How AI Search Is Recommending Meal Delivery Services: Monthly Trends

Executive Summary

Factor leads valid recommendation coverage in Meal Delivery Services at 78.6% in October 2026, holding the top position it also occupied in August 2026 (78.6%) and September 2026 (80.9%). Its lead over second-place HelloFresh (76.4%) stands at 2.2 points in October 2026, a narrower margin than the roughly 4.8 points recorded in September 2026 after HelloFresh moved back up 0.3 points month over month. CookUnity, the strongest riser across the series, climbed from 68.2% in August 2026 to 71.9% in October 2026, a gain of 3.7 points, with 0.8 points of that arriving between September and October 2026.

Blue Apron is the sharpest decliner, falling 8.6 points from 52.9% in August 2026 to 44.3% in October 2026. That move is significant and forms a two-month downtrend, with 3.2 points of the decline arriving between September and October 2026. Home Chef (down 6.1 points to 60.1%) and Purple Carrot (down 5.2 points to 57.6%) round out the three significant decliners for the full series.

October 2026 was a quiet month on its own terms. No brand crossed its significance threshold between September and October 2026, and the benchmark's category summary records no significant movement in the current month. The full-series picture is more eventful than the single-month one: three brands carry significant declines against the August 2026 baseline, and the contest at the top narrowed as HelloFresh recovered the rank-one ground it had surrendered in September 2026.

Each monthly run begins with 800 prompt-surface observations (460 unique questions in August 2026, 449 in September 2026, and 477 in October 2026) across the benchmark's defined AI/search surface universe. All 800 prompts in each month mentioned a tracked brand or competitor; 797 were relevant and 3 were irrelevant in August 2026 and September 2026, and 794 were relevant and 6 were irrelevant in October 2026. The public metrics use the qualified observations that survive both qualification stages: 696 in August 2026, 691 in September 2026, and 677 in October 2026.

Key Findings

AI Response Inconsistency Alerts

Two critical or high-severity factual inconsistencies were detected across three AI platforms (ChatGPT, Copilot, and Google AI Overviews) in the October 2026 collection. Both conflicts were rated high severity and carried a confidence of 0.9.

Blue Apron

When asked "What is the best meal kit USA today?", ChatGPT and Copilot provided conflicting information about Blue Apron's current business and fulfillment status. ChatGPT stated that Blue Apron is currently dealing with major fulfillment problems following its supplier's bankruptcy, including missing ingredients and delayed deliveries, and should be avoided for the moment. Copilot stated that Blue Apron remains a strong choice for overall quality and recipe creativity, delivering high-quality ingredients with creative recipes. These two claims cannot both describe the same company's current operational status. ChatGPT drew on a Wall Street Journal report on the meal-kit industry following a major supplier collapse, alongside a Good Housekeeping review, while Copilot cited a Mealfan ranking, a Business Insider buying guide, and an NBC News roundup. Two additional sources were flagged on the Copilot side of the conflict, including a Kitchn review describing Blue Apron as "Best Overall Meal Kit," with whole ingredients and generous portions.

EveryPlate

When asked "What's the average cost per meal for healthy delivery?", Google AI Overviews and Copilot gave conflicting answers on EveryPlate's per-serving price. Google AI Overviews stated that EveryPlate ranges from $6.00 to $11.00 per serving, while Copilot stated that EveryPlate costs $5.99 to $6.99 per serving. The upper bound of $11.00 contradicts the claim that EveryPlate meals top out at $6.99, as both describe the same company's per-serving price range. Google AI Overviews cited a Daily Harvest nutrition article, a Clean Eatz Kitchen blog post, and a Bon Appetit roundup; Copilot cited Mealfan, Top Meal Boxes, and The Global Statistics. Two sources were flagged on the Google AI Overviews side of the conflict, including a CNET buying guide placing the range at $6 to $8 per serving, and one flagged on the Copilot side, a Global Statistics page listing a budget tier of $6.99 per serving.

Benchmark Context

Questions This Section Answers

• How does the October 2026 qualified observation count differ from the raw 800 prompts collected?
• Why did recommendation-shaped answers rise so sharply while the qualified benchmark set fell?

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

The qualified observation count fell slightly from 696 to 677 across the two bookend months, while the composition of answers shifted far more than the raw count: a majority of qualified answers now take recommendation-shaped form, a change the table below quantifies.

Benchmark-Level Metrics

September 2026 sat between the two columns and matters for interpretation: recommendation-shaped answer share dipped to 45.2% that month, then moved sharply the other way in October 2026 as recommendation shortlists rose from 297 observations in September 2026 to 480 in October 2026. Ranked lists moved in the opposite direction over the same window, from 257 in September 2026 to 128 in October 2026. Pricing analysis observations stood at 30 in October 2026, up from 23 in September 2026.

AI Recommendation Trend

Questions This Section Answers

• Which brands carry significant recommendation-coverage declines, and who leads the category?
• Which brand posted the largest coverage gain across the three months?

Top-Tier Coverage Holds as the Lead Narrows and Decline Spreads Through the Mid-Pack

Valid recommendation coverage measures the share of qualified observations where a brand appears on a ranked recommendation list. Factor remains the category leader in October 2026, but the wider movement is downward: three brands carry significant declines against the August 2026 baseline, and no brand posted a significant rise.

No brand exceeded normal month-to-month variation between September 2026 and October 2026. The category-level change that appears against the August 2026 baseline is the combined product of three separately significant declines plus several smaller downward movements, not the result of any single brand resetting the field.

What Changed This Month

Questions This Section Answers

• How did Factor hold the top spot while its margin narrowed?
• Which brands declined significantly, and what does the Blue Apron-to-CookUnity gap show about the mid-pack?

Factor: Leadership Held With a Narrower Margin

Factor's valid recommendation coverage was unchanged across the three months at 78.6% in August 2026 and 78.6% in October 2026. The more useful read for a decision-maker sits in the month-to-month detail: coverage slipped 2.3 points from 80.9% in September 2026, a move within normal variation but enough to hold the lead by a smaller margin than the prior month.

Placement strengthened even as coverage flattened. Top-three rate rose 12.0 points from 27.9% in August 2026 to 39.9% in October 2026, a significant move, and rank-one rate rose 2.6 points to 13.2%. Raw mention presence was effectively flat at 87.7% in October 2026 against 86.8% in August 2026.

The distinction to note is that Factor is appearing at roughly the same rate while converting more of those appearances into high placements. The absolute counts sit in a healthy range: 532 valid recommendations out of 677 qualified observations in October 2026. Highest-priority diagnostic: which prompts moved Factor from the answer set between September and October 2026, and whether the top-three gain is concentrated in a subset of surfaces.

HelloFresh: Flat Coverage, Recovered Rank-One Capture

HelloFresh held valid recommendation coverage steady at 76.4% in August 2026 and 76.4% in October 2026, with a 0.3-point month-over-month rise. Against the September 2026 reading of 76.1%, the brand has held essentially flat across all three months while the distance to the leader narrowed from roughly 4.8 points in September 2026 to 2.2 points in October 2026.

Placement gains were the substantive change. Top-three rate rose 12.8 points from 45.4% in August 2026 to 58.2% in October 2026, a significant move, and rank-one rate rose 8.8 points to 24.2%, also significant. The rank-one recovery is notable because HelloFresh had already climbed to 20.5% in September 2026 and then extended further.

The distinction is that HelloFresh's breadth is unchanged while its position within answers has strengthened materially. With 517 valid recommendations in October 2026, this is a placement-quality story at scale. Highest-priority diagnostic: which prompt themes drive the rank-one gain, and whether the advance is concentrated in any single AI surface family.

CookUnity: The Series' Largest Riser

CookUnity's valid recommendation coverage rose 3.7 points from 68.2% in August 2026 to 71.9% in October 2026, with a 0.8-point gain arriving between September and October 2026. The overall move is within the brand's normal variation but forms a two-month upward streak.

Supporting metrics confirm the direction. Raw mention presence rose 5.7 points to 80.7% from 75.0% in August 2026, a significant gain, and top-three rate rose 11.1 points to 41.9%, also significant. Rank-one rate rose 1.5 points to 18.0%. At 487 valid recommendations in October 2026, the counts are large enough to read the placement shifts directly.

The distinction is that CookUnity is gaining on both presence and placement at the same time, which is the combination that widens gaps to rivals. Highest-priority diagnostic: which prompts and surfaces account for the top-three gain, and whether the advance is broadening or concentrated.

Blue Apron: Significant Decline and a Widening Gap

Blue Apron's valid recommendation coverage fell 8.6 points from 52.9% in August 2026 to 44.3% in October 2026, one of three significant declines in the category. The decline has run for two consecutive months and accelerates when read month over month: 3.2 points of the drop arrived between September and October 2026, against the 5.4-point decline recorded between August and September 2026.

Supporting metrics moved the same way. Raw mention presence fell 10.9 points to 51.7% from 62.6% in August 2026, a significant drop. Rank-one rate fell 4.9 points to 9.6%, also significant. Top-three rate fell 3.3 points to 26.1%, within normal variation. Presence (51.7%) still exceeds coverage (44.3%), meaning the brand appears in answers more often than it reaches recommendation lists, and 300 valid recommendations were recorded for Blue Apron in October 2026, the smallest among the leading brands.

The category consequence sits in the comparative gaps. Blue Apron now trails CookUnity by 27.6 points, up from 15.3 points in August 2026, a gap that widened in each month of the series. Highest-priority diagnostic: which competitor absorbs the recommendation on prompts where Blue Apron appears but is unranked, and whether those prompts cluster around availability or fulfillment language given the conflicting answers reported above.

Home Chef: A Significant Decline Carried by Presence

Home Chef's valid recommendation coverage fell 6.1 points from 66.2% in August 2026 to 60.1% in October 2026, a significant move and the second-largest decline in the field. The month-over-month picture is milder: coverage slipped 1.4 points from 61.5% in September 2026, within normal variation.

Presence carried the decline. Raw mention presence fell 6.2 points to 69.9% from 76.1% in August 2026, a significant drop, while top-three rate rose 3.5 points to 29.1% and rank-one rate slipped 1.1 points to 3.4%. Home Chef recorded 407 valid recommendations in October 2026.

The distinction is that Home Chef is being named less often but ranking similarly when it does appear, which points at answer-set inclusion rather than placement quality. The gap to CookUnity widened to 11.8 points in October 2026 from 2.0 points in August 2026. Highest-priority diagnostic: which prompt themes dropped Home Chef out of the answer set entirely.

Purple Carrot: Significant Decline With a Sharp Rank-One Drop

Purple Carrot's valid recommendation coverage fell 5.2 points from 62.8% in August 2026 to 57.6% in October 2026, the third significant decline in the category. In absolute terms the brand holds 390 valid recommendations in October 2026, so the coverage base remains broad.

The sharper signal sits in placement. Rank-one rate fell 1.4 points to 0.9% from 2.3% in August 2026, a significant move on a small count, and raw mention presence fell 7.5 points to 61.6%, also significant. Top-three rate slipped 0.8 points to 4.7%.

The distinction for the reader is that Purple Carrot retains wide coverage but has lost nearly all of its rank-one placements, with 6 rank-one counts in October 2026. The gap to CookUnity widened to 14.3 points from 5.4 points in August 2026. Highest-priority diagnostic: which narrow prompt set previously produced Purple Carrot's rank-one placements, and whether they survive in October 2026.

EveryPlate, Marley Spoon, Green Chef, Sunbasket: Smaller Movements in the Same Direction

The remaining brands moved without crossing their thresholds. EveryPlate fell 1.3 points to 62.3% against August 2026, with a 4.1-point month-over-month decline from 66.4% in September 2026, while its top-three rate rose 5.6 points to 16.7%, a significant placement gain. Marley Spoon fell 4.7 points to 34.7%, with rank-one rate holding at 13.3%, one of the highest in the field against the lowest coverage tier. Green Chef fell 1.6 points to 50.1%, with rank-one rate down 2.1 points to 0.9%, a significant move on small counts. Sunbasket fell 0.4 points to 33.7%, holding the lowest coverage in the category at 228 valid recommendations, while its top-three and rank-one rates rose modestly.

Highest-priority diagnostic: for each, whether the coverage movement tracks a change in answer-set inclusion or a change in placement once included, since those two conditions call for different investigations.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent cluster actually carries qualified observations for meal delivery brands?
• What can the benchmark say about pricing and head-to-head comparison questions?

The qualified observations in this category fall entirely within brand-recommendation prompts. Discovery and consideration is the only cluster carrying observations in any month of the series, and the comparison and pricing clusters return no qualified observations for any tracked brand. October 2026 recorded 30 pricing analysis observations, and the conflicting per-serving claims reported above show that price information is reaching users even where the benchmark's qualified set does not isolate it as a cluster.

The commercial consequence is that the public benchmark can speak to who AI recommends first for general discovery questions, and cannot yet answer the pricing, value, or head-to-head comparison questions that sit closer to a purchase decision. Those questions require a company-level analysis that can isolate the specific prompts and surfaces where cost and comparison language appears.

Brand Opportunity Summary

Questions This Section Answers

• Which brand carries a significant two-month decline in recommendation coverage?
• Which company-specific diagnostic should each brand investigate first?

The benchmark identifies where attention is warranted across all ten tracked brands; a company-level analysis is needed to explain why individual brands gain or lose recommendation credit.

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 Visibility Market Discovery research program.

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

Report-Specific Interpretation Notes

• Qualified observations (677 in October 2026) form the denominator for all recommendation coverage percentages, not the raw 800 prompts collected.
• Several rank-one rates in this category rest on small absolute counts, including Purple Carrot (6 rank-one counts in October 2026), Green Chef (6), Sunbasket (12), and EveryPlate (12); those figures should be read as directional rather than settled.
• Movement between months identifies changes worth investigating; it does not by itself establish cause. The significant declines reported here are measured against the August 2026 baseline, and no brand crossed its threshold between September and October 2026.

Next Step

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

The aggregate percentages say Factor leads at 78.6% in October 2026 and Blue Apron fell 8.6 points against August 2026. Beneath those numbers sit the questions that matter for strategy: which high-intent prompts each brand is winning or losing, which competitor takes the recommendation when a brand drops off the list, what attributes AI systems associate with each option, and which external sources shape those answers. The conflicting answers reported for Blue Apron and EveryPlate show that source selection and staleness matter here, not only ranking.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. It converts the benchmark's directional signals into concrete actions for the prompts and surfaces where recommendation share is slipping or ready to grow.

Request an AI visibility audit