The Vertical PR Map: Where to Pitch If You Want AI Citations, Not Just Backlinks in 2026

Most PR pitches in 2026 are optimised for the wrong metric. They chase domain rating and journalist bylines at publications that AI engines rarely cite. If you want your brand to appear when someone asks ChatGPT or Perplexity for a recommendation in your category, the publication list your agency handed you last year is probably wrong. Here's how to build one that actually works.

Why Your Current Media List Isn't Producing AI Citations

The publications AI engines cite are not always the publications PR teams target. The gap between those two lists is where most digital PR budgets disappear.

The clearest evidence of this problem comes from a real campaign audit. A four-facility behavioral health operator discovered that roughly three-quarters of their PR budget had been targeting publications that LLMs do not cite for treatment-related queries, according to an audit published by Webserv in June 2026. A $48,000 quarterly PR fee, producing placements that moved nothing in AI search. The publications looked impressive on a coverage report. They were invisible to the models that now answer patient queries.

That's not a behavioral health problem. It's a vertical PR problem, and it affects every sector. AI engines have strong publication preferences by topic category, and those preferences do not map neatly onto traditional media prestige or domain authority.

How AI Engines Actually Select Citations

Different AI engines pull from different source pools, and understanding that split is the first step in building a smarter pitch list.

ChatGPT Search retrieves live web content via Bing's index. That means Bing crawlability is table stakes: if Bing hasn't indexed a page, ChatGPT can't retrieve it. Perplexity runs its own crawler and averages 8.2 cited sources per answer, the highest citation density of any mainstream AI search engine, according to data from Omnibound. Claude uses Brave Search's index. Google AI Overviews draw heavily from Google's top organic results.

Each engine has a different retrieval bias, which means a single placement rarely covers all four surfaces. The brands building real AI citation authority are earning coverage across multiple source types, not betting everything on one Forbes feature.

B2B referral share data sourced from Enterprise DNA's 2026 AI referral analysis. Claude's 18.5% B2B referral share is disproportionate to its web traffic share, which tells you something about the research intent of Claude users and the value of appearing there.

What a Vertical PR Map Is and How to Build One

A vertical PR map is a publication list built by auditing what the AI engines in your category actually cite, not what your agency considers tier-one media. You build it by running your target queries across ChatGPT, Perplexity, Claude, and Google AI Overviews, then recording which sources appear in the citations.

The process has four stages:

1. Define your citation-worthy queries. These are the prompts your prospective customers are actually typing. "Best [category] software for [use case]," "[your category] compared to [alternative]," "who are the leading [category] providers" and similar recommendation-style queries. If you're not sure which prompts to start with, the intent taxonomy matters: category queries, comparison queries, recommendation queries, and problem-solution queries each pull different citation patterns. BrandPrompts generates research-backed prompt sets from real search data if you need a structured starting point.
2. Run each query across all four major engines. Screenshot or record every cited source. Do this for at least 20-30 queries per topic cluster. One-off spot checks produce misleading results because AI responses aren't deterministic.
3. Cross-reference with your existing coverage. Map every placement your PR activity has produced in the last 12-18 months against the citation list. Anything outside that list is backlink-building, not citation-building. Both have value, but you need to know which you're buying.
4. Build your target list from the citation audit, not from DA scores. A niche industry publication that ChatGPT cites consistently for your category is worth more for GEO than a DR90 generalist outlet that never appears in your category's AI responses.

Which Publication Types Win by Vertical

The citation patterns vary substantially by sector. Here's what we observe consistently across verticals when auditing AI citations:

Healthcare and wellness: Peer-reviewed adjacent sources, established medical journalism (Healthline, WebMD, medical school publications), and state licensing body resources appear frequently. Community forums and Reddit's health communities appear for Perplexity. Prestige business press rarely appears for clinical queries.

B2B SaaS and technology: G2, Capterra, and TrustRadius are heavily cited for comparison and recommendation queries across ChatGPT and Perplexity. Vertical trade publications (not just TechCrunch) appear for category-specific queries. Reddit's SaaS and programming subreddits surface consistently on Perplexity.

Finance and fintech: Investopedia, NerdWallet, and regulatory body pages dominate. Mainstream business press (Forbes, Bloomberg) appears but rarely for specific product recommendations. The citations skew toward authoritative explainers and comparison tools.

Legal and professional services: Avvo, FindLaw, and bar association resources appear. Practice area blogs from established firms appear for specific legal topic queries. Academic and think-tank content surfaces for Claude specifically.

Consumer retail and ecommerce: AI referral traffic to US retail sites grew 693% year-over-year during the 2025 holiday season, according to Mashable. For product categories, review platforms, specialist editorial (Wirecutter equivalents), and niche community sites dominate citations. Brand-owned content rarely appears for "best product" queries across any engine.

How to Pitch for AI Citations Specifically

Getting covered is necessary but not sufficient. The coverage needs to be structured in a way that AI engines can extract and reproduce. That means your pitch materials and your follow-through on the journalist's piece both matter.

When you're pitching data or expert commentary, give the journalist a quote or a statistic that's genuinely quotable in isolation. AI engines pull passages that make sense without full-page context. A vague paragraph about industry trends won't get extracted. A specific claim with a named source will.

Push for structured mentions. If you're pitching a comparison roundup, ask that your product be listed with specific feature claims alongside competitors, not just mentioned in a paragraph. List-format coverage is consistently more likely to surface in AI responses than prose-only mentions.

Target earned placements, not contributed content. AI engines are systematically biased toward third-party editorial over brand-authored content. A contributed article from your CEO may help brand awareness, but it carries less weight for AI citation than an independent journalist writing about your product without your byline on it.

Make sure the publications you target are crawlable by the right bots. Any publication that blocks OAI-SearchBot, PerplexityBot, or ClaudeBot in its robots.txt is producing backlinks, not citations. This is worth checking before you invest time in a pitch relationship.

The Compounding Problem with Wrong Publication Lists

Here's what makes this urgent: the brands building AI citation authority right now are creating a structural advantage that's getting harder to close. AI engines develop what amounts to a preferred source set for each topic category. Once a publication becomes a consistent citation source for a given query type, it tends to stay cited. Your competitors who figure this out before you do are building a citation position that takes months to displace.

The scale of the opportunity is real. Google AI Overviews now reach over 2 billion users monthly across more than 200 countries. ChatGPT crossed 1.1 billion monthly active users by May 2026. These aren't niche channels. They're the primary information surface for a growing portion of your market. Brands that appear there get considered. Brands that don't, often aren't.

The right response isn't to panic and rebuild your entire PR programme overnight. It's to audit first, understand which placements are producing AI visibility versus backlinks alone, and progressively shift your pitch list toward the publications your citation audit identifies. You can track whether it's working by monitoring your brand's citation share across engines over time. BrandPrompts can generate the prompt sets you need to run that monitoring systematically.

Frequently Asked Questions

Does a high domain authority publication always produce AI citations?

No. Domain authority correlates with citation likelihood in some cases, but the more important factor is topical relevance. A DR45 specialist publication that covers your exact category consistently outperforms a DR90 generalist outlet for AI citations in that category. Audit by topic, not just by authority score.

How often do I need to rebuild my vertical PR map?

Roughly every quarter. AI models are updated regularly, Perplexity's crawler behaviour shifts, and the publication space changes. A citation audit you ran six months ago may not reflect current model behaviour. Build the audit into a quarterly review cycle rather than treating it as a one-time project.

Should I pitch Reddit directly as part of my PR strategy?

Yes, but carefully. Reddit is a consistent citation source for ChatGPT and Perplexity, especially for product recommendation queries. The right approach is genuine participation: answer questions in relevant subreddits with specific, useful information. Direct promotion gets flagged by communities and rarely gets cited anyway. The goal is to appear in threads that AI engines pull as authoritative community consensus.

Does it matter which AI engine I prioritise for citation-building?

It depends on your audience. For B2B buyers, Claude's 18.5% share of B2B AI referrals makes it disproportionately useful relative to its overall web traffic share. For consumer categories with high search volume, ChatGPT and Google AI Overviews carry more weight given their scale. The safest position is to build coverage that appears across multiple engines, which generally means a mix of Bing-indexed editorial, Brave-indexed content, and strong organic Google rankings.

How do I know if a new placement actually produced an AI citation?

Run your tracking prompts across ChatGPT, Perplexity, Claude, and Google AI Overviews before and after a significant placement goes live. Record which sources appear in the citations. If the new publication shows up in subsequent query runs, the placement is working for AI visibility. If it never appears after several weeks, you have backlink value but not citation value. That's useful information for future pitch prioritisation.