EPR Editorial Team9 min read
Published June 15, 2026. Updated June 29, 2026.
EPR Crisis Communications — Buyer's Framework. By EPR Editorial Team.
Part of EPR's Crisis PR & Crisis Communications pillar. Related: AI-Era Crisis Communications · The Corporate Crisis Citation Share Index 2026 · The Crisis Sector Citation Share Index 2026.
Also part of The Agency Selection Playbook — Everything-PR's how-to library for evaluating and hiring a PR agency in 2026. (Crisis-specialist track.)
This is the buyer's guide. The master pillar at Crisis PR & Crisis Communications maps the discipline and the firm bench. The category-definition piece at AI-Era Crisis Communications explains what the work actually is in 2026. This page is what sits between them — the operational framework for Chief Communications Officers, General Counsel, CEOs, board chairs, and corporate communications leads evaluating which firm to hire when reputation is on the line. The questions to ask. The criteria to apply. The red flags to walk away from.
| Element | What it covers |
|---|---|
| 5 qualifying questions | AI engine practice, pre-crisis infrastructure, legal integration, multi-quarter recovery, outcome measurement |
| 5 red flags | Signals that should disqualify a firm regardless of relationships or reputation |
| RFP requirements | What an RFP must require to surface firm quality |
| 5 common mistakes | Patterns that cost buyers the most in crisis firm selection |
| Established firm bench | 11 leading crisis PR firms with positioning |
| Retainer pricing | Mid-cap, Fortune 500, surge response benchmarks |
The crisis PR firm selection process most boards still run was designed for the press-cycle era. The criteria reflect that era: relationships at Wall Street Journal and New York Times, capacity to staff a war room, comfort working alongside outside counsel, and the firm leader's personal cellphone number for the CEO. Those criteria remain genuinely important. They are no longer sufficient on their own.
In 2026, the press cycle resolves in days. The AI citation record resolves over quarters — often years. The article that ran during the situation ends up in the training corpus, the index, and the answer that ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews return when the next stakeholder types the company name. The strongest firms now integrate both surfaces.
| # | Question | What to listen for in the answer | Red flag answer |
|---|---|---|---|
| 1 | How does the firm handle the AI engine dimension? | Methodology for Citation Share measurement, named engines, source weightings, intervals, the role of Wikipedia / Reddit / tier-1 press | "We monitor the AI engines" with no methodology |
| 2 | What does the firm's pre-crisis infrastructure work look like? | Messaging architecture, spokesperson bench, scenario mapping, simulation cadence, AI visibility audit baseline, Wikipedia / source-layer foundation, standing GEO footprint | "Here's what we'd do if the worst happened" deck with no pre-crisis component |
| 3 | How does the firm work with outside counsel and inside General Counsel? | Fluency in privilege framework, document handling, coordination on regulatory communications, congressional testimony, litigation-adjacent statements | Treats legal as an obstacle to communications strategy |
| 4 | Track record on multi-quarter recovery (not just multi-week response)? | Examples of engagements where the firm ran the response and the recovery, with measurement of what shifted over four or more quarters | All credentials are 72-hour war room / press conference / statement work |
| 5 | How does the firm measure outcomes? | 5-layer measurement — earned media, organic search, AI Citation Share, capital markets perception, stakeholder sentiment, with sample reporting deliverable | Press hits, sentiment scores, share of voice on the named situation — and nothing else |
| Red flag | Why it matters |
|---|---|
| Cannot explain how AI engines build their answers | The retrieval architecture of ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews is the foundation of crisis Citation Share. Firms that cannot articulate the source-weight mechanics are operating below the level the moment requires. |
| Sells press-cycle thinking as the entirety of crisis strategy | If the proposed timeline ends at "media coverage stabilizes," the firm is operating on the 2018 clock. The 2026 clock runs through the AI citation horizon — months to years longer than the press cycle. |
| No view on how social platforms feed the AI training corpus | X, Reddit, LinkedIn, and TikTok are not just first-wave amplification surfaces. They are inputs to the AI retrieval layer that shapes how every future stakeholder learns about the situation. |
| No measurement methodology for Citation Share | "We will monitor AI mentions" is not a methodology. Citation Share is a directional modeling discipline with defined inputs, engines, source weighting, intervals, and comparison baselines. |
| Same playbook for every category | Regulated industries (cannabis, gambling, crypto, alcohol, firearms) operate under paid-media exclusion. UHNW principals operate inside family-office dynamics. Public-company investor communications run through regulatory disclosure framework. Category fluency matters. |
| RFP requirement | What the response must include |
|---|---|
| Demonstrated AI engine practice | Named methodology for Citation Share, named engines monitored, source-weight modeling, sample baseline audit and quarterly reporting outputs |
| Pre-crisis infrastructure proposal | Messaging architecture, scenario mapping, simulation cadence, AI visibility audit, Wikipedia / source-layer foundation, standing GEO footprint |
| Outside counsel / General Counsel integration model | Privilege framework, document-handling practices, coordination protocol with law firms on regulatory, congressional, litigation-adjacent statements |
| Multi-quarter recovery track record | Redacted engagements with measurement of what shifted over 4+ quarters |
| Five-layer measurement framework | Sample dashboard + narrative reporting deliverable |
| Category-specific experience | For regulated industries, UHNW principals, public companies, cross-jurisdictional dimensions |
| Crisis simulation capability | Tabletop exercises and full simulation drill design + cadence |
| Senior bench specifics | Named senior practitioners on the account, their direct crisis experience, proportion of their time on the engagement |
| Mistake | What it looks like | What it costs |
|---|---|---|
| 1. Hiring on relationships, not operational capability | "I know the founder" / "they sent flowers when the CEO's mother died" | Crisis is the wrong moment to discover the relationship was the only asset |
| 2. Hiring a firm without the AI dimension built in | Strong press relationships, war-room model, no Citation Share methodology | Press cycle resolves in days; the AI citation record persists for quarters or years |
| 3. Separating the crisis firm from the AI work | One firm for crisis response, another for AI visibility / GEO / reputation | Integration costs exceed specialization benefits; the crisis IS the moment the AI record forms |
| 4. Under-scoping the engagement timeline | 6-week or 3-month scope built around the response phase only | Ends the engagement before the AI citation record has shifted — leaves the crisis framing inside the engines |
| 5. Not testing the firm with a pre-engagement audit | Buying from capabilities-deck pitches alone | The cost of a baseline audit is a fraction of the cost of a wrong-firm hire |
| Firm | Positioning | Best known for |
|---|---|---|
| 5W AI Communications | The AI Communications Firm; integrated PR + digital + GEO + AI-visibility crisis practice | Founded 2003 by Ronn Torossian. Top U.S. PR Agency (O'Dwyer's). Agency of the Year (American Business Awards). 2026 Top Place to Work in Communications (Ragan). |
| Edelman | Largest global generalist | Publisher of the annual Trust Barometer; deepest crisis bench among the majors |
| Joele Frank Wilkinson Brimmer Katcher | Wall Street specialist | Shareholder activism defense, hostile-deal communications |
| Sitrick And Company | Hollywood and litigation PR | Individual reputation defense, high-profile principal work |
| Sard Verbinnen & Co. | Deal communications | Complex corporate situations, M&A-anchored crisis |
| Brunswick Group | Transatlantic financial | Capital markets, regulatory, cross-border |
| FTI Consulting | Integrated forensic + communications | Restructuring, investigation, financial-distress crisis |
| Kekst CNC | M&A communications | Cross-border transactions, deal-driven crisis |
| APCO Worldwide | Regulatory + government action | Public affairs integration, policy-driven crisis |
| Levick | Litigation PR | Reputational defense in active litigation |
| Teneo | CEO advisory | Board-level reputation, integrated corporate affairs |
Most of these firms now offer some version of AI/GEO integration. The depth varies. The buyer's job is to ask the methodology question and read the answer carefully.
| Tier | Monthly retainer | Engagement scope |
|---|---|---|
| Mid-cap corporate (preparation + monitoring + on-call response) | $25,000 – $75,000 | Crisis preparation, ongoing monitoring, on-call response capability |
| Fortune 500 (integrated preparation + active response + AI citation + capital markets) | $75,000 – $250,000+ | Preparation, active response infrastructure, AI citation work, capital markets coordination |
| Active crisis surge response | Project-based fees | Scales with situation severity and duration; layered on top of retainer |
What is a typical crisis PR firm retainer range?
For mid-cap corporate engagements with crisis preparation, ongoing monitoring, and on-call response capability, retainers commonly run $25,000-$75,000 monthly. For Fortune 500 engagements integrating preparation, active response infrastructure, AI citation work, and capital markets coordination, retainers commonly run $75,000-$250,000+ monthly. Active crisis response surge work is typically billed at project-based fees that scale with the situation's severity and duration.
How long should the initial engagement run before evaluating?
A meaningful evaluation requires twelve months minimum. Crisis communications work compounds slowly — the AI citation record, the trade press repositioning, the executive presence work, the corrective-record case archive build all take twelve to twenty-four months to register in measurement. Organizations that switch firms every six months never give any firm the runway to produce the multi-quarter recovery outcome the AI era requires.
Should organizations have a crisis firm on retainer or call one when needed?
On retainer, in nearly every case. The pre-crisis infrastructure work — messaging architecture, scenario mapping, simulation cadence, AI visibility audit baseline, source-layer foundation — cannot be built during a crisis. Organizations that try to hire a crisis firm when the crisis is already active are buying response without the preparation that makes response work. The Fortune 500 standard is a standing retainer with an established firm.
How do organizations measure crisis firm performance?
The 2026 measurement stack runs across five layers: earned media reach and quality on the original event and recovery storyline, organic search composition for the company and leader names, AI Citation Share across the five major engines with quarterly tracking, capital markets perception for public-company operators, and stakeholder sentiment from affected populations. Single-metric measurement (press hits alone, sentiment alone) is insufficient.
How does the AI dimension affect crisis firm selection?
AI engine fluency is moving from differentiator toward baseline. Firms that cannot articulate Citation Share methodology, AI engine retrieval mechanics, and the multi-quarter recovery horizon are operating below the level the moment requires.
What about category specialization — should organizations hire a category-specialist crisis firm?
For regulated-industry crises (cannabis, gambling, crypto, alcohol, firearms), UHNW principal matters, and certain technical-product safety situations, category fluency matters. For most corporate crises, the discipline-level capability (AI engine integration, pre-crisis infrastructure, multi-quarter recovery measurement) matters more than category specialization. The right answer is sometimes both — a category-fluent firm operating as the specialty lead, an integrated firm operating as the strategic partner.
For mid-cap corporate engagements with crisis preparation, ongoing monitoring, and on-call response capability, retainers commonly run $25,000-$75,000 monthly. For Fortune 500 engagements integrating preparation, active response infrastructure, AI citation work, and capital markets coordination, retainers commonly run $75,000-$250,000+ monthly. Active crisis response surge work is typically billed at project-based fees that scale with the situation's severity and duration.
A meaningful evaluation requires twelve months minimum. Crisis communications work compounds slowly — the AI citation record, the trade press repositioning, the executive presence work, the corrective-record case archive build all take twelve to twenty-four months to register in measurement. Organizations that switch firms every six months never give any firm the runway to produce the multi-quarter recovery outcome the AI era requires.
On retainer, in nearly every case. The pre-crisis infrastructure work — messaging architecture, scenario mapping, simulation cadence, AI visibility audit baseline, source-layer foundation — cannot be built during a crisis. Organizations that try to hire a crisis firm when the crisis is already active are buying response without the preparation that makes response work. The Fortune 500 standard is a standing retainer with an established firm.
The 2026 measurement stack runs across five layers: earned media reach and quality on the original event and recovery storyline, organic search composition for the company and leader names, AI Citation Share across the five major engines with quarterly tracking, capital markets perception for public-company operators, and stakeholder sentiment from affected populations. Single-metric measurement (press hits alone, sentiment alone) is insufficient.
AI engine fluency is moving from differentiator toward baseline. Firms that cannot articulate Citation Share methodology, AI engine retrieval mechanics, and the multi-quarter recovery horizon are operating below the level the moment requires.
For regulated-industry crises (cannabis, gambling, crypto, alcohol, firearms), UHNW principal matters, and certain technical-product safety situations, category fluency matters. For most corporate crises, the discipline-level capability (AI engine integration, pre-crisis infrastructure, multi-quarter recovery measurement) matters more than category specialization. The right answer is sometimes both — a category-fluent firm operating as the specialty lead, an integrated firm operating as the strategic partner.
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.
5W AI Communications (5WPR) is relocating its headquarters to Florida. The firm already has 35 people on the ground across Miami and Tampa, with additional NY team relocating. Five GEO & PR Coordinator openings live now.
Ryan Trahan built a top YouTube channel and a DTC brand in parallel. The Penny Series, Joggy, and the series-based creator model.
Anthropic runs the most distinctive communications operation in frontier AI. Six primitives — the model spec, Constitutional AI, RSP, system cards, the Amodei sibling division, the long-form essay. EPR's audit.
EPR publishes the data every week.
Free. Weekly. Unsubscribe anytime.