How AI Search Is Recommending Document Management Software: Monthly Trends

Executive Summary

Microsoft SharePoint remains the coverage leader in AI recommendations for document management software, but the category leader lost ground against July 2026. SharePoint's valid recommendation coverage stands at 50.1% in October 2026, down 8.9 points from its July 2026 baseline, a decline above normal month-to-month variation. The gap to the next brand, Box at 40.2%, is 9.9 points. That margin has narrowed from earlier in the series even as SharePoint held first place in all four months.

The strongest upward mover with material coverage is Dropbox Business, which rose 6.5 points from September 2026 (16.8%) to October 2026 (23.3%), a move above normal variation for the brand. Its valid recommendation coverage sits 1.9 points below its July 2026 baseline of 25.2%. FileHold registered the largest baseline-to-current gain, up 4.3 points from July 2026 (1.0%) to October 2026 (5.3%), a significant riser on a small base, now up three consecutive months.

The sharpest decliner from baseline is Laserfiche, down 12.0 points from July 2026 (34.5%) to October 2026 (22.5%). Box carries the second-largest baseline-to-current decline at 9.4 points, though Box recovered 6.4 points from September to October. Five brands sit below their July baseline at levels above normal variation: Box, Laserfiche, M-Files, Microsoft SharePoint, and Revver (eFileCabinet).

October 2026 is a partial rebound month. Most brands that fell sharply from July to August regained some coverage from September to October, but none returned to July levels. The category leader remains the same, the middle tier has compressed, and the baseline-to-current direction is still down for the majority of the field.

Each monthly run begins with 800 prompt-surface observations (499 unique questions in July, 566 in August, 575 in September, and 578 in October) across the benchmark's defined AI/search surface universe. Of those, all 800 mentioned a tracked brand or competitor in each month; 437 were relevant and 363 were irrelevant in July, versus 562 relevant and 238 irrelevant in August, 540 relevant and 260 irrelevant in September, and 524 relevant and 276 irrelevant in October. The public metrics use the 385 qualified observations in July and 395 in October, with 454 in August and 423 in September, that survive both qualification stages.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• What pricing inconsistency did the AI platforms produce for DocuWare, and how far apart were the cited figures?
• Does the flagged pricing conflict change any brand's recommendation coverage?

The benchmark detected 1 critical or high-severity factual inconsistency across 2 AI platforms in this period, involving conflicting pricing information from ChatGPT and Copilot.

DocuWare

When asked "How much does DocuWare cost per month?", the AI platforms provided conflicting information about DocuWare's entry-level monthly pricing, a high-severity pricing conflict. ChatGPT stated that DocuWare cloud pricing runs about £10–£60 per user per month, and that a 5-user setup could cost roughly $50–$300/month, citing the company's own pricing FAQ page. Copilot stated that DocuWare's monthly subscription starts around $300 per month for small teams, citing third-party pricing aggregators including checkthat.ai, softabase.com, and itqlick.com. The two responses are incompatible at the minimum price point, since one places a small-team entry as low as $50/month while the other sets the floor near $300/month.

Flagged sources on both sides carried supporting excerpts. A softabase.com page used by Copilot listed "Starting at $300/mo" with a confidence of 0.95, and a checkthat.ai page used by Copilot estimated $296/month for a Cloud 4 tier at authorized-partner rates with a confidence of 0.9. An itqlick.com page carried a wider range of $3,600–$12,000 annually, or $30–$100 per user per month, with a confidence of 0.7. The responses differed on where the price floor sits, and the cited sources do not converge on a single entry-level figure.

The distinction worth noticing is that this is a pricing conflict rather than a recommendation conflict. It does not by itself move any brand's recommendation coverage, but it does mean that a buyer asking a straightforward cost question receives materially different answers depending on which AI platform they use.

Benchmark Context

Questions This Section Answers

• How many prompt-surface observations survive qualification into the public benchmark set?
• Which denominator is used for the brand-level recommendation percentages in this report?

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 moved from 385 in July to 454 in August, down to 423 in September, and 395 in October. Every brand-level percentage in this report is calculated against those qualified sets, not against the raw collection of 800 prompts.

Benchmark-Level Metrics

The recommendation-shaped answer share reversed direction sharply in October. After falling from 37.4% in July to 34.6% in August to 29.6% in September, it rose to 57.5% in October, the highest level in the series. The valid recommendation shortlist share recovered from its September low of 51.3% back to 65.3%, essentially matching July.

AI Recommendation Trend

Questions This Section Answers

• Which brands lead the October 2026 coverage ranking, and how wide is the gap between first and second?
• How many brands declined from their July 2026 baseline by more than normal month-to-month variation?

Microsoft SharePoint holds first place as most of the category recovers part of its earlier decline

Microsoft SharePoint leads the category at 50.1% valid recommendation coverage in October 2026. The gap to the next brand, Box at 40.2%, is 9.9 points. The middle tier now clusters between 40.2% and 22.5% coverage.

Five brands registered baseline-to-current declines above normal variation, and the October recovery from September was broad enough that the category-level change came from the combination of several smaller movements rather than a single dominant event. Dropbox Business was the only brand to exceed normal variation on a September-to-October basis, and Laserfiche was the only brand to fall out of its prior rank position relative to Dropbox Business.

What Changed This Month

Questions This Section Answers

• Which brand posted the only above-variation gain from September to October, and how did it move in rank?
• Which brand saw both its top-three and rank-one placement quality fall above normal variation?

Microsoft SharePoint

Microsoft SharePoint's valid recommendation coverage sits at 50.1% in October 2026, down 8.9 points from its July 2026 baseline of 59.0%, a decline above normal month-to-month variation. The path was front-loaded: a 17.4-point drop from July to August, a 3.5-point recovery in September, and a further 5.0-point gain from September to October. The brand is up two consecutive months, though neither gain was large enough to register above month-to-month variation on its own.

Raw mention presence fell 8.8 points from 86.8% in July to 78.0% in October. Top-three rate held effectively flat at 36.7% versus 37.1%, and rank-one rate moved from 23.6% to 24.1%. The absolute rank-one count held at 91 in July and 95 in October, and top-three count rose from 143 to 145.

The distinction to notice: SharePoint lost presence across the series while holding its premium placements, and the recovery from August lows came from coverage within a smaller qualified set rather than expanded presence. The brand's 50.1% coverage remains the highest in the category, but the 8.9-point gap to July has not closed.

Highest-priority diagnostic: Which prompt categories still place SharePoint first, and does the presence decline reflect a narrower footprint in the questions being asked rather than reduced recommendation strength?

Laserfiche

Laserfiche carries the largest baseline-to-current decline in the category, down 12.0 points from July 2026 (34.5%) to October 2026 (22.5%). The drop was concentrated in July to August at 14.9 points, with a slight further fall to 19.4% in September and a 3.1-point partial recovery in October. The brand fell from fifth to sixth place, overtaken by Dropbox Business.

Raw mention presence fell 12.3 points from 48.8% to 36.5%. Top-three rate declined 2.6 points from 10.7% to 8.1%, and rank-one rate fell 0.5 points from 2.3% to 1.8%. The absolute valid recommendation count fell from 133 in July to 89 in October.

Presence fell and recommendations fell in near lockstep, which points to reduced appearance in AI answers rather than a change in how the brand is framed when it appears. Its net sentiment score slipped from 0.9 to 0.8, remaining positive.

Highest-priority diagnostic: Which question categories no longer surface Laserfiche at all, and does the brand retain strength in prompts tied to its core segments like government or financial services?

Box

Box's valid recommendation coverage sits at 40.2% in October 2026, down 9.4 points from July 2026 (49.6%), a decline above normal variation. The movement was front-loaded: a 14.6-point drop from July to August, a further 1.2-point fall in September, then a 6.4-point gain from September to October. The October gain was just under the brand's variation threshold and does not register as significant on its own.

Raw mention presence fell 7.2 points from 71.2% to 64.0%. Top-three rate rose 4.7 points from 13.8% to 18.5%, and rank-one rate dipped 0.6 points from 3.4% to 2.8%. The absolute valid recommendation count fell from 191 in July to 159 in October, while the top-three count rose from 53 to 73.

The distinction: Box gave back overall coverage while improving its top-three placement rate, meaning when the brand is recommended, it now lands higher on average. The brand held second place throughout and ranks second in October with a 9.9-point gap to the leader.

Highest-priority diagnostic: Which prompts still produce top-three placements for Box, and where did the mid-tier recommendations that carried its July coverage go?

M-Files

M-Files' valid recommendation coverage sits at 38.7% in October 2026, down 8.1 points from July 2026 (46.8%), a decline above normal variation. The decline was concentrated in July to August at 11.3 points, followed by a further 2.6-point fall in September and a 5.8-point gain in October that came in just under the brand's variation threshold.

Raw mention presence fell 9.8 points from 68.0% to 58.2%. Top-three rate fell 7.8 points from 26.5% to 18.7%, and rank-one rate fell 3.5 points from 6.5% to 3.0%, both of which register above normal variation. The absolute valid recommendation count fell from 180 in July to 153 in October.

The distinction: M-Files is the only brand in the category where placement quality declined above normal variation on both the top-three and rank-one measures. Its average recommended rank worsened from 2.88 to 3.23, meaning when recommended, it now appears lower on average. The brand holds third place.

Highest-priority diagnostic: Which prompt patterns drove the top-three and rank-one declines, and does the brand still win top placements in use-case questions like document workflow automation?

Dropbox Business

Dropbox Business is the notable October mover, up 6.5 points from September 2026 (16.8%) to October 2026 (23.3%), a move above the brand's normal variation. Against the July baseline of 25.2%, coverage sits 1.9 points lower, which is within normal range. The brand climbed from sixth to fifth place, overtaking Laserfiche.

Raw mention presence held at 36.5% against the July baseline of 40.5%, a 4.0-point decline within normal range. Top-three rate rose 0.7 points from 11.2% to 11.9%, and rank-one rate rose 0.8 points from 1.0% to 1.8%. The absolute valid recommendation count fell from 97 in July to 92 in October.

The distinction: Dropbox Business is the only brand whose September-to-October gain registered above normal variation, and it did so while its presence held roughly steady. The gain came from its coverage within the qualified set rather than from a wider footprint, and its top-three and rank-one positions both improved modestly across the series.

Highest-priority diagnostic: Which prompts drove the October coverage gain, and does the brand win recommendations in cloud file-sharing questions rather than full document management asks?

DocuWare and the remaining field

DocuWare's valid recommendation coverage sits at 37.7% in October 2026, down 5.9 points from July 2026 (43.6%), below the brand's variation threshold. Within that movement, DocuWare holds the category's strongest rank-one rate at 6.8%, up 1.6 points from July, and its top-three rate rose 4.5 points to 24.8%. The brand sits fourth and posts the highest single-brand rank-one rate in October.

Revver (eFileCabinet) sits at 10.6% in October, down 7.1 points from July 2026 (17.7%), a decline above normal variation. Its raw mention presence fell 11.2 points from 23.6% to 12.4%. The absolute valid recommendation count fell from 68 in July to 42 in October. OpenText is effectively flat against baseline at 16.0% versus 15.3% in July, with presence up 5.6 points to 36.2% and top-three rate up 3.4 points to 8.3%, both within normal range over a two-month upward streak.

FileHold rose 4.3 points from July 2026 (1.0%) to October 2026 (5.3%), a significant riser now up three consecutive months, with 21 valid recommendations versus 4 in July. Raw mention presence rose 5.8 points from 1.8% to 7.6%. Templafy holds 0.0% coverage in October with no valid recommendations, after reaching 0.0% in September, with only 2 raw mentions and a net sentiment that fell to 0.0. The small absolute counts for FileHold and Templafy make their percentage movements directionally informative but not commercially material.

Highest-priority diagnostic: For DocuWare, whether its premium rank-one placements are concentrated in a narrow prompt set; for FileHold, which prompt patterns produce recommendations; and for Templafy, which prompts produced the vanished recommendations.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent cluster did every qualified observation fall into across the series?
• What can the benchmark not yet explain about how pricing or direct brand comparisons shape AI recommendations?

Across all four months of the series, every qualified observation fell into the Brand Recommendation cluster. There were zero qualified observations in either the Pricing & Value or Multi-Brand Comparison clusters in July, August, September, or October 2026. The public benchmark can therefore speak to which brands AI recommends for document management software, but it cannot yet answer how price, value, or direct brand comparisons shape those recommendations. The pricing inconsistency flagged earlier in this report surfaced in AI answers, but it did not resolve into a qualified pricing-cluster observation within the benchmark set.

Brand Opportunity Summary

Questions This Section Answers

• What is the highest-priority diagnostic for each brand based on its October 2026 coverage signal?
• Which brands recovered without returning to July levels, and which carry the largest unresolved decline?

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 capturing the 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. The pricing inconsistency flagged in this report illustrates the boundary: conflicting claims and cited sources can be detected at the answer level, but explaining which sources shaped which answers requires a company-level audit.

About This Benchmark

This report is part of the CiteWorks Studio AI Visibility Industry Market research program.

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

Report-Specific Interpretation Notes

• Small-count movement: FileHold holds 21 valid recommendations and Templafy holds 0 in October 2026. Their percentage movements are directionally informative but not commercially material.
• Qualified denominator: All coverage percentages use the qualified observation count (395 in October 2026) as the denominator, not the raw collection of 800 prompts.
• Directional analysis: Month-over-month movement identifies changes worth investigating; it does not by itself establish the cause of those changes. The benchmark describes the output distribution, not its cause.

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 is a brand winning, which competitor takes the recommendation when a brand loses, what attributes does AI associate with each option, and which external sources shape those answers. The October 2026 data raises these questions for the five significant decliners in document management software and for the brands that recovered part of their earlier losses without returning to July levels.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. Where the public benchmark identifies a decline or a recovery, the audit identifies the mechanism and the path forward.

Request an AI visibility audit