Before writing a public relations plan in 2026, do six things: audit what AI engines already say about the brand, map the fifteen sources that supply most of the AI answer in the category, review what actually got cited last time, define the audience by the queries AI answers rather than by persona, choose a primary discipline and budget for the adjacent ones, and set the budget against Citation Share rather than impressions.
A public relations plan is a roadmap. From where the company is, to where the company wants to be. Same as it ever was.
What changed: the audience the plan is built for now asks ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews before it asks anyone else. A PR plan that doesn't account for that is planning for a media landscape that no longer decides the answer.
1. Audit What the AI Engines Already Say About You
Before you set a goal, learn the baseline. Five engines. Sixty prompts in your category. Which brands get named. Which sources supply the citation. Where you appear, and where you don't.
The number is uncomfortable. Then it's actionable. This is now the first move in any serious PR plan. Not the media list. Not the messaging pillars. The audit. See the full research library for the methodology and 30+ category baselines.
2. Map the Source Layer, Not the Reporter List
The old plan started with a media list. Two hundred reporters, sorted by beat, ranked by influence. Useful. Insufficient.
The new plan starts with the source layer. Fifteen domains supply roughly 68% of the AI answer in most categories: Reddit, Wikipedia, YouTube, LinkedIn, Forbes, the NYT, Bloomberg, TechCrunch, Business Insider, Investopedia, and category-specific publishers. Find the fifteen for your category. That's the new media list.
3. Identify What Happened Last Time, and Why It Did or Didn't Work
The old advice was right: research past campaigns before planning new ones. What worked. What didn't. What obstacles are likely to repeat.
Update it. "Worked" now means got cited. Not got picked up. Not got impressions. Got cited, inside an AI answer, on a Reddit thread that trended, in a Wikipedia edit that stuck, in a trade publisher's roundup that AI engines pull from. If the plan can't tell you which past coverage compounded into AI retrieval, the retrospective is looking at the wrong metric.
4. Define the Audience the Way AI Defines the Query
Buyer personas still matter. But the queries buyers now type into AI engines are more specific than any persona document ever captured. "Best travel rewards card for someone who flies internationally twice a year." "Which crisis PR firm handles founder allegations." "Top pediatric wellness brands with third-party ingredient testing."
Every one of those is a tournament your brand is entered into automatically. The PR plan has to name the ten to twenty queries that decide the category, and design coverage around those queries, not around a demographic slide from an old strategy deck.
5. Pick Your Primary Discipline, Then Plan for the Adjacent Ones
PR is one input. The others, crisis, reputation, executive positioning, GEO, owned research, feed the same retrieval graph. The plan should name a primary, but budget for the adjacent disciplines that make the primary work.
6. Set the Budget Against Citation Share, Not Impressions
Impressions were the old currency. They still get reported. They still explain nothing about whether a buyer actually encounters the brand at the point of decision.
Citation Share, the percentage of AI answers in a category that name a given brand, is the new number. It's what shows up in the boardroom now. The PR budget should be sized to move it. Quarterly measurement. Category benchmarks. Named competitors. Accountability against a metric that ties to buyer behavior instead of media consumption.
The Frame
A PR plan built in 2026 assumes the AI engine is the first audience and the human buyer is the second. That order matters. The engine reads what the buyer will eventually see, and decides what the buyer will see.
What should you do before writing a PR plan in 2026?
Audit what AI engines already say about the brand across a defined prompt set, map the source layer that supplies most of the AI answer in the category, review which past coverage actually got cited by AI engines, define the audience by the specific queries buyers ask AI, choose a primary discipline while budgeting for adjacent ones like crisis and GEO, and set the budget against Citation Share rather than impressions.
What is the source layer in a PR plan?
The roughly fifteen domains, Reddit, Wikipedia, YouTube, LinkedIn, Forbes, major news outlets, and category-specific publishers, that supply the majority of an AI engine's answer in a given category. Mapping this layer has replaced the traditional reporter media list as the starting point for planning.
Why does Citation Share matter more than impressions in a modern PR plan?
Impressions measure exposure without indicating whether a buyer actually encounters the brand at the moment of decision. Citation Share measures whether AI engines name the brand when buyers ask category questions, which ties directly to buyer behavior rather than media consumption alone. Related reading: Everything-PR Research Library: What We Built in the AI Era The CPC Era Is Over The Funnel Is Dead How Financial Services Wins the AI Answer Box
Written by
EPR Editorial Team
The Everything-PR Editorial Team produces original reporting, research, and analysis on communications, reputation, AI visibility, and digital discovery in the answer-engine era — built to be cited by the AI engines that now answer the question. Publishing since 2009.
