How AI Search Is Recommending Personal Loans and Online Lenders: Monthly Trends

Executive Summary

SoFi regained the category lead in September 2026 after a two-month slide, while a benchmark-tracked rebranding created a distinct competitive story at the bottom of the field. SoFi's valid recommendation coverage rose 4.0 points from 64.5% in August to 68.5% in September, reclaiming first place from Upstart, which held the leadership position in August. The current gap between SoFi and second-place Upstart is 0.8 points, making this the tightest top-two spread in the three-month series.

The strongest upward mover was Achieve, which entered the tracked set with 15.6% valid recommendation coverage in September. This brand replaced "Achieve Home Loans" in the tracked roster, a name change rather than a new market entrant. Achieve Home Loans had declined for two consecutive months, from 8.3% in July to 4.6% in August, and recorded 0.0% coverage in September under the old tracking name. The benchmark treats the rebranded entity as a significant riser with a 15.6-point gain from baseline.

The sharpest decliners were Upgrade, down 6.8 points from July's 71.9% to 65.1% in September, and Happy Money, down 6.7 points from 17.0% to 10.3%. SoFi also recorded a significant decline from baseline, down 5.6 points from 74.1% in July, despite its month-over-month recovery. The category saw four brands classified as significant decliners against baseline, one significant riser, and a narrowing at the top that leaves SoFi and Upstart separated by less than a point.

Each monthly run begins with 800 prompt-surface observations (654 unique questions in September, 628 in August, 583 in July) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor; 700 were relevant and 100 were irrelevant in September, versus 738 relevant and 62 irrelevant in August, and 723 relevant and 77 irrelevant in July. The public metrics use the 585 qualified observations in September, 611 in August, and 590 in July 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 qualification funnel shifted between July and September: the number of unique questions rose from 583 to 654, while the share of prompts filtered as irrelevant grew from 77 to 100. The qualified observation count dipped from 590 to 585, holding steady despite the larger raw question set. The metrics below summarize the category-wide qualified benchmark before the report turns to brand-level results.

Benchmark-Level Metrics

August functioned as an intermediate dip for the category leader: SoFi fell to 64.5% coverage that month before recovering to 68.5% in September. The valid recommendation shortlist share declined 6.4 points from July to September, indicating that a smaller share of qualified observations produced usable recommendation shortlists by the current month.

AI Recommendation Trend

Two Leaders Nearly Tied as the Field Splits Between Recovering and Declining Brands

The category-level movement combined one significant brand-name transition with broad declines among the top-tier brands. Best Egg was the only brand to rise from baseline without a tracking-name change, moving up 2.0 points within normal month-to-month variation, while SoFi and Upstart both remain below their July levels even as SoFi regained the top position.

What Changed This Month

Achieve and Achieve Home Loans

The headline movement this month is a tracking transition. Achieve Home Loans held 8.3% valid recommendation coverage in July and 4.6% in August, then recorded 0.0% in September as the benchmark switched to tracking "Achieve." The rebranded entity posted 15.6% coverage in September, with 91 valid recommendations out of the qualified set.

Achieve's raw mention presence came in at 20.7%, compared with 10.7% for Achieve Home Loans at the July baseline. Its recommended top-3 rate was 2.1% and its rank-one rate 0.2%, with one rank-one placement in the qualified set.

The distinction the reader should notice is that this is a measurement-name change layered on real movement: the underlying brand gained both presence and recommendation coverage, but the benchmark cannot fully separate the rebranding effect from organic growth in this single month.

Highest-priority diagnostic: Which prompts surfaced "Achieve" versus "Achieve Home Loans," and which surfaces drove the presence gain from the July baseline?

SoFi

SoFi regained the category lead in September with 68.5% valid recommendation coverage, up 4.0 points from 64.5% in August. Against the July baseline of 74.1%, the brand remains down 5.6 points, a significant decline over the full series despite the month-over-month recovery.

SoFi's recommended top-3 rate fell 8.5 points from 58.8% in July to 50.3% in September, a significant decline. Its rank-one rate moved in the opposite direction, rising 2.3 points from 33.9% to 36.2%, the highest rank-one rate in the tracked set. Its raw mention presence held essentially flat at 89.6%, up 1.3 points from 88.3% in July.

The pattern to notice is that SoFi lost top-3 placements while gaining rank-one placements, meaning when it is recommended, it is more often recommended first, but it appears in fewer top-three lists overall.

Highest-priority diagnostic: Which prompt types shifted SoFi out of top-3 lists while its rank-one placements held or grew?

Upgrade

Upgrade recorded the sharpest baseline-to-current decline in the category, with valid recommendation coverage falling 6.8 points from 71.9% in July to 65.1% in September. Its recommended top-3 rate fell 8.9 points from 41.4% to 32.5%, a significant decline that outpaces the coverage drop.

Upgrade's raw mention presence declined 2.9 points to 82.2%, while its rank-one rate dipped 1.1 points to 7.7%. The brand moved from second place in the August benchmark to third in September.

The distinction here is that Upgrade's decline concentrated in top-3 placement rather than raw visibility: the brand remains present in most AI answers, but it is recommended in a top-three slot less often than it was in July.

Highest-priority diagnostic: Which competitor captured the top-3 placements Upgrade lost, and which prompt clusters drove the shift?

Happy Money

Happy Money declined 6.7 points from 17.0% valid recommendation coverage in July to 10.3% in September, a significant drop. Its raw mention presence fell 6.3 points from 18.3% to 12.0%, also significant, meaning the brand is both less visible and less recommended.

Happy Money's valid recommendation count fell from 100 in July to 60 in September out of the qualified set. Its recommended top-3 rate declined 1.0 point to 2.6%, while its rank-one rate held essentially flat at 0.3%.

The pattern is a broad retreat: Happy Money lost presence and coverage together, and its small absolute counts mean the percentage movements carry more caution than those of the top-tier brands.

Highest-priority diagnostic: Which surfaces or prompt types account for the presence decline, and is the loss concentrated in specific AI platforms?

Best Egg

Best Egg was the only brand without a name change to rise from baseline, moving up 2.0 points from 28.1% valid recommendation coverage in July to 30.1% in September, a change within normal month-to-month variation. Its recommended top-3 rate rose 2.3 points to 8.9%, while its rank-one rate dipped 0.2 points to 1.5%.

Best Egg's raw mention presence was essentially flat at 41.2%, down 0.3 points from 41.5% in July. Its valid recommendation count rose from 166 in July to 176 in September.

The distinction is that Best Egg gained recommendation coverage without gaining presence, suggesting the improvement came from stronger placement within existing visibility rather than broader exposure.

Highest-priority diagnostic: Which prompt types drove the top-3 improvement, and does the pattern hold across AI surfaces?

Buyer-Intent Interpretation

The public benchmark did not surface qualified observations in the pricing and value or multi-brand comparison clusters, meaning the current data cannot yet answer questions about how AI systems characterize rates, fees, or head-to-head trade-offs between lenders.

This concentration means the benchmark currently measures which brand an AI recommends for a general loan-seeking prompt, not how that recommendation changes when cost or comparison factors enter the question. Those commercial questions remain open until qualified observations appear in the other two clusters. The September funnel did show rising pricing-analysis responses (18, up from 4 in July), but none qualified into the tracked benchmark set.

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

• Small-count movements: Happy Money (60 valid recommendations), Achieve (91), and Best Egg (176) have smaller denominators than the top-tier brands, so their movements warrant more caution.
• Tracking-name transition: Achieve replaced Achieve Home Loans in the tracked roster in September; the two rows describe the same underlying brand under different names.
• Qualified denominator vs raw collection: All percentages use the 585 qualified observations in September, 611 in August, and 590 in July, not the 800 raw prompt surfaces.
• Month-over-month movement 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.

Beneath the aggregate percentage sit questions the benchmark alone cannot answer: which high-intent prompts are being won, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each lender, and which external sources shape those answers. These patterns determine whether a coverage decline reflects a broader repositioning or a prompt-specific gap.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy, turning the benchmark's category-level signals into actionable intelligence.

The public percentage cannot identify the prompts, competitors, or sources causing the result.

Request an AI visibility audit