How AI Search Is Recommending Water Delivery Services: Monthly Trends

Executive Summary

The leader held, but the ground beneath it narrowed. Mountain Valley Spring Water remains the most-recommended brand in AI search for Water Delivery Services with 39.8% valid recommendation coverage in October 2026, a 9.6-point lead over second-place Primo Water at 30.2%. That margin is the tightest of the four-month series, down from 21.7 points in July 2026, as Primo Water and Culligan both climbed sharply in October while Mountain Valley Spring Water continued a three-month decline.

The sharpest movement this month belongs to Primo Water and Culligan. Primo Water rose 12.0 points to 30.2% from 18.2% in September 2026, a rise the benchmark flags as significant, and Culligan rose 11.7 points to 29.1% from 17.4%, also significant. Both brands reversed multi-month declines in a single month, and both now sit within roughly one point of each other in the second tier.

The largest baseline-to-current decline is Mountain Valley Spring Water, down 9.1 points from 48.9% in July 2026, a move the benchmark classifies as a significant decliner and a third consecutive monthly decline. Its raw mention presence fell 11.1 points to 54.9% over the same period, even as its top-three and rank-one rates held roughly flat.

Against the prior month, October 2026 is the category's most active month in the series. The gap between first and third place compressed to 10.7 points from 25.9 points in September 2026, and Culligan now sits just behind Primo Water rather than a full tier below the leader.

Each monthly run begins with 800 prompt-surface observations (548 unique questions in October 2026, against 492 in July 2026 across the series) across the benchmark's defined AI/search surface universe. Of those, 800 mentioned a tracked brand or competitor; 703 were relevant and 97 were irrelevant in October 2026, against 720 relevant and 80 irrelevant in July 2026. The public metrics use the 344 qualified observations in October 2026, against 368 in July 2026, that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Where did AI platforms give conflicting answers about Mountain Valley Spring Water's lawsuit status?
• Which ownership and acquisition facts about Mountain Valley Spring Water remain unresolved across platforms?
• Why do AI platforms disagree about whether Sparkletts and Primo Water are the same brand?

The benchmark detected 28 critical and high-severity factual inconsistencies across six AI platforms: ChatGPT, Copilot, Gemini, Perplexity, AI Overviews, and AI Mode. The conflicts cluster around three companies and concentrate heavily on lawsuit status, acquisition history, and brand ownership.

Mountain Valley Spring Water

Sixteen critical or high-severity conflicts involve Mountain Valley Spring Water, spanning every one of the six tested platforms. The dominant pattern is a split over the status of a federal class action. When asked "What is the Mountain Valley Spring Water scandal?", ChatGPT stated the lawsuit was voluntarily dismissed without prejudice in November 2025, citing a court filing and two legal-analysis pages, while Copilot stated the case was in active litigation with class certification pending as of mid-2026, citing three lawsuit-tracking pages. A parallel conflict on the same question saw ChatGPT's dismissal claim set against Gemini's statement that the case continues to move through the legal system as a proposed class action, citing a consumer cooperative page and two class-action news pages.

The same fault line reappears when users ask "What is happening with Mountain Valley Spring Water?" AI Mode stated the plaintiff voluntarily dismissed the complaint and the federal case was officially closed without prejudice, while Gemini stated the litigation is ongoing and the allegations have not been resolved in court. ChatGPT's dismissal position was also set against Copilot's "active litigation in 2026, with class certification hearings underway" and against Perplexity's statement that the case was filed in August 2025 and is proceeding in federal court. In each instance, responses drew on opposing source sets, with dismissal-side answers citing court records and active-litigation answers citing lawsuit-news and legal-commentary pages.

A second cluster concerns when and by whom the brand was acquired. When asked "What happened to Mountain Valley Spring Water?", ChatGPT stated Primo acquired the business in 2018, citing the brand's own site and an investor earnings deck, while AI Overviews stated Primo Water Corporation bought the brand in 2020, citing social video and retail pages. Perplexity, answering the same question, described a notable sale to DS Services, a subsidiary of Cott, in 2018. A related conflict saw AI Mode describe the brand's parent as Primo Water Corporation following its merger with BlueTriton to form Primo Brands, while Copilot described Mountain Valley Spring Co. as preparing its own IPO, implying an independent company going public.

The remaining conflicts concern acquisition year and contamination status. On ownership, Perplexity stated Primo Water announced acquiring Mountain Valley Water Company assets in 2020, while AI Mode stated the brand was acquired by Primo Water's DS Services division in October 2018, citing trade-press coverage of the original deal. One conflict is a purity disagreement: asked "What is the cleanest bottled water to drink in the USA?", AI Mode described Mountain Valley as a standout clean natural spring water with no contamination concerns, while Copilot stated the brand is currently facing lawsuits over arsenic and uranium contamination, citing a wellness review page.

Sparkletts

Eight high-severity conflicts involve Sparkletts, and all eight turn on the same question: whether Sparkletts and Primo Water are the same brand. Asked "Is Primo Water the same as Sparkletts?", Copilot stated that Sparkletts has been rebranded under the Primo Water name and that the Sparkletts name is being phased out, citing a branding-update blog and two Primo-owned pages. ChatGPT stated the two are separate water brands and that Sparkletts remains a distinct regional brand, citing Primo Brands' own about page. Against AI Overviews, Copilot's phase-out claim conflicted with the statement that Sparkletts is owned by the same parent but has historically maintained distinct regional branding. Against Gemini, the phase-out claim conflicted with the statement that the two brands are produced and owned by the same parent company, and with a second Gemini response noting that the Sparkletts name may still appear on regional products or delivery trucks. Perplexity appeared on both sides of the question across separate responses, once stating that Sparkletts is a brand Primo distributes and once stating the two remain distinct brand identities, citing a cooking site and a comparison page that sit on opposite sides of the rebranding question.

Culligan

Four high-severity conflicts involve Culligan, three of them pricing-related. The pricing conflicts show materially different price ranges for the same product. Asked "Is Culligan cheaper than bottled water?", ChatGPT stated Culligan lists 5-gallon bottled water at $18 per bottle on its national site, citing the brand's own product page, while AI Mode stated Culligan 5-gallon delivery jugs average around $6 to $8 per jug, citing a cost-comparison calculator. A parallel conflict saw AI Overviews state that traditional 5-gallon jug delivery often costs $10 to $14 per jug against AI Mode's $6 to $8 average. A third pricing conflict set ChatGPT's $18 per-bottle figure against Perplexity's statement that a 5-gallon Culligan bottle runs roughly $7 to $15. On equipment rather than water, ChatGPT stated that whole-house Culligan filtration media can often last several years, while Perplexity stated whole-house filters last roughly 6 to 12 months, citing the brand's own replacement guidance and a regional dealer page.

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 intermediate months moved within a narrow qualified range, from 411 observations in August 2026 to 379 in September 2026 and 344 in October 2026, the smallest qualified set of the series.

Benchmark-Level Metrics

Recommendation-shaped answer share dipped to 35.1% in September 2026 before returning to 43.3% in October, and valid recommendation shortlist share reached its highest level of the series this month.

AI Recommendation Trend

Questions This Section Answers

• Which brands lead AI recommendation coverage for water delivery, and how did their rank order shift from July to October 2026?
• How did the leader's margin over second place change across the four-month series?

Recommendation Coverage by Brand

The category moved from a single-brand lead to a compressed top tier, with the leader's margin over second place at its narrowest point of the series.

The category-level change this month came from the combination of one significant decline and two significant single-month rises rather than a single dominating shift.

What Changed This Month

Questions This Section Answers

• Why is Mountain Valley Spring Water's decline showing up as lost presence rather than lost ranking quality?
• What drove Primo Water's 12.0-point October jump, and did it come from placement or presence?
• How does Culligan's October position compare with its July baseline?

Mountain Valley Spring Water

Mountain Valley Spring Water's valid recommendation coverage fell to 39.8% in October 2026 from 43.3% in September 2026, a 3.5-point move within the benchmark's normal month-to-month range for the brand. The baseline-to-current position is the category's only significant decline: down 9.1 points from 48.9% in July 2026, now in a third consecutive month of decline.

Raw mention presence fell 11.1 points to 54.9% from 66.0% in July 2026. Top-three rate rose 3.4 points to 35.5% from 32.1% across the same period, and rank-one rate rose 0.5 points to 20.3% from 19.8%. Valid recommendation count fell to 137 in October from 164 in September and 180 in July. Net sentiment held at 0.8 across July and October.

The distinction to notice: the brand is being named less often across the qualified set while still placing near the top of the lists it does enter, which means its decline is showing up as lost presence rather than lost ranking quality.

Highest-priority diagnostic: Which prompt classes account for the three-month presence decline, and whether the loss is concentrated on specific AI surfaces.

Primo Water

Primo Water posted the largest single-month rise in the category, with valid recommendation coverage climbing to 30.2% in October 2026 from 18.2% in September 2026. The 12.0-point jump is significant and reverses a two-month decline that had taken the brand down 9.0 points from July.

Top-three rate rose 7.4 points to 29.1% from 21.7% between July and October, and rank-one rate rose 7.4 points to 18.3% from 10.9%. Raw mention presence rose 3.2 points to 54.6% from 51.4%, within normal range. Valid recommendation count rose to 104 in October from 69 in September and 100 in July. Net sentiment held at 0.6.

The distinction to notice: Primo Water now wins top-three placement and first-position recommendations in a materially larger share of answers than it did in July, so its gain is a placement shift, not just a presence shift.

Highest-priority diagnostic: Which surfaces and question types drove the October jump, and whether the placement gains hold in the next run.

Culligan

Culligan recorded the second-largest single-month rise, with valid recommendation coverage reaching 29.1% in October 2026 from 17.4% in September 2026. The 11.7-point move is significant and reverses a two-month decline from 24.5% in July.

Top-three rate rose 5.5 points to 26.7% from 21.2% between July and October, and rank-one rate rose 2.7 points to 8.7% from 6.0%, both within normal range. Raw mention presence rose 3.0 points to 45.4% from 42.4%. Valid recommendation count rose to 100 in October from 66 in September and 90 in July. Net sentiment held at 0.7.

The distinction to notice: Culligan's October position is now above its July baseline, so the two-month dip it carried through September has been fully recovered across both coverage and top-three placement.

Highest-priority diagnostic: Whether the recovery is concentrated in the pricing-related prompts where conflicting answers were also detected, or spread across the recommendation set.

The Compressed Top Tier

A notable category observation: the gap between the leader and third place narrowed to 10.7 points in October 2026 from 25.9 points in September and 24.4 points in July. The compression reflects Mountain Valley Spring Water's decline and the simultaneous October rises at Primo Water and Culligan, rather than any single brand's shift. The gap between Aquafina and Mountain Valley Spring Water has also narrowed every month across the series, from 44.6 points in July to 34.3 points in October.

Highest-priority diagnostic: Whether the compression reflects a durable redistribution of recommendations across the top tier or a single-month reordering that reverts next period.

Aquafina

Aquafina's valid recommendation coverage held flat month over month at 5.5% in October 2026 and September 2026. The baseline-to-current position shows a 1.2-point rise from 4.3% in July 2026, within normal range.

Raw mention presence fell 2.9 points to 43.6% from 46.5% between July and October. Top-three rate rose 1.9 points to 3.8% from 1.9%, and rank-one rate fell 0.2 points to 0.6% from 0.8%. Net sentiment remained negative at negative 0.2 in both months, the only negative reading among tracked brands.

The distinction to notice: Aquafina remains the category's most visible brand that is not being recommended, appearing in 43.6% of qualified observations while winning valid recommendations in 5.5%.

Highest-priority diagnostic: Which prompts produce Aquafina mentions without recommendations, and what the persistent negative sentiment is attached to.

Sparkletts

Sparkletts's valid recommendation coverage was essentially flat, moving down 0.6 points to 11.6% in October 2026 from 12.2% in July 2026, within normal range. The brand rose 4.0 points month over month from 7.6% in September, its second consecutive monthly rise.

Raw mention presence fell 6.5 points to 20.1% from 26.6% between July and October. Top-three rate rose 0.6 points to 9.6% from 9.0%, and rank-one rate rose 0.1 points to 2.6% from 2.5%. Valid recommendation count rose to 40 in October from 29 in September and 45 in July.

The distinction to notice: Sparkletts is being mentioned in fewer qualified observations while converting a larger share of those mentions into recommendations, so its coverage has stabilized even as its footprint has narrowed.

Highest-priority diagnostic: Whether the presence decline is surface-specific and whether the conversion improvement holds.

Absopure and DS Services

Absopure's valid recommendation coverage fell 0.8 points to 0.6% in October 2026 from 1.4% in July 2026, within normal range. Its valid recommendation count was 2 in October, against 3 in September and 5 in July. DS Services rose 0.6 points to 0.6% from 0.0% in July, a third consecutive monthly rise the benchmark classifies as a watch item because the cumulative movement remains within normal variation. Its valid recommendation count was 2 in October, against 2 in September and 0 in July.

The distinction to notice for both brands: these percentages rest on single-digit recommendation counts, so the direction is worth tracking but the magnitude is not yet a stable trend.

Highest-priority diagnostic: Which prompt types produced the two valid recommendations for each brand, and whether those prompt types recur.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters did the October qualified observations actually cover?
• What can the benchmark not yet measure about how AI positions water delivery brands on price or comparisons?
• Why is the pricing gap notable given the Culligan pricing conflicts?

The qualified observations in October 2026 fell entirely within the brand recommendation category, all 344 of them. The comparison and pricing clusters recorded zero qualified observations in the month.

The public benchmark therefore measures which brand AI systems recommend, but it cannot yet answer how AI positions brands on price, value, or head-to-head comparisons. That gap is notable given the Culligan pricing conflicts reported above, which show AI platforms giving materially different price ranges for the same product while the benchmark's qualified set contains no pricing-class observations to measure the effect.

Brand Opportunity Summary

Questions This Section Answers

• Which water delivery brand carries the most urgent diagnostic signal heading into the next run?
• What highest-priority diagnostic applies to each brand's current position in AI recommendations?

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 Visibility Market Discovery research program.

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

Report-Specific Interpretation Notes

• Small-count movement: Absopure's 0.6% and DS Services's 0.6% both rest on two valid recommendations in October 2026. These should be read as directional signals rather than stable trends.
• Qualified denominator vs raw collection: All coverage percentages use the qualified observation count (368 in July 2026, 344 in October 2026) as the denominator, not the 800 raw prompts. The funnel table shows the full journey from raw collection to qualified set.
• Directional analysis: Month-over-month movement identifies changes worth investigating; it does not by itself establish the cause of those changes. The significant October rises at Primo Water and Culligan, and the significant three-month decline at Mountain Valley Spring Water, indicate where attention is warranted, not why the changes occurred.

Next Step

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

The aggregate percentages leave the most decision-relevant questions unanswered: which high-intent prompts does a brand win, which competitor takes the recommendation when a brand loses, what attributes does AI associate with each option, and which external sources shape those answers. For Mountain Valley Spring Water, the question is where its three-month presence decline concentrated and whether it is surface-specific. For Primo Water and Culligan, the question is whether the October rises reflect durable placement gains or a single-month reordering. For every brand in the category, the pricing conflicts reported above raise a further question the aggregate set cannot answer: which price AI surfaces cite, and which source pages drive the range.

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 a list of concrete actions tied to the surfaces and prompts that matter most for a given brand.

Request an AI visibility audit