How AI Search Is Recommending Online Car Buying Sites: Monthly Trends

Executive Summary

CarMax holds the category lead in September 2026 with 86.0% valid recommendation coverage, a narrowing 0.6-point edge over Carvana at 85.4%. This is a leader shift under pressure: CarMax fell 4.9 points from its 90.9% August level, while Carvana dropped 4.7 points from 90.1%. The top cluster that held through the summer has loosened, with the three former leaders — CarMax, Autotrader, and CarGurus — all posting significant declines from their July baselines.

The strongest riser this month was Lithia Motors / Driveway, which rebounded 9.8 points to 9.8% coverage from a zero August baseline, a significant reversal after the entity tracking change that had removed it from measurement entirely in August. No brand posted a significant upward move outside that tracking-related recovery.

The sharpest decliner is Edmunds, falling 12.8 points from 75.0% in July 2026 to 62.2% in September 2026, a significant two-month slide driven primarily by a 7.8-point drop in raw mention presence. This is the category's most persistent downward trend, with Edmunds ranking 6th and drifting further from the top group each month.

Research scope: Each monthly run begins with 800 prompt-surface observations across the benchmark's defined AI/search surface universe (516 unique questions in July 2026, 555 in August 2026, and 576 in September 2026). All 800 mentioned a tracked brand or competitor in each month. Of those, 794 were relevant and 6 irrelevant in July 2026; 781 were relevant and 19 irrelevant in August 2026; 793 were relevant and 7 irrelevant in September 2026. The public metrics use the 731 qualified observations for July 2026, 726 for August 2026, and 733 for 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.

August 2026 sat between these two months in the series, with 800 source prompts, 555 unique questions, 781 relevant and 19 irrelevant prompts, and 726 qualified observations. With the qualification funnel established, the following metrics summarize category-wide performance.

Benchmark-Level Metrics

The valid recommendation shortlist share fell from 91.5% in July 2026 to 85.1% in September 2026, even as the qualified observation count held steady. The response mix shifted toward factual answers, which rose from 38 to 84 observations, and ranked lists, which rose from 111 to 138, suggesting AI systems returned fewer direct shortlists and more explanatory or ordered content this month.

AI Recommendation Trend

The Leadership Group Is Declining Together, Compressing the Distance Between Tiers

The category's defining feature in September 2026 is not a single leader pulling away but a broad pullback across the top tier. CarMax, CarGurus, and Autotrader each posted significant declines from their July baselines, while Carvana's drop fell just inside the normal variation threshold. The result is a compressed top cluster where the leader's 86.0% coverage sits only 7.6 points above the 4th-ranked brand, Edmunds' decline has widened its gap from the leaders, and the middle tier of Cars.com and TrueCar has lost ground at a slower pace.

The category-level change came from the combination of six significant baseline-to-current declines rather than from any single brand's movement. Only CarsDirect and DriveTime moved upward, and both remained within normal variation. Edmunds' 12.8-point decline is the largest single movement in the category, but the leadership group's simultaneous pullback is the more consequential pattern for competitive positioning.

What Changed This Month

Edmunds Posts the Category's Steepest Two-Month Decline

Edmunds fell 12.8 points in valid recommendation coverage from 75.0% in July 2026 to 62.2% in September 2026, a significant decline spanning both months. The sharper single-month drop came in August (9.6 points), while September added a further 3.2-point decline that was not itself significant.

Edmunds' raw mention presence fell 7.8 points from 82.8% to 75.0%, with 550 of 733 observations mentioning the brand in September. Its valid recommendation count fell from 548 in July to 456 in September, and its top-3 placement rate dropped 5.8 points from 10.3% to 4.5%, with only 33 top-3 placements.

The distinction to notice is that Edmunds is losing at both the presence and placement stages simultaneously. Unlike a brand that remains visible but falls in recommendation rank, Edmunds appears in fewer answers and is placed lower when it does appear. This is a compound erosion rather than a single-stage problem.

Highest-priority diagnostic: Which prompts and AI surfaces account for the 7.8-point drop in Edmunds' mention presence, and which brands are taking the recommendation slots in the answers where Edmunds no longer appears?

Autotrader Falls to Fourth as Coverage Declines for a Second Month

Autotrader fell 6.9 points in valid recommendation coverage from 89.3% in July 2026 to 82.4% in September 2026, a significant decline with a 4.8-point drop in the final month. This two-month slide moved Autotrader from 2nd to 4th in the category ranking.

Autotrader's raw mention presence held essentially stable at 97.7%, with 716 of 733 observations mentioning the brand. Its top-3 placement rate, however, fell 9.1 points from 46.7% to 37.6%, while rank-one placement eased 1.6 points to 9.7%.

The distinction to notice is between presence and placement: Autotrader remains nearly universally visible, but AI systems are recommending it as a top-three choice far less often. The brand is losing ground at the recommendation stage, not the mention stage.

Highest-priority diagnostic: Which prompt categories drove the 9.1-point drop in Autotrader's top-3 placement, and which competitors are taking the top positions in the answers where Autotrader still appears?

CarMax Holds the Lead While Losing Ground

CarMax entered September 2026 with 86.0% valid recommendation coverage, down 4.6 points from 90.6% in July 2026, a significant decline. The leader posted a 4.9-point drop in the final month, the largest single-month move among the tracking-stable brands.

CarMax's raw mention presence actually rose 0.7 points to 98.2%, with 720 of 733 observations mentioning the brand. But its top-3 placement rate fell 3.7 points to 23.7%, and its valid recommendation count fell from 662 in July to 630 in September. CarMax retains the leadership position largely because its competitors declined alongside it.

The distinction to notice is that CarMax's coverage lead is a function of relative stability within a declining category. Its presence remains the most consistent in the benchmark, but its rank-one rate of 5.9% is among the weakest in the top tier, meaning it is recommended less often as the single best answer.

Highest-priority diagnostic: Which high-intent prompts are converting CarMax's near-universal presence into rank-one recommendations, and which competitors take the first slot when CarMax does not lead?

CarGurus Declines While Retaining the Strongest Top Placement

CarGurus fell 4.4 points in valid recommendation coverage from 89.1% in July 2026 to 84.7% in September 2026, a significant decline. The brand posted a 3.7-point drop in the final month that was also significant.

CarGurus' top-3 placement rate fell 6.3 points from 53.9% to 47.6%, still the highest in the category by a wide margin. Its rank-one rate eased 1.8 points to 24.3%, also the category lead. Its raw mention presence rose 1.0 point to 97.0%.

The distinction to notice is that CarGurus is declining from a position of placement strength. Even after the drop, it leads the category in top-3 rate by more than 10 points over the next brand and continues to record the highest rank-one rate as well, suggesting its recommendations remain prominent even as their breadth narrows.

Highest-priority diagnostic: Which prompt categories account for the 6.3-point drop in CarGurus' top-3 rate, and is the remaining placement strength concentrated in specific AI surfaces?

Carvana's Decline Is Sharper Month-Over-Month Than Across the Full Series

Carvana fell 3.2 points in valid recommendation coverage from 88.6% in July 2026 to 85.4% in September 2026, a decline that sits just inside the normal variation threshold. The sharper movement came in the final month: Carvana dropped 4.7 points from 90.1% in August 2026, a significant single-month decline.

Carvana's top-3 rate fell 4.8 points from 37.5% to 32.7%, and its rank-one rate fell 2.8 points to 15.8%, with 116 rank-one recommendations in September. Its valid recommendation count fell from 654 in August to 626 in September.

The distinction to notice is that Carvana's August gains were largely reversed in September. The brand had risen to a 90.1% coverage peak in August, only to give back most of that gain, leaving it 2nd in the category but with momentum pointing downward.

Highest-priority diagnostic: Which surfaces or prompt types drove Carvana's 4.7-point single-month drop, and did the decline concentrate in the same areas that produced its August rise?

Lithia Motors / Driveway Rebounds After the Tracking-Change Dip

Lithia Motors / Driveway rose 9.8 points to 9.8% valid recommendation coverage in September 2026 from a 0.0% August baseline, a significant recovery. The combined entity had been tracked at 16.1% in July 2026, dropped to 0.0% in August as Driveway was split out as a standalone brand, and now registers 9.8% for the combined entity.

Its raw mention presence rose 10.4 points in September to 10.4%, with 76 of 733 observations mentioning the combined entity. The brand recorded 72 valid recommendations, with a top-3 rate of 1.6% and no rank-one placements. Its net sentiment score of 0.99 was the highest in the category.

The distinction to notice is that this recovery follows a structural tracking change, not a purely competitive shift. August's zero baseline reflected the measurement split, and September's 9.8% reflects the combined entity regaining partial attribution. Driveway as a standalone brand registered 0.0% coverage in September after an 8.7% August appearance.

Highest-priority diagnostic: How should the September recovery of Lithia Motors / Driveway be reconciled with Driveway's standalone zero, and what is the accurate combined trajectory for the parent and subsidiary across the three months?

TrueCar Continues a Gradual Slide

TrueCar fell 7.0 points in valid recommendation coverage from 46.1% in July 2026 to 39.1% in September 2026, a significant decline across the two-month period. The final-month drop of 4.7 points was not itself significant, indicating a steady rather than abrupt erosion.

TrueCar's raw mention presence fell 6.0 points from 62.1% to 56.1%, with 411 of 733 observations mentioning the brand. Its valid recommendation count fell from 337 in July to 287 in September, and its rank-one rate fell 0.7 points to 0.5%, with only 4 rank-one placements.

The distinction to notice is that TrueCar's decline is broad-based rather than concentrated in a single stage. Presence, coverage, and top placement all moved downward together, suggesting a category-level repositioning away from the brand rather than a specific weakness in one type of AI answer.

Highest-priority diagnostic: Which prompts and competitor pairings account for TrueCar's 6.0-point presence decline, and is the brand being replaced by a specific competitor or by a broader shift in recommendation patterns?

Small-Count and Stable Brands Show Limited Movement

CarsDirect held at 2.2% coverage with 16 valid recommendations, and DriveTime moved within normal variation to 3.8% with 28 valid recommendations. Cars.com held essentially stable at 78.4% coverage, down from 81.7% in July but within the normal variation threshold.

Cars.com's top-3 rate actually rose 1.5 points to 19.0%, and its raw mention presence rose 0.6 points to 93.9%. The brand remains present in nearly all answers but is not converting that presence into top placement, with a rank-one rate of 2.3%.

These small-count movements should be interpreted with caution given the limited observations. The pattern across the lower tier is one of presence without placement: these brands appear in AI answers but are rarely positioned as the leading recommendation.

Highest-priority diagnostic: Are the modest presence and coverage patterns for CarsDirect and DriveTime stable across surfaces, or do they reflect isolated query patterns that could shift with the prompt mix?

Buyer-Intent Interpretation

In September 2026, all 733 qualified observations fell into the Brand Recommendation cluster, with zero observations classified under Pricing & Value or Multi-Brand Comparison. The public benchmark can currently speak to which online car buying sites are present and recommended in AI answers, but it cannot yet answer questions about how AI systems frame pricing differences, value trade-offs, or direct head-to-head comparisons between brands.

The response-type distribution does show that 112 of 733 observations were comparison analyses and 3 were pricing analyses, indicating the raw prompt set includes comparative and pricing-intent queries. These observations were not separated into distinct buyer-intent clusters in the qualified output, so the commercial questions about price and comparison remain open for company-level analysis.

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

• Category-wide pullback: Six of 11 tracked brands posted significant declines from their July 2026 baselines. The leader's coverage fell 4.6 points, and no brand posted a significant upward move outside the tracking-related Lithia Motors / Driveway recovery.
• Tracking structure changes: The Lithia Motors / Driveway recovery and Driveway's standalone zero reflect how the benchmark attributes recommendations between the parent and subsidiary brands. These movements should be read together with the August tracking change.
• Small-count caveats: CarsDirect (16 valid recommendations), DriveTime (28 valid recommendations), Driveway (0 valid recommendations), and Lithia Motors / Driveway (72 valid recommendations) have small observation counts. Their percentage movements are valid but warrant caution in interpretation.
• Qualified denominator: All brand-level percentages are calculated within the 733 qualified observations for September 2026, not the 800 raw prompts collected.
• Directional analysis: 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 percentages sit the questions that matter for competitive positioning: which high-intent prompts are won and lost, 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 September 2026 data shows a category-wide pullback with the leadership group declining together, but the public benchmark cannot explain the prompt-level dynamics behind these shifts or why presence is no longer translating into placement for several top brands.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. This moves from knowing a brand's coverage percentage to understanding the specific queries, surfaces, and evidence sources that drive it.

Request an AI visibility audit