How AI Search Is Recommending Language Learning Software: Monthly Trends

Executive Summary

The category leadership picture held in October 2026, but the composition underneath it changed materially. Duolingo remains the coverage leader at 85.5% valid recommendation coverage, a 4.0-point edge over Babbel's 81.5%. Both brands sit in a tier of their own; Pimsleur, at 72.0%, is the clear third-place challenger and now sits 9.5 points behind Babbel and 25.4 points ahead of italki. The leader is stable, but the spread between the top tier and the middle of the field widened again.

Babbel was the strongest significant riser of the quarter. Its coverage climbed 6.9 points from 74.6% in July 2026 to 81.5% in October 2026, exceeding the significance threshold, and it rose 4.6 points from September's 76.9%. Its top-3 placement rate jumped 13.7 points over the same period, from 56.4% to 70.1%, while rank-one presence softened slightly, down 2.2 points to 19.1%.

Pimsleur recorded the second significant baseline-to-current gain, up 8.9 points from 63.1% in July 2026 to 72.0% in October 2026, its third consecutive monthly increase. Top-3 placement rose 23.7 points to 50.8%, and rank-one presence climbed 4.7 points to 7.0%. Rosetta Stone recorded the sharpest decline in the series, down 14.0 points from 48.5% in July 2026 to 34.5% in October 2026, with its raw mention presence falling 15.6 points over the same period. Memrise was the other significant decliner, down 6.9 points from 32.5% to 25.6% and down 5.4 points against September's 31.0%.

On a baseline-to-current basis, four brands exceeded the significance threshold this quarter: Babbel and Pimsleur rising, Memrise and Rosetta Stone declining. Month-over-month, only Memrise's September-to-October decline cleared that threshold; the rest of the field's single-month moves fell within normal variation. The category-level position was mixed: two significant risers, two significant decliners, and six brands within normal variation across the quarter.

Each monthly run begins with 800 prompt-surface observations across the benchmark's defined AI/search surface universe. In July 2026, unique questions totaled 464; by October 2026, unique questions totaled 520. All 800 observations in each month mentioned a tracked brand or competitor. Relevant prompts totaled 800 in July 2026 and 795 in October 2026; irrelevant prompts totaled 0 and 5 respectively. The public metrics use the 520 qualified observations in July 2026 and the 539 qualified observations in October 2026 that survive both qualification stages. August 2026 (522 qualified observations) and September 2026 (558) sit between them.

Key Findings

AI Response Inconsistency Alerts

One critical or high-severity factual inconsistency was detected across two AI platforms in the current measurement period. This is a category-level observation; the benchmark identifies where an inconsistency exists but does not establish which claim is correct.

Babbel

A high-confidence pricing inconsistency was detected between Copilot and Google AI Mode. When asked "Is paying for Babbel worth it?", Copilot stated that subscriptions start around $8.95 per month, citing FluentU, LanguaVibe, and PCMag. Google AI Mode stated that Babbel costs around $14 to $18 per month if paid monthly, citing a Reddit thread, a Medium review, and a YouTube video. The two claims cannot both describe the same starting monthly subscription price. The gap between the two figures is roughly twofold, which matters for any buyer comparison shopping on monthly cost. The responses relied on different source classes: Copilot referenced editorial review sites, while Google AI Mode drew on community and video sources. The benchmark flags the divergence; it does not adjudicate which price is accurate.

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.

These qualified counts anchor every brand-level percentage below; the raw collection volume is shown for context only and is not used as a denominator anywhere in the report.

Benchmark-Level Metrics

August 2026 (522 qualified observations, 66.9% recommendation-shaped share) and September 2026 (558 qualified observations, 55.4%) are intermediate months in this series. The recommendation-shaped answer share fell through September before recovering sharply in October, and the valid recommendation shortlist share followed the same trough-and-recovery shape.

AI Recommendation Trend

Questions This Section Answers

• Who leads recommendation coverage in October 2026, and how big is the gap to the next brands?
• How did the tier below the leader reorder between July and October 2026?

The leader held; the tier below it reordered twice

October 2026 preserved the commercial structure of July: Duolingo and Babbel in a tier of their own, Pimsleur as the clear third, and a broader middle that continues to shift. What changed is the size of the separation. Babbel closed most of the gap to Duolingo while Pimsleur pulled further ahead of the fourth-place cluster.

Four brands exceeded the significance threshold on a baseline-to-current basis, so October's category change is not purely the sum of small movements. Babbel and Pimsleur carried the upward side; Rosetta Stone and Memrise carried the downward side.

What Changed This Month

Questions This Section Answers

• Which brands moved beyond normal monthly variation, and in which direction?
• Why is Babbel's top-3 placement rising while its rank-one presence softens?
• What does Rosetta Stone's decline in mentions but firmer top-3 rate say about its visibility?

Babbel: Significant riser with a placement-quality shift

Babbel's valid recommendation coverage rose 6.9 points from 74.6% in July 2026 to 81.5% in October 2026, exceeding the significance threshold, and rose 4.6 points from September's 76.9%. This is the second consecutive monthly gain and Babbel's highest coverage level in the tracked series.

Top-3 placement rose 13.7 points from 56.4% to 70.1%, and valid recommendation count reached 439, up from 388 in July. Rank-one presence moved in the other direction, down 2.2 points from 21.3% to 19.1%. Raw mention presence rose 1.7 points to 86.3%.

The distinction to notice: Babbel is being recommended in more responses and placed in top-3 positions far more often, but it is being named first slightly less often than at baseline. The 70.1% top-3 rate against the 19.1% rank-one rate describes a brand that is almost always in the shortlist conversation but is winning the single-answer question less frequently than its shortlist dominance would suggest.

Highest-priority diagnostic: Which prompt categories added Babbel to top-3 positions without converting to rank-one placement, and which competitor takes the rank-one slot in those prompts?

Pimsleur: Third consecutive monthly gain

Pimsleur's valid recommendation coverage rose 8.9 points from 63.1% in July 2026 to 72.0% in October 2026, a significant baseline-to-current gain, and rose 5.2 points from September's 66.8%. This is the third consecutive monthly increase in the series, the longest active upward streak in the category.

Top-3 placement climbed 23.7 points from 27.1% to 50.8%, and rank-one presence rose 4.7 points from 2.3% to 7.0%. Raw mention presence rose 5.1 points to 75.3%. Valid recommendation count reached 388, up from 328 in July.

The distinction to notice: Pimsleur's gains are showing up in placement rather than simple presence. Its mention rate rose modestly across the quarter, but its top-3 rate nearly doubled, which means prompts that already mentioned Pimsleur are increasingly ranking it near the top. That is a different pattern from a brand winning new mentions.

Highest-priority diagnostic: Which surfaces are moving Pimsleur up within shortlists it already appears in, and does the pattern hold if any single surface is excluded?

Rosetta Stone: Sharpest decline in the series

Rosetta Stone's valid recommendation coverage fell 14.0 points from 48.5% in July 2026 to 34.5% in October 2026, the sharpest baseline-to-current decline in the category, and fell 2.1 points from September's 36.6%. Its raw mention presence fell 15.6 points from 58.3% to 42.7%.

Top-3 placement moved slightly the other way, up 1.5 points from 10.4% to 11.9%. Rank-one presence fell 0.4 points from 1.5% to 1.1%. Valid recommendation count fell from 252 in July to 186 in October. The largest single-month movement was 12.3 points, between July and August.

The distinction to notice: Rosetta Stone's decline is concentrated in visibility, not in how it is treated when it appears. Its top-3 rate firmed slightly even as its mention rate dropped sharply, which means the brand is appearing in fewer responses but holding its position reasonably well in the ones that still surface it. Net sentiment remains positive at 0.8.

Highest-priority diagnostic: Which surfaces or prompt categories stopped mentioning Rosetta Stone between July and August, and which evidence sources are those systems relying on instead?

Memrise: Second consecutive significant drop

Memrise's valid recommendation coverage fell 6.9 points from 32.5% in July 2026 to 25.6% in October 2026, a significant baseline-to-current decline, and fell 5.4 points from September's 31.0%, which also exceeds the threshold. This is the second consecutive month of significant decline.

Raw mention presence fell 10.7 points from 40.6% to 29.9%. Top-3 placement rose 2.6 points from 5.0% to 7.6%. Rank-one presence was flat at 0.4%. Valid recommendation count fell from 169 in July to 138 in October.

The distinction to notice: like Rosetta Stone, Memrise's decline is a presence story rather than a placement story. When Memrise appears, it is placed in top-3 positions more often than at baseline, but it appears in a materially smaller share of responses. Absolute counts are modest, so single-prompt movements register as visible percentage shifts.

Highest-priority diagnostic: Which prompt categories dropped Memrise between September and October, and did the same categories shift toward a specific competitor?

Category divergence: the Pimsleur and Rosetta Stone gap

The gap between Pimsleur and Rosetta Stone widened from 14.6 points in July 2026 to 37.5 points in October 2026, an increase of 22.9 points, and widened in every month of the series. The Babbel-to-Rosetta Stone gap widened by 20.9 points over the same period, and the Duolingo-to-Rosetta Stone gap by 17.8 points.

This is a category-observation about competitive repositioning rather than any single brand's problem. Three of the four brands at the top of the coverage table are now separated from one mid-table brand by a margin that grew every month of the tracked series, which means the shape of the category, not just the ordering, is changing.

Highest-priority diagnostic: Whether the widening is driven by a shared surface or prompt category, or by three independent brand-specific patterns that happen to point the same direction.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters are actually represented in the qualified observations?
• Why can't the benchmark yet answer questions about pricing, value, or head-to-head comparisons?

In October 2026, all 539 qualified observations fell into the Brand Recommendation cluster. There were no qualified observations in the Pricing & Value or Multi-Brand Comparison clusters in July 2026 or October 2026, meaning the public benchmark cannot yet answer questions about how AI systems position these brands on price, value, or head-to-head comparison. The current evidence is confined to direct brand recommendation behavior. This is notable because the inconsistency alert above concerns a pricing question, which sits in a cluster the benchmark does not currently qualify.

Brand Opportunity Summary

Questions This Section Answers

• Which brand carries the strongest current signal, and what diagnostic matters most for it?
• Which brands declined significantly, and what should be investigated first for each?

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 CiteWorks Studio 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 (539 in October 2026) form the public denominator; raw collection volume was 800 prompts. Brand-level percentages are calculated only within the qualified set.
• Small absolute counts matter. Mondly's 5.2% coverage represents 28 qualified recommendations, and Mango Languages' 8.5% represents 46. Movement at these counts is less stable than for brands with larger bases, though these are valid observations for the category.
• Significant movement is identified when the baseline-to-current change exceeds the threshold for normal month-to-month variation. It indicates a change worth investigating, not a proven cause.
• This analysis identifies where attention is warranted; it does not establish why a brand gained or lost recommendation credit.

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: 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. The October 2026 data shows Babbel's significant rise, Pimsleur's three-month streak, and Rosetta Stone's sharp quarter-long decline, but it does not show which prompts changed or which evidence sources drove the shifts. The pricing inconsistency between Copilot and Google AI Mode illustrates the same limit: the benchmark can flag that two platforms disagree on Babbel's monthly cost, but only a prompt-level review can show which sources each platform weighted.

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