How AI Search Is Recommending Small Business Loans: Monthly Trends

Executive Summary

The small business lending category saw the most consequential reshuffle of the three-month series in September 2026, driven by an instrument-level change in how two major banks are identified. Bluevine remains the category leader at 59.9% valid recommendation coverage, but that position now comes with a 7.6-point lead over onDeck at 52.3% — the narrowest leadership margin measured since the benchmark began in July 2026.

The defining movement this month is the disappearance of Bank of America and Chase from the tracked-brand set under their prior names, replaced by Bank of America Corp. at 18.5% and Chase Credit Journey at 0.2%. The benchmark classifies both prior entries as significant decliners with two-month downward streaks, but the mechanism is a naming-entity shift rather than a loss of AI recommendation share: Bank of America's coverage fell 43.2 points from August 2026, and Chase fell 36.2 points over the same period. Readers should treat these movements as an instrument break in brand tracking, not as evidence that either bank lost commercial ground.

Setting the banking-entity shift aside, the competitive story is the rise of the fintech tier. Fundbox climbed 10.2 points to 51.5% in a significant move, onDeck rose 10.8 points to 52.3%, and Lendio gained 7.4 points to 33.5% — all three now sit at or near their strongest coverage levels of the series. Bluevine remains on top but has declined 9.9 points from its July 2026 baseline of 69.8%, a significant two-month pullback that has tightened the entire field.

Each monthly run begins with 800 prompt-surface observations (565 unique questions in July 2026, 590 in August 2026, and 533 in September 2026) across the benchmark's defined AI/search surface universe. All 800 observations in each month mentioned a tracked brand or competitor; 476 were relevant and 324 were irrelevant in July 2026, 464 were relevant and 336 were irrelevant in August 2026, and 651 were relevant and 149 were irrelevant in September 2026. The public metrics use the 437 qualified observations from July 2026, 414 from August 2026, and 421 from September 2026 that survive both qualification stages.

Key Findings

Benchmark Context

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

The raw collection universe held steady at 800 observations in both baseline and current months, while the relevance mix shifted sharply: the September 2026 run qualified 651 relevant prompts versus 476 in July 2026, with irrelevance falling from 324 to 149. The qualified surface breadth held at six AI surface families across all three months, with the highest number of relevant prompts occurring in the current month.

Benchmark-Level Metrics

The benchmark operated on the same instrument version across the series, but the September 2026 tracking set renamed two entities: Bank of America became Bank of America Corp., and Chase became Chase Credit Journey. The August 2026 interim month saw valid recommendation shortlist share fall to 65.9% before recovering to 71.3% in September 2026.

AI Recommendation Trend

Leadership held, but the field tightened as the fintech tier surged

Bluevine retained first place, yet the competitive picture changed materially: its lead over onDeck narrowed from 23.8 points in July 2026 to 7.6 points in September 2026, and three brands now sit above 50% coverage compared with one at the start of the series.

The category-level movement in September 2026 came from a combination of several significant shifts: the two banking-entity replacements, the significant rises at Fundbox and onDeck, and Bluevine's significant two-month decline from baseline. Bank of America and Chase, which ranked 2nd and 3rd in July 2026, no longer appear as trackable entities under those names in the current set. Lendio's 7.4-point prior-month rise was also significant, though its baseline-to-current movement of 2.1 points sits within normal variation.

What Changed This Month

Bluevine: significant two-month decline from baseline

Bluevine's valid recommendation coverage fell from 69.8% in July 2026 to 59.9% in September 2026, a 9.9-point decline that the benchmark classifies as significant. The prior-month movement was a smaller 3.6-point drop from August 2026, within normal variation.

The top-three rate actually improved against baseline, rising from 39.4% to 40.4%, and the rank-one rate strengthened from 14.4% to 18.3%. Bluevine produced 77 rank-one appearances in September 2026, up from 63 in July 2026, even as its overall coverage narrowed.

The distinction is between recommendation breadth and recommendation depth. Bluevine wins more first-position placements than ever, but appears across fewer qualifying answers overall. Raw mention presence held relatively steady at 72.2% versus 77.3% in July 2026, a modest shift, meaning the brand remains highly discoverable. The coverage decline is concentrated in the middle of the recommendation set rather than at the top.

Highest-priority diagnostic: Which answer contexts still name Bluevine but no longer place it within the valid recommendation shortlist, and which brands now occupy those slots?

Bank of America and Chase: brand-entity tracking change, not coverage loss

Bank of America's coverage series reads 56.5% in July 2026, 43.2% in August 2026, and 0.0% in September 2026 under that name, with Chase following the same pattern at 49.0%, 36.2%, and 0.0%. The benchmark flags both as significant decliners with two-month downward streaks, but the September 2026 movement reflects a naming-entity change rather than a loss of AI recommendation share.

The banks now appear under new tracked identities: Bank of America Corp. at 18.5% valid recommendation coverage and Chase Credit Journey at 0.2%. Bank of America Corp. records a 28.5% raw mention presence rate, a 6.7% top-three rate, and a 1.9% rank-one rate in September 2026, with 78 valid recommendations. Chase Credit Journey appears in only 3 of 421 observations, with a single valid recommendation.

The commercial interpretation must wait for a stable tracking entity. The August-to-September movement from the old names to the new names is an instrument change, not evidence that either bank lost or gained AI recommendation ground. What the benchmark can say is that Bank of America Corp. enters the series as the 5th-ranked brand, well behind the four fintech leaders.

Highest-priority diagnostic: How does Bank of America Corp.'s 18.5% coverage under the new entity compare with Bank of America's 43.2% under the prior name, once equivalent prompt sets are matched?

Fundbox: significant rise to 51.5%

Fundbox's valid recommendation coverage rose from 39.8% in July 2026 to 51.5% in September 2026, an 11.7-point increase that the benchmark classifies as significant. The prior-month movement was a 10.2-point climb from August 2026, also significant, giving the brand a two-month upward streak.

Raw mention presence rose from 47.6% to 61.8%, a 14.2-point gain, and the top-three rate climbed from 21.1% to 30.2%. Fundbox recorded 36 rank-one appearances in September 2026 versus 29 in July 2026, with the rank-one rate moving from 6.6% to 8.6%.

The pattern is broad-based growth across presence, top-three placement, and rank-one frequency. Fundbox is being surfaced more often, recommended in the top three more often, and named first more often. Net sentiment held steady around 0.9, indicating the expansion is not accompanied by any shift in how the brand is framed.

Highest-priority diagnostic: Which lending scenarios joined Fundbox's recommendation set, and which competitors lost those placements?

onDeck: significant prior-month rise to 52.3%

onDeck's coverage moved from 46.0% in July 2026 to 52.3% in September 2026, a 6.3-point gain that the benchmark classifies as stable against baseline. The sharper story is the prior-month movement: onDeck rose 10.8 points from 41.5% in August 2026, a significant jump.

The supporting metrics show consistent strength. Raw mention presence rose from 54.0% to 65.6%, and the top-three rate climbed from 24.9% to 34.9%. The rank-one rate more than doubled against baseline, from 5.7% to 10.9%, taking onDeck from 25 to 46 first-position appearances.

The distinction is between the baseline comparison and the current momentum. onDeck's rise is concentrated in the most recent month, not spread evenly across the series. The brand now holds 2nd place at 52.3%, its strongest position of the three-month window.

Highest-priority diagnostic: Which surface or prompt category drove the August-to-September acceleration, and is the new coverage level durable across prompt types?

Lendio: significant prior-month recovery to 33.5%

Lendio's coverage sits at 33.5% in September 2026, up 2.1 points from the 31.4% July 2026 baseline, a movement the benchmark classifies as stable. The prior-month shift tells a different story: Lendio rose 7.4 points from 26.1% in August 2026, a significant recovery after two months of decline.

Raw mention presence rose from 39.6% to 47.3%, and the top-three rate climbed from 9.8% to 14.7%. Lendio recorded 62 top-three placements and 17 rank-one appearances in September 2026, up from 43 and 16 respectively in July 2026.

The August 2026 dip to 26.1% now reads as a single-month softness rather than a trend, given the September 2026 recovery. Lendio's rank-one frequency held essentially flat across the series while its top-three placement expanded significantly.

Highest-priority diagnostic: Which competitor lost top-three share to Lendio in September 2026, and which evidence sources supported the shift?

Notable gap: Bank of America and Fundbox fully converge and cross

The gap between Bank of America and Fundbox moved from 16.7 points in July 2026 to 1.9 points in August 2026 and then to negative 51.5 points in September 2026, as the banking entity change removed Bank of America from the comparable set. The series widened by 68.2 points across the three months, a category observation about competitive repositioning rather than a single brand's gain.

The same pattern appears in the Bank of America Corp.-Chase comparison, which moved from negative 49.0 points in July 2026 to positive 18.5 points in September 2026. These gap movements are artifacts of the tracking-entity change and should not be read as evidence about either bank's market position.

Highest-priority diagnostic: Once equivalent prompt sets can be matched across the entity change, which of the fintech leaders captured the coverage share previously attributed to the two banks?

Buyer-Intent Interpretation

All 421 qualified observations in September 2026 fell into the Brand Recommendation class. The benchmark found no qualified observations in the Pricing & Value or Multi-Brand Comparison classes, meaning pricing discussions and direct comparison requests did not surface in the qualified set.

The public benchmark cannot yet answer price sensitivity, value positioning, or head-to-head preference questions. The response-type distribution in September 2026 included 1 pricing analysis response and 105 comparison analysis responses, but those did not qualify into the separate buyer-intent clusters, leaving those commercial questions unanswered at the benchmark level.

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.

About This Benchmark

This report is part of the LLM Authority Index AI Market Discovery research program.

• AI Industry Market Discovery
• AI Industry Market Discovery Methodology
• AI Industry Market Discovery Metrics
• AI Industry Market Discovery Standards

Report-Specific Interpretation Notes

The September 2026 tracking set renamed two entities. Bank of America now appears as Bank of America Corp., and Chase appears as Chase Credit Journey. Movements between the old and new names reflect an instrument change in brand tracking, not commercial gains or losses. Any comparison involving these brands across the entity boundary should be treated as directional only.

Small-count movement is meaningful in this vertical. Brands like Funding Circle, Chase Credit Journey, QuickBridge, and Biz2Credit operate on valid recommendation counts in the single digits to low twenties; their percentage movements can appear volatile even when the underlying count changed by a handful of observations. Chase Credit Journey's 0.2% coverage represents a single valid recommendation out of 421 qualified observations.

The qualified denominator differs from the raw collection universe. All metrics in this report use the qualified benchmark observations (421 in September 2026), not the 800 raw prompt-surface observations. Directional analysis identifies movement worth investigating; it does not by itself establish cause, and no claim in this report should be read as attributing a brand's movement to any specific action by that brand or its partners.

Next Step

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

Beneath the aggregate percentage sit the questions that determine competitive position: which high-intent prompts a brand wins, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each option, and which external sources shape those answers.

The public percentage cannot identify the prompts, competitors, or sources causing the result. A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. Where the public benchmark shows movement, the audit identifies the mechanism and the path to remediation.

Request an AI visibility audit