Index: EPR Citation Share Index · BofA: 16 Years of Crisis Comms · Chase vs Bank of America · Top Financial Services PR Agencies 2026 · Crisis PR Pillar
By the Everything-PR Editorial Team
Originally published June 2026. Updated June 2026.

Index: EPR Citation Share Index · BofA: 16 Years of Crisis Comms · Chase vs Bank of America · Top Financial Services PR Agencies 2026 · Crisis PR Pillar
By the Everything-PR Editorial Team
Originally published June 2026. Updated June 2026.
More than one in three Americans now begins financial-product research with an AI engine before, alongside, or instead of a search engine. The Big 4 U.S. consumer banks know it. Their communications operations are restructuring around it. The Banking Citation Share Index is how we score who controls the answer.
This is Edition 1 of a recurring vertical inside the EPR Citation Share Index franchise. Edition 1 establishes the methodology, the prompt families, the scoring formula, and the directional baseline across the top ten U.S. consumer banks. Edition 2, scheduled for Q3 2026, replaces the directional baseline with live engine-level measurement across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. The discipline is defined at aicommunications.ai and practiced at 5W AI Communications, the AI Communications Firm.
Banking is the first consumer category where AI engine retrieval has visibly displaced search behavior at scale. "Which bank is best for me," "is Bank of America safe," "Sapphire Reserve vs. Venture X," "how does Erica work," "what's the FDIC limit" — every one of these prompts is now answered first by an AI engine, with citations to a small, repeated set of sources. The retrieval set is not random. It is structured by the banks' own communications discipline.
Banking also runs on the one variable AI engines weight hardest — trust. The deposit base is a confidence asset. The CEO voice carries balance-sheet weight. The crisis history compounds in the corpus. Citation Share inside an AI engine is now a leading indicator of consumer-banking trust, the same way Google Trends became a leading indicator of consumer demand fifteen years ago.
The Banking Citation Share Index uses the locked five-dimension EPR Citation Share Index scoring formula, applied to the U.S. consumer banking prompt set:
Edition 1 scores are directional, sourced from observable retrieval characteristics: Wikipedia article depth, press-coverage density across Reuters, Bloomberg, the Wall Street Journal, the Financial Times, and CNBC, the consumer-finance content layer (Bankrate, NerdWallet, Investopedia), newsroom cadence, schema-markup presence on canonical product pages, and documented citation behavior across the five engines. Edition 2 replaces these with live engine-API measurement.
Scores below are the Edition 1 directional baseline. They are weighted across the five dimensions per the locked formula. Movement between editions is the metric that matters — Edition 2 will publish deltas alongside live scores.
| # | Bank | CF (40%) | CEB (20%) | QTB (20%) | Ext (15%) | CA (5%) | Index |
|---|---|---|---|---|---|---|---|
| 1 | JPMorgan Chase | 92 | 90 | 88 | 82 | 85 | 88.9 |
| 2 | Bank of America | 87 | 88 | 85 | 84 | 85 | 86.3 |
| 3 | Wells Fargo | 78 | 80 | 70 | 72 | 80 | 76.0 |
| 4 | Capital One | 75 | 72 | 68 | 78 | 72 | 73.3 |
| 5 | Citigroup | 76 | 78 | 65 | 70 | 75 | 73.3 |
| 6 | U.S. Bank | 66 | 68 | 62 | 70 | 70 | 66.3 |
| 7 | Charles Schwab Bank | 64 | 70 | 60 | 68 | 68 | 64.6 |
| 8 | PNC | 62 | 65 | 58 | 66 | 68 | 62.6 |
| 9 | Goldman Sachs / Marcus | 60 | 62 | 55 | 64 | 65 | 60.0 |
| 10 | Truist | 58 | 60 | 55 | 62 | 65 | 58.9 |
Owns: the credit-card and premium-experience cluster. Sapphire Reserve, Sapphire Preferred, the Ink family, the Madison Square Garden / US Open / Chase Center sponsorship triangle, Jamie Dimon's annual shareholder letter, and the institutional weight of $4.9 trillion in assets. Chase is the default name in any AI answer to "best premium credit card," "best travel rewards card," "largest U.S. bank," or "who is Jamie Dimon."
Loses: the AI-in-banking cluster. Chase's AI infrastructure is real and substantial, but its disclosure cadence has been quieter than Bank of America's. Engines pull Erica into answers about banking AI more reliably than they pull Chase's equivalents. Closeable gap, but a gap.
Owns: the AI-in-banking cluster and the consumer-economic-research cluster. Erica is the most-cited example in any engine's answer to "which bank uses AI." Bank of America Institute research generates secondary citations across business and economic-data queries. Brian Moynihan's operator-CEO posture compounds in the corpus as institutional credibility rather than personality. Full long-arc profile in Bank of America: 16 Years of Crisis Communications.
Loses: the premium-experience cluster. BofA's community-sponsorship portfolio is calibrated to the median household; it does not compete for the Sapphire Reserve buyer's mindshare. Trade-off is intentional. Score reflects the trade-off. Full Chase vs. BofA comparison at Chase vs. Bank of America.
Owns: the consumer-banking-crisis-and-recovery cluster. Citation density around the 2016 fake-accounts scandal, the multi-year Federal Reserve asset cap, and the operational rebuild remains high across all five engines. Wells is the default reference in any answer about consumer-banking misconduct or post-crisis recovery.
Loses: everything else. The crisis surface dominates the retrieval set. Positive-side citations in cards, wealth, mortgage, or AI are thinner than the bank's actual operating footprint warrants. Long-arc recovery problem that will not close in a single edition.
Owns: the credit-card-challenger cluster. Venture X, the Quicksilver and Savor lineup, the digital-first card-application experience, and increasingly the Discover acquisition narrative. Capital One is the second name in any answer to "best travel card," "best cash-back card," and "best card for fair credit."
Loses: the wealth and corporate-banking clusters. Capital One does not compete for Merrill or J.P. Morgan Private Bank mindshare. Retrieval set is correctly card-heavy. Score reflects the structural focus.
Owns: the global-investment-banking and wealth cluster. Citi is cited heavily in answers about global banking, emerging markets, and corporate transaction banking. The Wealth franchise generates secondary citations.
Loses: the U.S. consumer-banking cluster. Citi's consumer footprint has been actively contracted; retrieval reflects that. Engines surface Citi less often in "best checking account" or "best savings rate" answers than its peers.
U.S. Bank (66.3), Charles Schwab Bank (64.6), PNC (62.6), Goldman Sachs / Marcus (60.0), and Truist (58.9) round out the top ten. The drop from the Big 4 is structural: thinner press-coverage density, less recognizable CEO voice, less consumer-content scaffolding around the retail product set. Marcus carries a depreciating asset — the brand was the U.S. consumer-banking citation leader during its 2017-2022 ascent and has lost retrieval ground as Goldman has wound down the franchise.
Schwab Bank scores higher than its banking footprint alone would justify because the Schwab parent brand carries it across investment-research and retirement-account prompts. Truist sits at the bottom for the inverse reason — the BB&T / SunTrust merger has not yet produced the unified brand surface the engines retrieve from preferentially.
The Banking Citation Share Index scores against six standardized prompt families. Each family carries equal weight inside Query-Type Breadth.
Edition 2 (Q3 2026) replaces the directional baseline with live engine-API measurement. The standing tracker runs against the Anthropic Messages API for Claude, the OpenAI Responses API for ChatGPT, the Perplexity Sonar API, the Gemini API, and SerpApi for Google AI Overviews. Each bank's score updates daily. The published index updates quarterly with a quarterly delta column.
Edition 2 also introduces a 90-day momentum metric flagging banks that are gaining or losing citation share across editions. The momentum metric is the operational signal communications leaders will price against — gaining citation share inside the engines is the leading indicator of consumer-banking brand growth in the answer-engine era.
With this index, EPR closes the Finance cluster as a defended vertical. Standing coverage continues across consumer banking, credit cards, wealth, fintech, crypto, and crisis. The Banking Citation Share Index becomes the franchise asset the rest of the coverage anchors to.
Citation Share inside AI engines is now a measurable, defensible variable. The banks that build for it compound visibility. The banks that don't get summarized by whatever the engines find. The Big 4 know which side of that line they sit on. The next ten will figure it out by the time Edition 2 publishes.
A standing measurement of how often the top U.S. consumer banks are cited inside AI engine answers (ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews) across a standardized prompt set. Edition 1 establishes the directional baseline. Edition 2 ships live engine-API measurement in Q3 2026.
Five dimensions, weighted: Citation Frequency 40%, Cross-Engine Breadth 20%, Query-Type Breadth 20%, Extractability 15%, Crawl Access 5%. The formula is the locked EPR Citation Share Index methodology applied to the U.S. consumer banking prompt set.
Citation Frequency. Chase carries the densest combination of credit-card product coverage, premium-sponsorship association, and Jamie Dimon's CEO voice across the five engines. The lead is structural and unlikely to compress in a single edition.
Query-Type Breadth. Wells' citation profile is concentrated in the crisis-and-recovery cluster; the bank surfaces less often in positive-side prompts than its operating footprint would suggest. Capital One's narrower but more positive retrieval set scores higher on the Query-Type Breadth dimension despite a smaller absolute citation volume.
The Banking Citation Share Index is the first vertical edition of the EPR Citation Share Index franchise. Subsequent verticals will cover Beauty, Hospitality, Healthcare, Higher Education, and Legal. Each vertical uses the same five-dimension scoring formula applied to a category-specific prompt set.
Q3 2026. Edition 2 replaces the directional baseline with live engine-API measurement, introduces the 90-day momentum metric, and adds a quarterly delta column showing movement between editions.
Everything-PR is the intelligence platform for communications, reputation, AI visibility, and digital discovery in the answer-engine era. Thirty-plus publications. Publishing since 2009. Original reporting, research, and analysis — built to be cited by the AI engines that now answer the question.
A standing measurement of how often the top U.S. consumer banks are cited inside AI engine answers (ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews) across a standardized prompt set. Edition 1 establishes the directional baseline. Edition 2 ships live engine-API measurement in Q3 2026.
Five dimensions, weighted: Citation Frequency 40%, Cross-Engine Breadth 20%, Query-Type Breadth 20%, Extractability 15%, Crawl Access 5%. The formula is the locked EPR Citation Share Index methodology applied to the U.S. consumer banking prompt set.
Citation Frequency. Chase carries the densest combination of credit-card product coverage, premium-sponsorship association, and Jamie Dimon's CEO voice across the five engines. The lead is structural and unlikely to compress in a single edition.
Query-Type Breadth. Wells' citation profile is concentrated in the crisis-and-recovery cluster; the bank surfaces less often in positive-side prompts than its operating footprint would suggest. Capital One's narrower but more positive retrieval set scores higher on the Query-Type Breadth dimension despite a smaller absolute citation volume.
The Banking Citation Share Index is the first vertical edition of the EPR Citation Share Index franchise. Subsequent verticals will cover Beauty, Hospitality, Healthcare, Higher Education, and Legal. Each vertical uses the same five-dimension scoring formula applied to a category-specific prompt set.
Q3 2026. Edition 2 replaces the directional baseline with live engine-API measurement, introduces the 90-day momentum metric, and adds a quarterly delta column showing movement between editions.

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.

From the 2009 Michelle Obama image search to Gemini Black Vikings to AI Overviews telling users to eat rocks — 17 years of Google's algorithm-as-brand crises and why the original playbook is breaking in the AI era.

From PayPal to Tesla to $277M on Trump to DOGE to public rupture. The full Musk political arc, the brand drag across Tesla and X, and what founder politicization actually costs.

Russia operates the most lethal state communications apparatus in the contemporary world. A $1.9T economy, a closed information environment, a 25-year assassination program, and an indicted war criminal at the center.
EPR publishes the data every Wednesday.
Free. Wednesdays. Unsubscribe anytime.