How AI Search Is Recommending Gravel, Adventure and All-Terrain Bikes: Monthly Trends

Executive Summary

In October 2026, every one of the 613 qualified observations in the gravel, adventure, and all-terrain bike benchmark was classified as a factual answer. Recommendation-shaped answer share fell to 0.0% from 40.3% at the July 2026 baseline, and valid recommendation shortlist share fell to 0.0% from 62.9% over the same span. Because no observation this month carried a recommendation shape, valid recommendation coverage reads 0.0% for all ten tracked brands.

Trek enters the month as the series leader by baseline coverage at 63.5% in July 2026, with Specialized second at 62.8%, Giant third at 59.7%, and Cannondale fourth at 56.9% on the primary metric. The September 2026 readings placed Trek at 66.8%, Specialized at 65.9%, Giant at 61.2%, and Cannondale at 56.7%. Those were the last comparable recommendation-coverage values before October's answer shapes changed.

The series significance gate classifies eight brands as significant decliners against baseline this month: Cannondale, Cube Bikes, Giant, Marin Bikes, Orbea, Specialized, Surly Bikes, and Trek. Niner Bikes and Spot Brand classify as stable, since both already held minimal or zero coverage across the series. Every decliner moves from a positive baseline to zero, a pattern that describes a shared change in the category's output distribution this month rather than a shift in relative brand standing.

Against September 2026, the category posted no recommendation activity at all this month. The October run recorded 613 qualified observations, all classified as factual answers. The benchmark tracked ten companies; none registered presence, coverage, placement, or rank in October.

Each monthly run begins with 800 prompt-surface observations (539 unique questions in July 2026, 588 in August 2026, 572 in September 2026, and 580 in October 2026) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor in each month; July 2026 recorded 778 relevant and 22 irrelevant, August 2026 recorded 764 relevant and 36 irrelevant, September 2026 recorded 771 relevant and 29 irrelevant, and October 2026 recorded 772 relevant and 28 irrelevant. The public metrics use the 685 qualified observations in July 2026, 628 in August 2026, 624 in September 2026, and 613 in October 2026 that survive both qualification stages.

Key Findings

Benchmark Context

Questions This Section Answers

• How much did the qualified observation set change between July and October 2026?
• Did the benchmark's surface breadth and tracked company set stay consistent across the series?

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set. The October 2026 run produced no recommendation-shaped answers, so the recommendation metrics are empty rather than weak.

The qualified surface breadth held at six canonical AI surface families in both July 2026 and October 2026, and the company universe of ten tracked brands stayed the same across the series, so the two months remain measured against a consistent surface and company set.

Benchmark-Level Metrics

The series moved 685 (July 2026) to 628 (August 2026) to 624 (September 2026) to 613 (October 2026) qualified observations. August 2026 recorded the highest recommendation-shaped answer share at 43.8% and the highest valid recommendation shortlist share at 69.7%. September 2026 held near the July baseline. October 2026 recorded 613 factual answers and zero recommendations of any shape, the first month in the series with no recommendation signal.

AI Recommendation Trend

Questions This Section Answers

• Which brands were leading recommendation coverage before October 2026?
• Why did all ten tracked brands drop to zero coverage in October?

No recommendation leadership can be stated for October 2026, since every qualified observation this month was a factual answer rather than a recommendation

Eight of the ten tracked brands register a significant decline against baseline on the series significance gate; Niner Bikes and Spot Brand register as stable, since both already held minimal or zero coverage across the series. The decline for every brand coincides with the month's qualified observations consisting entirely of factual rather than recommendation-shaped answers. The benchmark records the output distribution for October; it cannot establish from this data alone whether the factual-answer pattern persists.

What Changed This Month

Questions This Section Answers

• What exactly changed in the October answer shapes compared to July?
• Did the prompt collection itself collapse, or only the recommendation layer?

The category-wide signal: answer shapes collapsed to factual answers only

Across the October 2026 run, all 613 qualified observations were classified as factual answers. Recommendation-shaped answer counts and shares both read zero, as did valid recommendation shortlist counts and shares. In July 2026, 276 of 685 qualified observations carried a recommendation shape (40.3%), and 431 carried a valid recommendation shortlist (62.9%). September 2026 held 249 recommendation-shaped answers (39.9%) and 403 valid shortlists (64.6%).

The benchmark recorded ten companies in both months. In October 2026, no company was present in any qualified observation, no company was recommended, and no company appeared in a top-three, top-ten, or rank-one position. Sentiment, placement, and coverage all read zero across the board.

The distinction to notice is that presence, sentiment, recommendation, and placement are separate signals, and only the recommendation layer moved to zero. The prompt collection itself did not collapse: 800 source prompt-surface observations were collected, 772 were relevant, and 613 qualified after both stages. The benchmark collected and qualified the material, but the AI surfaces answered factual questions rather than recommending products.

Highest-priority diagnostic: Whether the October answer shapes reflect a durable shift in how AI systems respond to this category's queries, or a one-month variation in the surface mix.

Cannondale: Steady baseline coverage, no recommendation activity in October

Cannondale recorded 56.9% valid recommendation coverage in July 2026, the fourth position among the four major tracked brands (Trek, Specialized, Giant, and Cannondale). The brand held 56.7% in September 2026 with 354 valid recommendations, and its raw mention presence stood at 92.2% in September. October 2026 recorded 0.0% coverage with zero valid recommendations.

Against the July 2026 baseline, coverage is down 56.9 points; against September 2026, down 56.7 points. Raw mention presence, top-three placement, rank-one rate, and net sentiment all fell to zero for the month.

The distinction to notice is that Cannondale's baseline coverage was carried by breadth of mentions, not by rank-one strength. Even in July, the brand's rank-one rate was 1.6%, meaning it appears frequently but rarely takes the top slot. Zero coverage in October removes the entire signal, both the breadth and the occasional top placement.

Highest-priority diagnostic: Which prompt and surface combinations produced Cannondale recommendations in September and whether any of those prompts reacquire a recommendation shape in November.

Trek and Specialized: The two strongest trackers both fall to zero

Trek recorded 63.5% valid recommendation coverage in July 2026 with 435 valid recommendations (top-three rate 36.2%, rank-one rate 15.0%, raw presence 99.7%). Specialized recorded 62.8% coverage in July with 430 valid recommendations (top-three rate 36.5%, rank-one rate 18.5%, raw presence 98.4%). In September 2026, Trek held 66.8% and Specialized 65.9%.

October 2026 records 0.0% coverage for both brands, with zero valid recommendations, zero top-three placements, and zero rank-one placements. Raw mention presence also reads zero for each.

The distinction to notice is that both brands were near-ceiling on presence for the entire series, and October removed that presence entirely. A category-level change in answer shapes is a more parsimonious reading than two leading brands independently losing every mention on the same run.

Highest-priority diagnostic: Whether the October surfaces answered the same prompts with factual descriptions instead of naming brands, or whether the prompt mix itself shifted toward specification-style questions.

Giant: Third traceable position also lost in October

Giant recorded 59.7% valid recommendation coverage in July 2026 with 409 valid recommendations and 663 present observations. September 2026 held 61.2% coverage with 382 valid recommendations.

October 2026 reads 0.0% coverage, zero valid recommendations, and zero presence across 613 qualified observations.

The distinction to notice is that Giant's coverage sat 5.4 points below Trek in July and 5.6 points below in September, a stable relative position across the series. October carries no relative information because the metric itself is empty.

Highest-priority diagnostic: Whether the surfaces that previously placed Giant in the top three revert to that behavior in the next run, and on what prompt types.

Orbea, Marin Bikes, Surly Bikes, and Cube Bikes: The climbing tier resets to zero

Orbea recorded 10.9% coverage in July 2026 and reached 18.9% by September 2026, the strongest upward move in the prior series, with 118 valid recommendations in September. Marin Bikes held 5.5% in July and 6.1% in September with 38 valid recommendations in September. Surly Bikes recorded 2.6% in July and 3.0% in September with 19 valid recommendations in September. Cube Bikes recorded 1.8% in July and 3.4% in September with 21 valid recommendations in September.

October 2026 reads 0.0% coverage for all four, with zero valid recommendations each.

The distinction to notice is that these brands rest on small absolute counts, and small counts are more easily affected by an answer-shape change than large ones. Orbea's September position represented 118 valid recommendations; Marin 38; Cube 21; Surly 19. When the recommendation class of answer disappears, the small-count brands lose all measurable signal immediately.

Highest-priority diagnostic: Whether the prompts that surfaced these brands earlier in the series were the ones answered differently in October, and whether any of those prompts retain factual-name mentions without recommendation framing.

Niner Bikes and Spot Brand: No movement against baseline

Niner Bikes recorded 0.1% coverage in July 2026 and 0.0% in September 2026 with zero valid recommendations. Spot Brand recorded 0.0% coverage in every prior month of the series with no presence recorded in any month.

October 2026 reads 0.0% coverage for both, matching the September pattern for Niner and the entire-series pattern for Spot.

The distinction to notice is that for both brands, the classification is stable, not decliner. Their baselines were already at or near zero, so October's zero does not represent a change relative to their own trajectory.

Highest-priority diagnostic: Whether either brand appears in any qualified observation during a month that restores recommendation-shaped answers, and under what prompt conditions.

Buyer-Intent Interpretation

Questions This Section Answers

• Did buyer-intent clusters like brand recommendation, pricing, or head-to-head comparison appear in October?
• What does the absence of recommendation-shaped answers mean for buyers exploring gravel and all-terrain bikes?

In October 2026, the 613 qualified observations were all classified as factual answers with no recommendation-shaped, comparison, or pricing component. No observation fell into the Brand Recommendation cluster in a way that supports a coverage reading, and no head-to-head or price-value comparison observations are recorded in this month's qualified set.

The public benchmark therefore cannot speak to brand recommendation, price positioning, or direct comparison outcomes for October 2026. Buyers exploring gravel, adventure, and all-terrain options received informational answers rather than direct recommendations during this run.

Brand Opportunity Summary

Questions This Section Answers

• Which brands need the most urgent diagnostic follow-up, and what question does each one raise?

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why each brand sits where it does.

Evidence Behind the Benchmark

Questions This Section Answers

• What kinds of prompt-level data can a company-level analysis examine when the aggregate metrics are empty?

The aggregate metrics are built from prompt-level observations covering the query, surface, recommendation outcome, rank, sentiment, and citations where exposed. Company-level analysis can go deeper into which prompts a brand wins, which competitor takes the recommendation when a brand loses, and which external sources shape those answers. Source presence is not automatically treated as proof of causation. The October 2026 run recorded no recommendation outcomes to analyze at the aggregate level; the underlying observations remain available for company-level examination.

About This Benchmark

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

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

Report-Specific Interpretation Notes

• October 2026 recorded zero recommendation-shaped answers across the full qualified set, which produces zero valid recommendation coverage for every tracked brand regardless of prior position.
• The qualified denominator has moved 685 (July 2026) to 628 (August 2026) to 624 (September 2026) to 613 (October 2026); percentages in each month are calculated within that month's own qualified set.
• 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 hide the detail that matters for action: 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. October 2026 shows the benchmark returning factual answers across the qualified set; the aggregate view cannot say which prompts those answers covered, which surfaces produced them, or what caused the shift from the prior months.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It answers the questions the public benchmark cannot, and turns the movement identified here into a concrete inspection plan.

Request an AI visibility audit