses — and citation frequency has doubled.
What we should do this week … build a thought leadership publishing system around your senior brand voices. Four posts per month. Minimum 90-day commitment. Depth over volume.
02
AI Citation Solution
Only 15% of pages ChatGPT retrieves are ever cited.
What we should do this week … run a citation audit using ChatGPT and Perplexity. Rewrite your three lowest-performing pages as direct expert answers — named author, original insight, answer-first structure, credible outbound links.
03
Paid Media Solution
AI ad platforms now control targeting and bidding.
What we should do this week … our edge has moved entirely to creative quality, first-party data, and brand distinctiveness. Audit our top five campaigns this week against those three criteria.
04
Google Ad Solution
Google has signalled ads inside Gemini are coming.
What we should do this week … build a conversational ad creative framework now — before the formats are confirmed. Test helpfulness-led copy in existing channels as preparation.
The brands that will win the next three years of marketing are not the ones with the biggest budgets. They are the ones that understand AI systems are now the primary gatekeepers of discovery — across paid media, organic search, social platforms, and e-commerce — and who built their infrastructure for that reality before their competitors recognized the shift.
Every one of the six problems above has a solvable fix. None of them require a large budget. All of them require the willingness to update the playbook.
About the Author ::
Nicola Ziady is a CMO with over two decades of experience in higher education and healthcare marketing. Connect with Nicola on LinkedIn.
Article Information ::
Published: March 21 2026. Updated April 10 2026.
Frequently Ask Questions
According to Madison & Wall’s 2026 US ad spend forecast, $57 billion will flow through AI-powered advertising platforms in 2026 – a 63% year-on-year increase. Human-managed ad spend is growing at just 5% by comparison, signalling a decisive shift toward AI-automated campaign management via platforms like Google Performance Max and Meta Advantage+.
Meta’s AI agents, deployed in March 2026 following its acquisition of Manus, perform three functions across its marketing platform. Inside Meta Ads Manager, the agent reads live campaign performance data and surfaces specific optimisation recommendations. On Instagram’s Creator Marketplace, it evaluates audience-creator compatibility. On WhatsApp Business, it drafts client replies and manages project communications – all without requiring manual input from the marketer.
Google has not officially confirmed advertising inside Gemini, but at the IAB NewFronts in March 2026, Google’s VP and General Manager of Ads & Commerce, Vidhya Srinivasan, declined to rule it out — a signal industry analysts interpret as confirmation that Gemini ad placements are in development. The most likely format is a contextually integrated sponsored response appearing native within an AI-generated answer, rather than a traditional display or search unit.
According to research reported by MarketingProfs in March 2026, LinkedIn has become the most frequently cited professional domain in responses from ChatGPT, Claude, and Google Gemini, with citation frequency doubling in recent months. AI language models preferentially cite publicly indexed LinkedIn content- posts, long-form articles, newsletters, and company pages – because it demonstrates topical depth, named authorship, and consistent publishing history. This makes LinkedIn publishing a form of AEO (Answer Engine Optimisation) infrastructure for B2B marketers, not just a social media activity.
Research published in March 2026 found that ChatGPT retrieves content from many pages when formulating a response but cites only 15% of them in the final answer. The remaining 85% receive ghost traffic – read by the AI system but never surfaced to the user. Pages that get cited share four properties: specificity (they answer a precise question), authority (named author, trusted domain), originality (unique data, analysis, or perspective not found elsewhere), and extractability (clear headers, concise paragraphs, answer-first structure). Generic brand copy and pages without named authors are retrieved but rarely cited.
SEO (Search Engine Optimisation) focuses on ranking in traditional search results like Google. AEO (Answer Engine Optimisation) focuses on being cited in AI-generated answers from systems like ChatGPT, Perplexity, and Google Gemini. GEO (Generative Engine Optimisation) focuses on ensuring brand visibility within the outputs of generative AI models broadly. In 2026, these three disciplines are converging – the content signals that earn AI citations are largely the same signals Google’s helpful content system rewards for traditional search ranking: named authorship, original insight, structured answers, and genuine topical depth.
To optimise content for AI citation in 2026, focus on four factors:
(1) Answer specificity – each piece of content should directly answer a single precise question our target audience would ask an AI system.
(2) Named authorship – content with a visible, credentialed author is significantly more likely to be cited than anonymous or brand-attributed content.
(3) Original insight – proprietary data, frameworks, or analysis that cannot be found elsewhere gives AI systems a reason to cite the source.
(4) Answer-first structure – leading every section with the direct answer, supported by evidence, in clear concise paragraphs with descriptive headers. Testing content performance in ChatGPT and Perplexity is the fastest way to identify our citation gaps.
