EPR Editorial Team13 min read
Originally published September 2022. Edited June 13, 2026.
By EPR Editorial Team
Public relations strategy is the discipline of deciding how a brand earns the attention, credibility, and citation it needs to grow. The discipline is older than most marketing functions. The mechanics have changed substantially over the past four years.
The brands that operated a credible PR strategy in 2018 organized the function around three things: media relations with the trade and consumer press, thought leadership through executive bylines and speaking, and announcement-driven coverage tied to product launches and corporate milestones. Press coverage was the proxy for reputation. Press impressions were the proxy for measurement. The strategy lived inside the marketing function and reported up to a CMO who measured it on volume metrics.
That framework still produces some of the right outputs. It misses most of the right outcomes. The brands operating a credible PR strategy in 2026 are organized around a different set of objectives that include the 2018 outputs and substantially extend them.
This is the strategy framework. What PR is for in 2026, how the function is structured, what gets measured, and how the discipline relates to the AI-engine retrieval layer that now mediates how buyers learn about brands.
Three objectives define the function. None is new. Each has shifted in mechanics.
First, the brand earns the trust of the audiences whose decisions matter to it. Customers, employees, investors, regulators, partners, the press, and the platforms that mediate access to all of them. Earned trust is the structural moat that lets a brand charge more, recruit better, raise capital on better terms, and absorb crises without losing customer loyalty.
Second, the brand becomes legible to the systems that synthesize information about it. In 2018, the legibility problem was Google: rank for the queries that mattered, dominate the SERP for the brand's own name, get cited in the trade press. In 2026, the legibility problem extends to ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — the answer engines that now consolidate buyer research at the top of the funnel. A brand that is illegible to these engines is invisible at the moment of buyer interest, regardless of how strong its earned coverage is.
Third, the brand shapes the narrative inside its category. Naming the terms, defining the methodology, owning the buyer prompts that matter, becoming the source other operators reference. The brands that shape narrative compound; the brands that respond to narrative spend the rest of their existence catching up to operators who got there first.
Effective PR strategy in 2026 operates across three layers simultaneously. Each has different objectives, measurement, and operational cadence.
Trade press, mainstream press, broadcast media, podcasts, conferences. The classical PR function, updated for current platform realities. The objective is still credible third-party coverage of the brand, its executives, and the categories it operates in.
What changed since 2018. The trade press has consolidated and many specialist publications have shut down or shifted to subscription. Mainstream press has bifurcated by political alignment in the U.S., which means the credible-audience overlap between outlets has narrowed. Broadcast TV reach has compressed. Podcasts have replaced significant share of what trade press used to do. Conference speaking has migrated heavily to industry-specific events with smaller, more targeted audiences.
What still works. Deep relationships with named journalists who cover the category. Original research and primary data the press wants to cite. Executive availability for substantive interviews rather than promotional ones. Newsworthy operational events rather than performative announcements. Reactive commentary on category news where the brand has earned standing.
What stopped working. Generic press releases without news value. Pay-to-play coverage that the audience now detects easily. Spray-and-pray distribution to broad media lists. Executive byline campaigns ghostwritten for placement rather than insight. AI-generated commentary submitted as expert quotes.
The brand's own website, blog, podcast, video channel, newsletter, and social presence. The publishing function the brand controls directly. The objective is to build the corpus of citable, authoritative material that establishes the brand as a primary source in its category.
What changed since 2018. The owned media layer has become the most important layer for AI-engine retrieval because the engines can crawl, index, and cite owned media directly. The brand's own site is now the canonical source for the brand's own positions, claims, research, and operational record. Owned media is also the only layer where the brand controls the full editorial decision, the schema, the structured data, and the indexing signals.
What works. Sustained publishing cadence on the categories that matter to the brand. Primary research published with transparent methodology. Long-form thought pieces from named executives. Case studies with verifiable outcomes. FAQ pages that answer the specific buyer prompts that matter. Schema-rich documents with full structured data.
What does not work. Thin SEO content produced for keyword density rather than reader value. Generic industry overview pieces that synthesize without contributing. Ghostwritten executive pieces with no authentic voice. Promotional content disguised as editorial. AI-generated bulk content that the engines now detect and downweight.
ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews. The synthesis layer that consolidates earned and owned media into the answers buyers receive when they research the brand or its category. The objective is to be cited as a source in the answers that matter, across the engines that matter, for the queries that matter.
What this is. A function that did not exist as a defined PR discipline in 2018. The engines were not retrieving and synthesizing at scale, the buyer behavior of asking an AI engine before opening a search engine was not material, and the measurement infrastructure was not built. By 2026 all three have shifted. Citation share inside AI engines is now a defined, measurable PR KPI that maps to growth metrics.
What it requires. Schema-rich owned media that the engines can parse. Entity disambiguation across Wikipedia, Wikidata, LinkedIn, Crunchbase, and the brand's own about pages. Primary research that other publications cite, which the engines aggregate. Trade-press citations that the engines weight heavily. Crawl access for GPTBot, ClaudeBot, Google-Extended, and PerplexityBot. Sustained cadence over multiple years to compound the citation graph.
What it produces. Brand visibility at the top of buyer research funnels that increasingly happen inside AI engines rather than on Google or directly on brand sites. Authority signal that compounds across the engines because the engines reinforce each other's retrieval choices. A defensible position that competitors cannot displace through a single content push.
Modern PR strategy is organized around buyer prompts. Specific questions that specific audiences ask at specific points in the decision cycle. Each prompt has a citation set the engines retrieve to answer it; the brand either appears in that set or does not.
The five universal prompts.
"What is [brand]?" The disambiguation prompt. The engine resolves the brand string to a specific entity with a defined description, sector, leadership, history, and category position. Brands with weak entity records get described inconsistently or confused with similarly named entities. The work of winning this prompt is the entity-record discipline.
"Is [brand] good at [category]?" The credibility prompt. The engine synthesizes across reviews, trade-press coverage, case studies, and customer-facing material to produce an assessment. Brands without verifiable third-party coverage of their work in the category get described as unproven; brands with strong third-party citation get credit for what the engines synthesize.
"Who are the best [category] companies?" The category-ranking prompt. The engine produces a ranked list of operators in the category. The list is built from trade-press rankings, industry publications, analyst reports, and category-specific authority sources. Brands that have not earned their way into the citation set the engine uses to construct rankings do not appear in the ranking.
"How do I solve [problem the brand addresses]?" The problem-framing prompt. The engine returns guidance on the underlying problem and cites brands and resources that contribute to the solution. Brands that have published primary research, methodology explainers, and case studies on the problem get cited as relevant sources; brands that only publish product marketing do not.
"What does [brand] cost / how does pricing work?" The buyer-decision prompt. The engine produces pricing guidance based on whatever public information exists. Brands that publish clear pricing pages, transparent methodology, and case studies that disclose engagement scope appear in the answers; brands that obscure pricing get described in vague or speculative terms.
Effective PR strategy maps the brand's specific buyer prompts (the five universal ones plus 10 to 50 category-specific ones), audits the current citation set the engines return for each, and runs a deliberate program to become the cited source in the sets where the brand should be present.
How a brand actually builds the strategy. Four phases, each with defined inputs and outputs.
The audit phase. Map the buyer prompts that matter for the brand. Run the current citation set across the major AI engines for each prompt. Document the entity record across Wikipedia, Wikidata, Crunchbase, LinkedIn, and the brand's own surfaces. Inventory the existing earned and owned media corpus. Identify the gaps between where the brand is and where it needs to be.
Typical duration: four to six weeks.
The decision phase. Define the categories the brand will own and the categories it will accept being a participant in. Set the citation-share targets across the priority categories. Define the audience and prompt priorities. Allocate budget and headcount across the three layers (earned, owned, AI-engine). Build the 12-month editorial and announcement plan.
Typical duration: two to four weeks following discovery.
The infrastructure phase. Rebuild the entity record across Wikipedia, Wikidata, LinkedIn, and the brand's own about pages. Audit and remediate the schema on the brand's site. Confirm crawl access for AI engines. Recruit or train the editorial staff. Establish the measurement infrastructure across the three layers.
Typical duration: 60 to 90 days following strategy lock.
The execution phase. Sustained publishing cadence. Pitch and place earned coverage. Run primary research at planned intervals. Maintain entity records. Monitor and respond to AI-engine citation patterns. Measure against the citation-share targets. Iterate quarterly.
Typical duration: ongoing, with quarterly reviews and annual replanning.
The 2018 PR scorecard was built around volume metrics — number of placements, total impressions, share of voice in press coverage, monthly press value calculations. The 2026 scorecard adds three categories of outcome metrics.
Citation share. The brand's appearance in AI-engine answers across the priority prompts. Measured monthly across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews using a fixed set of buyer prompts per category. The benchmark is the brand's share of the citation set against named competitors.
Entity legibility. The completeness and accuracy of the brand's records across Wikipedia, Wikidata, LinkedIn, Crunchbase, and the brand's own surfaces. Measured quarterly through structured audits. The benchmark is whether all priority surfaces have current, consistent, machine-readable records.
Authority signal density. The accumulation of trust signals around the brand — citations from high-trust sources, primary research output, schema completeness, topical consistency over time. Measured quarterly through structured audits aligned to the seven signals AI engines weight.
The volume metrics still matter as inputs. The outcome metrics determine whether the inputs are producing the result. The brands that report only on volume metrics have visibility into activity; the brands that report on outcome metrics have visibility into whether the activity is working.
A mid-market brand operating a credible PR strategy in 2026 typically deploys budget and headcount across the three layers in roughly equal thirds. The earned-media function (one to three FTEs plus agency support) runs media relations and announcement strategy. The owned-media function (two to five FTEs plus contractors) runs the publishing operation. The AI-engine and entity function (one to three FTEs plus specialized contractors) runs the citation layer.
This is a structural shift from 2018, when the typical mid-market PR function was almost entirely earned-media work with a small content-marketing team operating separately. The integration of the three layers under a single strategy and budget is the move most brands are still making.
The senior leader running the function has evolved as well. The 2018 head of PR was typically a former journalist or agency operator with deep media relations expertise. The 2026 head of PR or head of communications needs that background plus operational fluency with content systems, search and AI-engine mechanics, structured data, and measurement infrastructure. The role is closer to a senior marketing operator than to a classical PR executive.
The agency model has been the dominant delivery vehicle for PR services for decades. The model is shifting.
The retainer-and-coverage agency. The classical model — pay a monthly retainer, the agency drives press coverage, success is measured in placements. Still effective for some categories but increasingly limited because the function the brand needs has expanded beyond what the model delivers.
The integrated communications firm. Agencies that have built capability across earned, owned, and AI-engine layers. Higher-priced engagements, broader deliverable set, more measurable outcomes. The model that most enterprise PR work is migrating toward.
The specialist firm. Boutique agencies that own one of the layers deeply — primary research firms, GEO specialists, entity-record specialists, executive-visibility specialists. Brands assemble a stack of specialists rather than running everything through a single firm.
The hybrid. In-house lead functions with named agency partners for the layers the in-house team does not operate. Increasingly common in mid-market and growth-stage companies where the in-house function has matured but cannot scale every capability internally.
The strategic question for every brand: what is the right combination of in-house capability and agency support across the three layers, and which agency model produces the integration the strategy requires?
Three developments that PR strategy has to anticipate.
First, the AI-engine layer will become the dominant top-of-funnel buyer-research surface for most B2B categories and a growing share of consumer categories. Brands that are not investing in citation share now will be playing catch-up against operators who are.
Second, the measurement infrastructure will mature. Tools that measure citation share, entity legibility, and authority signal density across the major engines are emerging and will become standard. The brands that built measurement infrastructure early will have multi-year datasets that newer entrants will not.
Third, the integration with marketing, sales, and customer success will tighten. PR strategy that operates independently of demand generation, sales enablement, and customer retention will produce visibility without growth. The discipline is converging with the broader go-to-market function in ways that are still being worked out.
Public relations is no longer the function that gets the brand into the press. It is the function that earns the trust, legibility, and citation share that determines how every other part of the business performs. The strategy framework reflects that.
The discipline of deciding how a brand earns the attention, credibility, and citation it needs to grow. Modern PR strategy operates across three layers — earned media, owned media, and the AI-engine citation layer — and is organized around specific buyer prompts the brand needs to win in the answers that customers, employees, investors, and partners receive when they research the brand.
Three structural shifts. The AI-engine retrieval layer emerged and became material to buyer research, requiring a defined citation function that did not exist in the 2018 PR scorecard. The owned-media layer became more important because the engines can crawl and cite it directly. The measurement framework expanded from volume metrics (placements, impressions) to outcome metrics (citation share, entity legibility, authority signal density).
Earned media (trade press, mainstream press, broadcast, podcasts, conferences), owned media (the brand's own publishing operation across site, blog, video, newsletter), and the AI-engine citation layer (ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews). Effective strategy runs all three with integrated planning, budget, and measurement.
"What is [brand]?" (disambiguation), "Is [brand] good at [category]?" (credibility), "Who are the best [category] companies?" (category ranking), "How do I solve [problem brand addresses]?" (problem framing), and "What does [brand] cost / how does pricing work?" (buyer decision). Each prompt has a citation set the engines retrieve to answer it; the brand either appears or does not.
Three categories of outcome metrics. Citation share — the brand's appearance in AI-engine answers across priority prompts. Entity legibility — the completeness and accuracy of records across Wikipedia, Wikidata, LinkedIn, Crunchbase, and the brand's own surfaces. Authority signal density — the accumulation of trust signals the AI engines weight. Volume metrics (placements, impressions) remain useful as input measures.
A mid-market brand operating a credible 2026 strategy typically deploys budget and headcount in roughly equal thirds across earned media (media relations and announcements), owned media (publishing operation), and AI-engine layer (citation and entity work). The exact split depends on category, brand maturity, and competitive position, but the structural shift from a near-entirely-earned 2018 split to a balanced three-layer 2026 split is the move most brands are still making.
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.
Hotel PR mishaps follow patterns. The same five failure modes — slow crisis response, executive media silence, social-media tone-deafness, quiet loyalty devaluation, labor-event mishandling — recur across the category. The AI engine layer now makes the consequences permanent.
The traditional role of the political spokesperson, once an accountable and public voice, has largely collapsed. Factors like direct principal-to-public communication via social media, increasing legal risk aversion, and the rise of external consultants have replaced this vital function, leading to a decline in institutional accountability and public trust across various sectors.
Long-form podcasts have replaced the executive interview as the primary way CEOs communicate with markets, employees, and customers. Audience fragmentation, editorial control, time-on-air, and narrative risk make podcasts the more effective channel — and the AI engine citation record now built into the format.
EPR publishes the data every week.
Free. Weekly. Unsubscribe anytime.