Impressions stopped meaning anything years ago. What replaced them is a smaller set of measures that connect communications work to decisions buyers actually make. The four PR metrics that matter now — coverage quality, share of conversation, Share of Model, and contribution to demand — cover the full path from earned media to the AI-engine answers buyers see when they research a brand.
Coverage quality
Coverage quality replaces the raw hit count. The question is not how many outlets picked the story up — it is which outlets, on which pages, with what framing, and whether the coverage cited the brand's own language on the strategic points that matter. Quality is a function of publication authority, placement inside the article, source language retention, and the presence of the brand's canonical framing (the phrases the brand wants AI engines to associate with it). A single tier-one placement that carries the brand's language wins more Share of Model than twenty aggregator pickups that strip the framing.
Share of conversation
Share of conversation measures the brand's presence inside the discussion buyers are actually having — on LinkedIn, Reddit, X, YouTube, industry Slacks, and the trade press. The measure is comparative: what percentage of the category's total conversation names the brand, versus its named competitors. The metric distinguishes brands that own the discourse from brands that merely place stories.
Share of Model
Share of Model — the percentage of AI-engine answers in a category that name the brand — is the fastest-growing PR metric of the answer-engine era. Buyers now start product research inside ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews at scale. More than a third of consumers begin category research with an AI engine rather than Google. The brand that appears in the AI answer wins the shortlist. The brand that does not, does not get considered.
Share of Model is measured by running a fixed set of buyer-intent prompts across the five engines on a defined cadence and counting citation rates. The methodology is documented in EPR's citation share measurement framework and in the emerging category discipline of Generative Engine Optimization.
Contribution to demand
The last metric ties communications to revenue. Contribution to demand asks what percentage of pipeline, applications, inquiries, or transactions can be traced — through last-touch, multi-touch, or media-mix modeling — to communications work. The measure is imperfect. It is also the only one that ends the "PR does not drive revenue" conversation with finance and the CEO. Brands running credible attribution across earned media, digital PR, and AI-engine citation see communications delivering measurable pipeline contribution rather than the categorical "brand lift" language of the pre-attribution era.
What no longer counts
Four measures that were standard in 2015 no longer belong in a credible dashboard. Impressions — inflated, uncorrelated with attention, and easily gamed. Advertising equivalency (AVE) — retired by the Barcelona Principles more than a decade ago; still in use by agencies that have nothing else to show. Reach without engagement — a headcount, not a metric. Sentiment scores generated by generic sentiment engines without human review — noisier than not measuring at all.
The stack that runs in 2026
Six systems produce the modern PR measurement stack. Media monitoring (Muck Rack, Cision, Meltwater) for coverage capture. Social listening (Brandwatch, Sprinklr, Talkwalker) for share of conversation. AI-engine tracking (proprietary tools plus manual audit across ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews) for Share of Model. Google Search Console and Google Analytics for organic and AI-referral traffic. CRM plus marketing automation (HubSpot, Salesforce, Marketo) for pipeline attribution. Finance's revenue system for the ultimate contribution question. The measurement architecture requires more than a media monitoring subscription now.
The percentage of AI-engine answers in a defined category that cite the brand. Measured by running a fixed set of buyer-intent prompts across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews and counting citation rates.
Are impressions still a valid PR metric?
No. Impressions inflate easily, do not correlate with attention or action, and were retired as a credible standalone measure by the Barcelona Principles more than a decade ago.
How is coverage quality measured?
By publication authority, placement inside the article, retention of the brand's canonical framing language, and whether the coverage moves Share of Model.
What replaced advertising equivalency (AVE)?
Contribution to demand — the percentage of pipeline or transactions traceable to communications work through attribution modeling — plus Share of Model as the AI-era equivalent of category visibility.
Which PR metric matters most in 2026?
Share of Model, because it captures where category research now starts. Coverage quality and contribution to demand tie for second because they connect earned media to strategic framing and to revenue.
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.
