How AI Search Is Recommending Business Checking Accounts: Monthly Trends

Executive Summary

Bluevine became the coverage leader in October 2026 at 53.2% valid recommendation coverage, moving ahead of the July 2026 baseline leader as Bank of America's tracked entity ceased reporting. Mercury holds second at 48.5%, a 4.7-point gap to the leader, and Relay sits third at 44.2%. The category's competitive set expanded sharply to 51 tracked brands from 10 in September 2026, reintroducing a long tail of community banks and credit unions that had dropped out of the qualified set in August and September.

The strongest upward mover among continuing entities was Relay, up 23.8 points to 44.2% from 20.4% in July 2026, a significant rise across the series, with top-three recommendations in 24.1% of qualified observations and 12 rank-one placements. Axos Bank rose 12.0 points to 32.4% from 20.4%, also significant against baseline, holding 60 top-three placements and 16 rank-one placements. Bluevine, Mercury, Found, Capital One, U.S. Bancorp, Wells Fargo & Co., Current, Citi, and Bank of America Corp. also registered significant rises against the July baseline.

The sharpest decliner was the entity tracked as Bank of America, down 43.9 points to 0.0% from 43.9% in July 2026, a significant drop that reflects its replacement in the tracked set by Bank of America Corp., which entered at 18.0%. The July-entity decliners also included Wells Fargo, down 31.8 points to 0.0%, and U.S. Bank, down 40.2 points to 0.0%, both replaced by successor entities at 11.1% and 21.8% respectively. Among continuing entities, Truist Bank fell 4.5 points to 3.5% from 8.0%, a significant decline against baseline, while PNC Bank eased 4.6 points to 12.5% from 17.1%, a move short of the significance threshold.

October reverses the shrinking-denominator pattern of the prior three months. Qualified observations rose to 423 from 144 in September 2026, and qualified surface breadth held at 6 canonical AI surface families. The category summary is mixed: eleven significant risers and six significant decliners measured against July, with no watch items flagged.

Each monthly run begins with 800 prompt-surface observations across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor; 478 were relevant in October 2026 against 145 in September and 273 in July, with 322 filtered out as irrelevant in October versus 655 in September and 527 in July. The public metrics use the 423 qualified observations from October 2026, 144 from September 2026, and 264 from July 2026 that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which brands and products triggered the high-severity factual inconsistencies?
• What kinds of claims, such as APY figures or sole-proprietor eligibility, did AI platforms contradict each other on?
• How did platforms cite sources differently when giving conflicting answers about the same product?

The benchmark detected 10 critical or high-severity factual inconsistencies across 5 AI platforms (ChatGPT, Copilot, Gemini, Google AI Mode, and Perplexity). All ten involve conflicting responses about pricing or eligibility, and all were assessed at high confidence.

Bluevine

Eight of the ten inconsistencies concern Bluevine Business Checking, and every one involves the advertised annual percentage yield (APY). AI platforms provided conflicting information about the same product across nearly every surface tested.

When asked "What is the best small business bank account?", ChatGPT stated Bluevine offers "up to 1.3% APY on qualifying balances up to $250,000," citing Bluevine's own support pages and NerdWallet, while Copilot stated "up to 3.7% APY, FDIC coverage up to $3M," citing Forbes Advisor, The College Investor, and SuperMoney. A second conflict on the same question saw Perplexity state the advertised yield can be "up to 3.00%," citing Axos Bank and Bluevine, while Copilot again stated "up to 3.7% APY." Source pages flagged in the analysis included two Beancount.io posts describing APY as "1.5% (Standard) up to 3.7% (Premier)" and "1.3% Standard, 1.75% Plus, 3.0% Premier."

When asked "Which bank is best for an online business account?", Google AI Mode stated "1.3%–3.0% APY (tiered)," citing Bluevine, Airwallex, and TowneBank, while Copilot stated "Up to 1.3% APY on balances," citing CNBC Select, TechRepublic, and Fit Small Business. The analysis flagged Bluevine's own business checking page as supporting both sides, carrying the excerpt "Get 1.3% APY with our Standard plan if you meet a monthly activity goal. Or, earn up to 3.0% APY with an upgraded plan."

When asked "What are the best online business checking accounts?", ChatGPT stated "1.30% APY on eligible checking balances up to $250,000," citing NerdWallet, Bluevine, and Finder, while Copilot stated a "1.5%–3.7% APY" range, citing Business Insider, Beancount.io, and other aggregators. A separate conflict on "Which is the best business bank account for a small business?" saw Gemini state "Up to 1.30% APY on the free tier if qualifying activity is met, and higher on paid plans," citing Relay, Wise, and Found, while Copilot stated "Bluevine Premier offers 3.00% APY on balances up to $3M," citing WalletHub, Rho, and Xero.

When asked "Which bank is best for an online business account?", Google AI Mode stated "1.3%–3.0% APY (tiered)," citing Bluevine's own checking page, while Copilot stated "Up to 1.3% APY on balances," citing CNBC Select, TechRepublic, and Fit Small Business. When asked "What is the best online bank for small business?", ChatGPT stated "1.3% APY on up to $250,000," citing Bluevine and Mercury, while Copilot stated "up to 3.7% APY," citing Beancount.io and Business Insider. When asked "Which Bank is best for an LLC business account?", ChatGPT stated "1.30% interest on qualifying balances up to $250,000," citing Fit Small Business and NerdWallet, while Copilot stated "Up to 3.7% APY," citing Beancount.io and StartupWise.

Mercury

Two inconsistencies concern Mercury, both about eligibility for sole proprietorships without a formal entity. AI platforms provided conflicting information about whether Mercury permits sole proprietors to open accounts.

When asked "Can I open a business bank account online without an LLC?", Gemini stated "Mercury routinely allows sole proprietors to open accounts entirely online using an SSN or EIN," citing Reddit, Baselane, and Wise, while Google AI Mode stated "Mercury typically requires formal incorporation documents like an LLC or Corp," citing Reddit, Business.com, and Wise. Both responses cited Wise's guide on opening a business bank account without an LLC.

When asked "What's the best online business bank account?", ChatGPT recommended Mercury for "startups/online companies" without excluding sole proprietorships, citing Bluevine and Mercury, while Google AI Mode stated "Mercury does not permit sole proprietorships," citing Airwallex, NerdWallet, and Bankrate.

Benchmark Context

Questions This Section Answers

• How did the qualified observation count and recommendation-shaped answer share change between July and October 2026?
• What does the difference between the raw collection universe and the qualified benchmark set mean for reading brand-level percentages?

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set.

Benchmark-Level Metrics

The following benchmark-level metrics summarize how the qualified prompt mix changed between the two months.

The qualified observation count reversed its three-month decline, rising to 423 in October 2026 after falling to 197 in August and 144 in September. The recommendation-shaped answer share more than doubled its July level to 70.9%, and the valid recommendation shortlist share reached 82.7%, its highest point in the series. The expansion pulled the long tail of community banks and credit unions back into the tracked set, which had narrowed to just 10 entities in August and September.

AI Recommendation Trend

Questions This Section Answers

• Which brands led the October 2026 coverage ranking, and how tight was the top three?
• Which neobanks and incumbents made the largest significant moves against the July baseline?

The Category Repriced After a Two-Month Compression, With a Long Tail Returning

October 2026 marks a structural reset rather than a simple leadership shift. Bluevine leads at 53.2% valid recommendation coverage, Mercury follows at 48.5%, and Relay sits at 44.2%, a tightly clustered top three separated by 9.0 points. The business-neobank cohort (Bluevine, Mercury, Relay, Found, Axos Bank) now occupies five of the top six coverage positions, with Capital One the highest-placed incumbent at 36.9%.

Eleven brands exceeded normal month-to-month variation against the July baseline. The category-level change was driven by the combination of several larger movements (Relay, Mercury, Capital One, Bluevine, U.S. Bancorp, Axos Bank) and the addition of new entrants, not by any single brand's shift.

What Changed This Month

Bluevine

Bluevine rose to 53.2% valid recommendation coverage in October 2026, up 22.9 points from 30.3% in July 2026, a significant gain across the series and the strongest movement of any continuing entity. The brand now leads the category, 4.7 points ahead of Mercury.

Raw mention presence rose 21.1 points to 53.7% from 32.6% in July. The top-three rate climbed 25.5 points to 45.9%, and the rank-one rate rose 12.2 points to 27.0% (114 rank-one placements, up from 39 in July). Valid recommendations reached 225 of 423 qualified observations, up from 80 of 264.

Bluevine's move combines both broader presence and stronger placement, a rarer pattern than coverage gains that come at lower list positions. The brand appears in 227 of 423 qualified observations and converts more than half of them into valid recommendations. The eight APY inconsistencies reported above sit directly on Bluevine's most-recommended product, which is the distinction the reader should hold alongside the coverage figure.

Highest-priority diagnostic: Which prompt clusters drove Bluevine's rank-one surge, and how do the flagged APY source pages map to the surfaces where Bluevine leads?

Mercury

Mercury holds 48.5% valid recommendation coverage in October 2026, up 48.5 points from its July 2026 baseline of 0.0%, when the brand was not in the tracked set. This is a debut-driven gain, not an organic month-over-month move, and the brand entered the comparison series only in August.

Raw mention presence was 49.4%, with a 28.1% top-three rate and a 2.8% rank-one rate (12 rank-one placements). Valid recommendations reached 205 of 423 qualified observations, second only to Bluevine in absolute count.

Mercury's profile shows broad recommendation coverage alongside weak top-of-list placement. The brand carries 48.5% coverage but only a 2.8% rank-one rate, a wider presence-to-placement gap than Bluevine, Relay, or Capital One. The two eligibility inconsistencies reported above sit on questions about sole proprietorships without an LLC.

Highest-priority diagnostic: Which prompt types generated Mercury's 205 valid recommendations, and why did rank-one placements lag despite near-parity coverage with the leader?

Relay

Relay rose to 44.2% valid recommendation coverage in October 2026, up 23.8 points from 20.4% in July 2026, the largest baseline-to-current rise among continuing entities. The brand also jumped 44.2 points from September 2026, when it recorded no valid recommendations in the compressed 10-entity set.

Raw mention presence rose 23.5 points to 45.1% from 21.6% in July. The top-three rate climbed 14.3 points to 24.1%, and the rank-one rate rose to 2.8% (12 rank-one placements). Valid recommendations were 187 of 423 qualified observations.

Relay's October profile shows a brand that regained both presence and placement after a two-month absence from the qualified set. The reinstatement is a set change as much as a measured rise, and the reader should hold the July baseline comparison as debut-resumption rather than continuous growth.

Highest-priority diagnostic: Which prompts returned Relay to the qualified set, and does its 24.1% top-three rate cluster on specific commercial-intent questions?

Axos Bank

Axos Bank rose to 32.4% valid recommendation coverage in October 2026, up 12.0 points from 20.4% in July 2026 and up 15.7 points from September 2026, both significant moves. The brand had fallen in September, dropping below its July baseline, before this October recovery.

Raw mention presence rose 9.6 points to 33.1% from 23.5% in July. The top-three rate climbed 6.2 points to 14.2%, and the rank-one rate rose 3.0 points to 3.8% (16 rank-one placements). Valid recommendations were 137 of 423 qualified observations.

Axos Bank's October gain reverses the September decline that had taken the brand below its July baseline. The recovery is broad-based, spanning presence, top-three, and rank-one placement.

Highest-priority diagnostic: Which surfaces drove the September decline and the October recovery, and is the rebound concentrated on the same prompt types that had previously delivered recommendations?

Capital One

Capital One enters the tracked set in October 2026 at 36.9% valid recommendation coverage, up 36.9 points from its July baseline of 0.0%, when the brand was not tracked under this entity name. Its predecessor entity, Capital One Auto Finance, recorded 3.4% in July and 0.0% in October.

Raw mention presence was 55.3%, the highest of any brand in the set. The top-three rate was 19.6% and the rank-one rate was 7.6% (32 rank-one placements). Valid recommendations were 156 of 423 qualified observations.

Capital One's October profile shows the widest presence-to-coverage gap among the top five, at 18.4 points. The brand appears in more than half of qualified observations but converts to a 36.9% coverage rate. As with Mercury, this is a debut figure rather than a continuing movement.

Highest-priority diagnostic: Which prompts mention Capital One without recommending it, and which competitor takes the recommendation in those answers?

Bank of America, U.S. Bank, and Wells Fargo

Three brands recorded significant declines at 0.0% in October 2026: Bank of America, down 43.9 points from 43.9% in July, U.S. Bank, down 40.2 points from 40.2%, and Wells Fargo, down 31.8 points from 31.8%. None recorded valid recommendations in the current month.

Successor entities entered the tracked set at the same time: Bank of America Corp. at 18.0%, U.S. Bancorp at 21.8%, and Wells Fargo & Co. at 11.1%. Each is a first-appearance entry rather than a continuing series.

These declines are set changes, not measured declines in brand quality. The July entities were replaced by successor entities in the tracked list for October. The percentages are not directly comparable across the entity boundaries, and the reader should read each pair (Bank of America and Bank of America Corp., U.S. Bank and U.S. Bancorp, Wells Fargo and Wells Fargo & Co.) as a tracking change rather than brand movement.

Highest-priority diagnostic: Do the successor entities' October figures cover the same prompt and surface distribution as the July entities, or has entity resolution altered the addressable question set?

PNC Bank

PNC Bank eased to 12.5% valid recommendation coverage in October 2026, down 4.6 points from 17.1% in July 2026, a move below the significance threshold. The brand also fell 5.6 points from September 2026.

Raw mention presence fell 11.4 points to 29.1% from 40.5% in July, and the top-three rate fell 5.0 points to 3.3% from 8.3%. The rank-one rate eased 1.6 points to 0.7% (3 rank-one placements). Valid recommendations were 53 of 423 qualified observations.

PNC Bank's October profile shows a category leader from earlier in the series now losing both presence and top-three placement while its coverage percentage holds just inside the mid-tier. The presence decline is the sharper supporting signal.

Highest-priority diagnostic: Which prompts dropped PNC Bank from the qualifying set, and does the decline correlate with the return of the neobank cohort that had been absent in August and September?

Truist Bank

Truist Bank fell to 3.5% valid recommendation coverage in October 2026, down 4.5 points from 8.0% in July 2026, a significant decline. The brand rose 3.5 points from September, when it held 0.0%.

Raw mention presence fell 10.7 points to 16.6% from 27.3% in July. The top-three rate was 0.7% and the rank-one rate was 0.2% (1 rank-one placement). Sentiment was the lowest of any continuing brand at 0.2.

Truist Bank holds 3.5% coverage on 15 valid recommendations against 423 qualified observations, with only a single rank-one placement. The brand's placement quality and sentiment both sit at the bottom of the continuing set.

Highest-priority diagnostic: Which prompts produced Truist Bank's 15 valid recommendations, and does the low sentiment score cluster on a specific question type?

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent cluster did the October qualified observations fall into?
• What can the benchmark not yet say about pricing, fee, or head-to-head comparison questions?

The October 2026 qualified observations fell entirely into the Brand Recommendation cluster, matching the August and September pattern. All 423 qualified observations were classified under this cluster, with the Pricing & Value and Multi-Brand Comparison clusters recording zero observations. The public benchmark measures which brands receive recommendation credit in direct choice questions, but it cannot yet answer questions about price positioning, fee competitiveness, or head-to-head comparison outcomes, even though pricing inconsistencies were detected on individual surfaces.

Brand Opportunity Summary

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations (query, surface, recommendation outcome, rank, sentiment, and citations where exposed). Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Source presence is not automatically treated as proof of causation.

This report is part of the broader AI Visibility Industry Market research program:

• AI Visibility Industry Market Methodology
• AI Visibility Industry Market Metrics
• AI Visibility Industry Market Standards

Interpretation Notes

• The qualified observation count reversed its three-month decline, rising to 423 in October 2026 from 144 in September, 197 in August, and 264 in July. Brand-level percentages are calculated within each month's qualified set, so points cannot be read as brand movement without accounting for the changing denominator.
• Fifteen brands that recorded measurable coverage in July 2026 fell to 0.0% by October 2026. Of those, three (Bank of America, U.S. Bank, Wells Fargo) were replaced by successor entities with new coverage at 18.0%, 21.8%, and 11.1% respectively. The remaining declines reflect set changes as the tracked list evolved, not measured declines in brand quality.
• The August and September qualified sets narrowed to 10 entities before reopening to 51 in October. Brands that entered in August or September but are absent in October cannot be compared on the same basis as continuing entities.
• The October 2026 qualified observations fell entirely into the Brand Recommendation cluster, matching August and September. Pricing and comparison questions were not present in the qualified set, so category-level pricing and head-to-head conclusions cannot be drawn even though pricing inconsistencies were flagged at the individual surface level.
• Movement between months identifies changes worth investigating. It does not by itself establish the cause of those changes.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

The aggregate percentages in this report answer where. Bluevine leads at 53.2%, Mercury follows at 48.5%, and Relay at 44.2%. The neobank cohort regained the qualified set, the incumbent entities reset under new names, and a long tail of community banks and credit unions returned. Ten factual inconsistencies across five AI platforms all concern two brands' most-recommended products. Beneath those figures sit the questions that matter for strategy: which high-intent prompts each brand is winning, which competitor gains the recommendation when a brand loses it, what attributes AI systems associate with each option, and which external sources shape those answers.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. The public benchmark identifies where attention is warranted; the audit explains why and what to do about it.

Request an AI visibility audit