How AI Search Is Recommending Short Term Health Insurance: Monthly Trends

Executive Summary

UnitedHealthcare (Golden Rule) remained the coverage leader in October 2026. Its valid recommendation coverage rose to 55.4%, up 4.1 points from the July 2026 baseline of 51.3%, a cumulative change that stays within normal month-to-month variation. The more notable move was within the month itself: coverage jumped 16.6 points from a September low of 38.8%, a significant month-over-month rebound that rebuilt the brand's margin over second-place Pivot Health to 12.9 points, up from a compressed 7.2-point margin in September.

eHealth was the one brand whose movement exceeded normal month-to-month variation on both counts. Its coverage climbed to 13.5% in October 2026 from 5.9% in September 2026, a significant 7.6-point increase that recovered most of the ground lost in a 9.6-point decline the prior month, though it did not return all the way to August's 15.5% level. Measured against the July 2026 baseline of 7.4%, eHealth is up 6.1 points, also a significant cumulative move. Pivot Health added 10.9 points from September 2026 to October 2026, reaching 42.5%, though against its July 2026 baseline it remains down 2.1 points, a change inside normal variation.

No brand recorded a significant decline from the July 2026 baseline this month. Everest moved up 5.6 points from September 2026 to October 2026, recovering to 32.6% coverage, though it remains 2.5 points below its July 2026 starting level. The category's aggregate movement this month came from a cluster of upward moves rather than a single dominant shift, and only one brand, eHealth, exceeded normal month-to-month variation on its own.

Each monthly run begins with 800 prompt-surface observations (624 unique questions in October 2026) across the benchmark's defined AI/search surface universe. Of those, 796 mentioned a tracked brand or competitor; 321 were relevant and 475 were irrelevant. The public metrics use the 193 qualified observations in October 2026 that survive both qualification stages. This compares with 271 qualified observations in July 2026, 264 in August 2026, and 237 in September 2026.

Key Findings

Benchmark Context

Questions This Section Answers

• How does the raw prompt collection narrow to the qualified benchmark set?
• Why is the qualified denominator smaller than the total collected prompts?

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

These two views describe the same monthly funnel from different angles. The research-stage table shows how each month's raw collection narrows to the public qualified set, while the benchmark-level metrics below summarize category-wide rates computed within that qualified set.

Benchmark-Level Metrics

The September 2026 benchmark recorded 237 qualified observations, and the qualified set contracted again to 193 in October 2026, the smallest of the four-month series. Despite the smaller denominator, the share of answers shaped as recommendations rose sharply, meaning a larger proportion of a smaller qualified set took recommendation form.

AI Recommendation Trend

Questions This Section Answers

• Which short term health insurance brands held or improved their recommendation coverage in October 2026?
• How did the leader's margin over the second-place brand change from September to October 2026?

UnitedHealthcare (Golden Rule) Holds the Lead as Several Brands Rebound From September

The leader held its position as several brands with material coverage rebounded from September lows. Pivot Health remained second and Everest third, with the margin between first and second rebounding to 12.9 points in October after narrowing to 7.2 points in September.

The category-level change came from the combination of several smaller upward movements layered on top of the leader's gain. Only eHealth exceeded normal month-to-month variation over the full baseline-to-current series; no other brand's baseline movement was classified as significant.

What Changed This Month

Questions This Section Answers

• Why did UnitedHealthcare (Golden Rule) rebound from its September coverage low?
• What drove eHealth's significant increase in recommendation coverage?
• Why did Pivot Health's presence fall even as its top-three placement held?

UnitedHealthcare (Golden Rule)

UnitedHealthcare (Golden Rule) lifted its valid recommendation coverage from 51.3% in July 2026 to 55.4% in October 2026, up 4.1 points. The sharper story is the month-to-month reversal: after falling to 38.8% in September 2026, the brand rose 16.6 points in October, a significant move against the prior month. That recovery rebuilt the lead over Pivot Health to 12.9 points from a September margin of 7.2 points.

Raw mention presence held essentially flat at 98.5% in October versus 99.3% in July, so the brand was already near-universal in visibility. The top-three rate rose 7.8 points from 43.5% to 51.3%, with 99 top-three placements in October, up from 79 in September. The rank-one rate held flat at 34.7% versus July, with 67 rank-one recommendations, up from 63 in September.

The distinction to notice is that the leader's gain came from converting its already-near-universal presence into higher placement, not from winning new visibility. The valid recommendation count rose to 107 in October from 92 in September. Its average recommended rank stood at 1.59.

Highest-priority diagnostic: Which prompt types and surfaces drove the October recovery in top-three placement after the September pullback?

eHealth

eHealth was the clear upward mover, lifting coverage from 7.4% in July 2026 to 13.5% in October 2026, up 6.1 points, a significant cumulative move. The month-to-month move is sharper still: coverage more than doubled from 5.9% in September 2026 to 13.5% in October, an increase of 7.6 points, also significant.

Raw mention presence slipped slightly from 49.5% in July to 47.1% in October, down 2.4 points, so the gain is not a visibility story. The top-three rate rose 8.7 points from 2.2% to 10.9%, with 21 top-three placements in October versus 1 in September. The rank-one rate rose 5.6 points from 2.2% to 7.8%, with 15 rank-one recommendations in October, up from none in September.

The distinction is that eHealth snapped back from visible-but-not-recommended to visible-and-recommended. The valid recommendation count climbed to 26 in October from 14 in September. Its average recommended rank was 1.61, among the strongest in the field.

Highest-priority diagnostic: Which prompts and surfaces account for the return of valid recommendations, and are those the same prompts the brand lost in September?

Pivot Health

Pivot Health slipped 2.1 points over the full series, from 44.6% coverage in July 2026 to 42.5% in October 2026, a move inside normal variation. The month-to-month story runs the other way: coverage rose 10.9 points from 31.6% in September 2026 to 42.5% in October, a significant prior-to-current move.

Raw mention presence fell 10.1 points from 59.8% in July to 49.7% in October. The top-three rate rose 4.5 points from 36.9% to 41.4%, with 80 top-three placements in October, up from 67 in September. The rank-one rate was essentially flat, up 0.5 points to 4.2%, with 8 rank-one recommendations.

The key distinction is a separation between presence and placement. The brand is being mentioned in fewer qualified answers than in July, yet it is converting a higher share of the answers where it appears into top-three placements. The valid recommendation count rose to 82 in October from 75 in September. Its average recommended rank was 2.11.

Highest-priority diagnostic: What is driving the multi-month decline in raw mention presence even as placement holds, and which prompts lost the mention?

Everest

Everest moved down 2.5 points over the series, from 35.1% coverage in July 2026 to 32.6% in October 2026, inside normal variation. The brand rose 5.6 points from September 2026 to October 2026, extending an upward streak to two consecutive months.

Raw mention presence fell 6.7 points from 46.1% in July to 39.4% in October. The top-three rate rose 2.0 points from 29.1% to 31.1%, with 60 top-three placements in October. The rank-one rate rose 6.3 points from 1.5% to 7.8%, with 15 rank-one recommendations in October versus 7 in September.

The distinction for Everest is that its ability to win the top spot has strengthened even as its overall presence softened. The valid recommendation count was 63 in October, flat against September. Its average recommended rank improved to 2.44 from 2.63.

Highest-priority diagnostic: Which surfaces are producing the rise in rank-one placements, and can that pattern be tied to specific prompt wording?

National General

National General rose from 8.5% coverage in July 2026 to 13.0% in October 2026, up 4.5 points. The month-to-month move was 4.1 points from 8.9% in September 2026 to 13.0% in October, both inside normal variation.

Raw mention presence rose 4.9 points from 12.2% to 17.1%. The top-three rate rose 1.4 points from 3.3% to 4.7%, with 9 top-three placements in October. The rank-one rate slipped 0.4 points to 0.0%, with no rank-one recommendations in October versus one in July. The valid recommendation count was 25 in October, flat against September. Average recommended rank was 3.64.

The distinction is that National General's growth is a coverage and presence gain, not a top-placement gain. It appears in more qualified answers but rarely reaches the first position.

Highest-priority diagnostic: Where is the incremental coverage coming from, and why does it not convert to top-three placement as often as peers with similar presence?

The Long Tail

IHC Group rose 0.9 points over the series to 1.6% coverage, extending an upward streak to three consecutive months, though the cumulative movement remains inside normal variation and is classified as a watch item. Raw mention presence fell 1.5 points to 2.6%, and the valid recommendation count was 3, unchanged from September.

Independence American rose 1.4 points to 2.1% coverage, with a valid recommendation count of 4 in October versus 3 in September. Companion Life rose 0.9 points to 1.6% coverage, with a valid recommendation count of 3, up from 2 in September. For all three brands the absolute counts are very small, so single observations can move their percentages materially.

LifeShield recorded 0.0% coverage in October 2026, matching its zero valid recommendation count in July 2026 and September 2026. Agile Health Insurance recorded 0.0% coverage for a fourth consecutive month.

Highest-priority diagnostic: For each long-tail brand, does the small presence cluster around a consistent prompt type or surface, or is it scattered?

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters produced qualified observations across the four-month series?
• What commercial questions can the public benchmark not yet answer?

Across all four months of the series, the qualified observations fell entirely into the Brand Recommendation cluster. There were no qualified observations in the Pricing & Value or Multi-Brand Comparison clusters. This means the public benchmark can currently measure which brands are being recommended, but it cannot yet answer commercial questions about which brands are winning on price, value, or direct comparison. The October response-type mix, which included 91 recommendation shortlists, 26 ranked lists, 11 comparison analyses, and 7 pricing analyses, suggests these intents exist in the raw collection but have not qualified for the public benchmark set.

Brand Opportunity Summary

Questions This Section Answers

• Which brands show the strongest or weakest signals in the October 2026 benchmark?
• What diagnostic question should each brand investigate next?

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

Evidence Behind the Benchmark

Questions This Section Answers

• What data underlies the aggregate recommendation metrics?
• Does source presence establish causation for the observed brand movements?

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 CiteWorks Studio AI Visibility Industry Market research program.

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

Report-Specific Interpretation Notes

• eHealth was the only brand whose full series movement exceeded normal month-to-month variation, rising 6.1 points from July 2026 to October 2026. No brand recorded a significant baseline decline this month.
• The qualified denominator (193 in October 2026) is smaller than the raw collection (800), and brand-level percentages are based only on the qualified set.
• Movement between months identifies changes worth investigating; it does not by itself establish the cause of those changes.
• Small observation counts for LifeShield, IHC Group, Independence American, and Companion Life mean their movements are sensitive to single observations.

Next Step

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

Beneath the aggregate percentage lies a set of unanswered questions: which high-intent prompts are won or lost, which competitor takes the recommendation when a brand is not selected, what attributes AI systems associate with each option, and which external sources shape those answers. The October 2026 data shows a category moving upward on a smaller qualified set, with the leader's margin rebounding and eHealth recovering sharply, but the benchmark alone cannot explain what drove those shifts.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy.

Request an AI visibility audit