Shared research study link

Miro Collaboration Tool Customer Study

Understand how knowledge workers use digital whiteboards, what frustrations they have with collaboration tools, and what features matter most for remote team brainstorming.

Study Overview Updated Jan 28, 2026
Research question: Understand how knowledge workers use digital whiteboards, their collaboration frustrations, and which features matter most for remote brainstorming across access, AI, and switching criteria.
Research group: 6 remote knowledge workers (PM, government budget analyst, school counselor, construction manager, and developer/IT personas) spanning office, hybrid, and field contexts.
What they said: Biggest blockers are workflow failures-SSO/permission friction mid-session, latency/sync lag, multi-user cursor chaos, and canvases that don’t become durable, auditable artifacts-amplified by hybrid inequity, hardware mismatch, and compliance that pushes teams to “boring” formats like Slides/Docs/PDFs or repo-friendly text/diagram tools (e.g., Mermaid, VS Code Live Share, Procore/PlanGrid in the field).
AI sentiment: Wanted only when it eliminates grunt work via accurate, action-oriented summaries (decisions, owners, dates, links) under strict data governance; diagram generation and generic suggestions are low-value unless source-linked and near error-free, with modest, conditional willingness to pay (~$5–$8/seat or 5–10% uplift) after pilots prove minutes saved.

Main insights: Retention hinges on team adoption/network effects, integrations and data portability, reliability/performance (including offline/low-bandwidth), and admin/security/governance, with ease-of-use and search as the glue; price and templates are tiebreakers, while trust violations, performance hits, and forced redesigns trigger switching.
Product focus: Ship zero-friction access (guest tokens, SSO preflight), sub-150 ms realtime with calm facilitation controls, and a canvas lifecycle that yields durable artifacts (one-click exports to PDF/Markdown/SVG/Mermaid, versioning/diffs, board-to-ticket bridges to Jira/SharePoint/Drive).
Takeaways: Launch governance-first AI summaries with citations and push-to-systems, run short ROI pilots and bundle or price lightly; prioritize accessibility and multilingual fidelity, and for field/developer segments invest in offline-first mobile resilience and open, repo-friendly formats to win and keep teams.
Participant Snapshots
6 profiles
Jennifer Zapata
Jennifer Zapata

Jennifer Zapata, 35, is a Houston-based Senior Product Manager, married and a stepmother. A pragmatic, tech-savvy homeowner, she prioritizes durable, time-saving products and seamless integration, balancing career leadership with organized routines, fitness…

Eric Reynolds
Eric Reynolds

37-year-old St. Paul construction site lead. Single homeowner, no kids. Pragmatic, time-focused, and safety-first. Pays for durability, rejects hype. Enjoys woodworking, hockey, and North Shore trips. Values local trades, clear warranties, and reliability.

Marsha Casanova
Marsha Casanova

Bilingual K-12 curriculum leader in Lakewood city, CA. Married, childfree, high-income household, pragmatic and equity-focused. E-bikes locally, blends Latino and Hindu traditions, values measurable outcomes, privacy, and time-efficient, reliable products a…

April Carney
April Carney

Warm, organized NASA program analyst in rural Maryland; married with one child. Faith-forward, science-loving, and practical. Teleworks, sings in church, gardens, stargazes, and values reliability, time savings, and community impact.

Kamrin Smith
Kamrin Smith

Rural New York systems support pro, 33, single and uninsured, living simply with Kamrin's rescue cat. Frugal, privacy-minded, and community-oriented, Kamrin values durability and transparency, Kamrin bikes to town, and fixes both networks and neighborly pro…

Caroline Whitaker
Caroline Whitaker

A 28-year-old data engineer in Cary, NC, Caroline is values-driven, accessibility-minded, and financially savvy. She lives simply, bikes often, mentors others, and chooses durable, privacy-first products that reduce friction and support a calm, focused life.

Overview 0 participants
Sex / Gender
Race / Ethnicity
Locale (Top)
Occupations (Top)
Demographic Overview No agents selected
Age bucket Male count Female count
Participant locations No agents selected
Participant Incomes US benchmark scaled to group size
Income bucket Participants US households
Source: U.S. Census Bureau, 2022 ACS 1-year (Table B19001; >$200k evenly distributed for comparison)
Media Ingestion
Connections appear when personas follow many of the same sources, highlighting overlapping media diets.
Questions and Responses
3 questions
Response Summaries
3 questions
Word Cloud
Analyzing correlations…
Generating correlations…
Taking longer than usual
Persona Correlations
Analyzing correlations…

Overview

Across 18 respondents, knowledge workers converge on pragmatic, workflow-first requirements: friction around access/permission and SSO kills momentum; real-time latency and edit-conflict behavior are primary blockers for synchronous ideation; and whiteboards fail to deliver long-term value without durable, versioned exports and ticketable artifacts. Role and context shape priorities: field/mobile workers demand offline-first, rugged-device workflows and simple role models; developers require human-diffable plain-file exports and lightweight, keyboard-first clients; public-sector and education respondents foreground compliance, retention and approved storage; product/PM roles want deep two-way integrations and measurable time-savings from AI. Willingness to pay for AI is modest and conditional on measurable ROI, strong data governance, and exportability. Accessibility and multilingual fidelity are cross-cutting but disproportionately decisive for specific users.
Total responses: 18

Key Segments

Segment Attributes Insight Supporting Agents
Field / Mobile Operators
occupations
  • Construction Manager
locale
  • rural / outdoor sites
  • cold climates
age range
mid-30s
device profile
  • iPad/pencil
  • dusty Androids
  • spotty LTE
Offline resilience, fast image/photo markup, reliable single-file exports (PDF/CSV), and deterministic, simple role models are non-negotiable. Web-first canvases are perceived as fragile; willingness-to-pay is small and contingent on AI that works offline and saves admin time. Eric Reynolds
Developer / Technical-support Personas
occupations
  • IT Support Specialist
  • Computer/Math
locale
  • rural users on WISPs
  • cross-platform environments
preferences
  • Linux friendly
  • repo-centric workflows
income bracket
$25–199k (varies)
Priority is on plain-file, human-diffable exports (Mermaid/Markdown/SVG), lightweight/non-Electron clients, keyboard-first flows and transparent APIs. Developers distrust opaque AI; provenance, offline tolerance and integration to repos/VS Code matter more than canvas bells. Kamrin Smith, Caroline Whitaker
Public-sector / Government Professionals
occupations
  • Budget Analyst
industry
  • Government Administration
locale
  • rural/field-adjacent
age range
mid-40s
Compliance (FedRAMP/ATO/records retention), approved storage (SharePoint) and audit logs trump many UX gains. Procurement thresholds mean that without explicit governance assurances, adoption and budgets will not follow. April Carney
Education / School-adjacent Roles
occupations
  • School Counselor
language
  • Spanish speakers present
concerns
  • student data privacy
  • principal/teacher adoption
Simplicity for onboarding, exportable/durable artifacts for records, and strict student-privacy controls determine adoption. Multilingual/search/font fidelity (accents, diacritics) materially affects perceived value for bilingual users. Marsha Casanova
Product Managers / Mid-career Info Workers
age range
early-to-mid 30s
occupations
  • Product Manager
industry
  • Information Services
tool stack expectation
  • Jira
  • Notion
  • Figma
They prioritize deep, two-way integrations, low-latency multi-user editing, clear permission/admin ergonomics, and AI that produces measurable minutes saved (action items, owners). Payment is pragmatic and tied to demonstrated ROI. Jennifer Zapata
Accessibility- and Inclusion-focused Users
traits
  • hearing aids / reliance on captions
  • multilingual needs
occupations
  • Computer/Math
  • IT
  • Counselor
Accurate live captions, searchable transcripts, keyboard navigation and correct font/diacritic handling are high-value features that can justify departmental or personal spend; accessibility requirements often map directly to procurement decisions. Caroline Whitaker, Marsha Casanova

Shared Mindsets

Trait Signal Agents
Access & permissions friction SSO/guest flows, license limits and permission prompts repeatedly interrupt meetings and prevent spontaneous collaboration-solving these is a top adoption lever across roles. Jennifer Zapata, April Carney, Marsha Casanova, Kamrin Smith, Caroline Whitaker, Eric Reynolds
Real-time performance and latency sensitivity Even small cursor lag or 1–3s delays kills synchronous ideation; users prefer robust conflict-resolution primitives and low-latency experiences over richer but slower features. April Carney, Marsha Casanova, Jennifer Zapata, Eric Reynolds, Kamrin Smith
Canvas lifecycle / exportability Boards that cannot be exported to versioned, human-diffable artifacts are treated as ephemeral. Durable exports (PDF/Markdown/Mermaid/SVG) and ticket integrations are required for work to 'count' in many workflows. Kamrin Smith, Eric Reynolds, Jennifer Zapata, Marsha Casanova, Caroline Whitaker
Conditional, modest willingness-to-pay for AI AI features attract interest only if they measurably save time, respect data governance (no org-data training by default) and allow exportability; price tolerance is low and ROI must be demonstrable. Jennifer Zapata, Marsha Casanova, Caroline Whitaker, Eric Reynolds, April Carney, Kamrin Smith
Preference for pragmatic integrations over novelty Stable integrations with Jira/Slack/SharePoint/Drive and clean admin controls reduce fragmentation and friction more than flashy UI or experimental features. Jennifer Zapata, April Carney, Marsha Casanova, Caroline Whitaker
Hybrid / hardware equity concerns Hybrid meetings privilege physical-room participants unless tooling intentionally equalizes experience; mismatched input devices (stylus vs mouse) reduce sketch fidelity and downstream export quality. Jennifer Zapata, Marsha Casanova, April Carney

Divergences

Segment Contrast Agents
Field/mobile operators vs Product Managers Field users prioritize offline-first resilience and rugged-device support over deep SaaS integrations; PMs privilege two-way integrations and low-latency cloud collaboration. Eric Reynolds, Jennifer Zapata
Developers / IT vs Accessibility-focused Users Developers emphasize plain-text, repo-friendly exports and lightweight clients; accessibility-focused users prioritize accurate captions, transcripts and keyboard navigation-even willing to pay for those features. Kamrin Smith, Caroline Whitaker, Marsha Casanova
Public-sector / Government vs Private-sector Info Workers Government professionals require FedRAMP/ATO-level assurances, SharePoint retention and auditability ahead of UX improvements; private-sector workers may accept more functionality if governance barriers are lower. April Carney, Jennifer Zapata
Education (bilingual users) vs General cohort Bilingual/font/search fidelity (accents/diacritics) is a showstopper for some education users, whereas most other respondents note localization as preferable but not blocking. Marsha Casanova, Caroline Whitaker
Creating recommendations…
Generating recommendations…
Taking longer than usual
Recommendations & Next Steps
Preparing recommendations…

Overview

Respondents need collaboration that eliminates access friction, delivers low-latency sync, calms multi-user chaos, and turns canvases into durable, auditable artifacts tied to existing systems (Jira/SharePoint/Drive/git). AI is welcome only when it kills grunt work (actionable summaries with owners/dates/citations) under strict data governance. Retention hinges on integrations, reliability (incl. offline), security/compliance, and accessibility; price/templates are tiebreakers, not drivers.

Quick Wins (next 2–4 weeks)

# Action Why Owner Effort Impact
1 Zero-friction guest access and session preflight Permissions/SSO interruptions are a top flow-killer; a pre-join check and time-bound guest tokens prevent mid-call stalls. Platform Auth + Security Med High
2 Facilitator & Calm Mode Multi-cursor chaos forces turn-taking; add Follow Presenter, temporary object locks, spotlight pointer, and hide-cursors-on-demand to reduce cognitive load. Product + Client Eng + Design Low High
3 One-click Durable Export Pack Boards ‘rot’ unless exported; bundle PDF (by frames), Markdown decision log, CSV of stickies, SVG, and Mermaid where possible. Client Eng Med High
4 Board → Actions Bridge (no-AI v1) Traceability gap wastes time; map tagged items to Jira/Planner/Asana and Docs/Confluence with owners/dates/links. Integrations Eng + PM Med High
5 Join-health checks (network/device/permissions) Early detection of weak Wi‑Fi, blocked guests, or missing licenses reduces live-session churn. Client Eng Low Med
6 Accessibility & Multilingual fidelity pass Captioning, keyboard nav, high-contrast, and diacritic-safe fonts are decisive for adoption and compliance. Design Systems + Frontend Med Med

Initiatives (30–90 days)

# Initiative Description Owner Timeline Dependencies
1 Realtime Performance & Sync Overhaul Deliver sub-150ms p95 edits with resilient conflict resolution (CRDT tuning), server-side fanout, and edge presence. Add degraded/offline queues for low bandwidth and structured reconciliation to avoid dupes. Eng-Infra (Realtime) Phase 1 (0–3 mo) metrics & hotspots; Phase 2 (3–6 mo) engine revamp; Phase 3 (6–9 mo) global rollout Observability & perf budgets, Edge/CDN provider, CRDT/OT library upgrades, Load test harness
2 Permissions, Governance & Enterprise Readiness Time-bound guest tokens, domain allowlists, role templates, audit logs, retention/ediscovery, SCIM/SSO hardening, and approved storage (SharePoint/Drive) with autosave. Platform + Security/Compliance 0–3 mo essentials; 3–6 mo audit/retention; 6–9 mo ATO/FedRAMP-lite path IdP integrations (Okta/AAD), Legal/policy review, DLP/Retention services, Microsoft/Google storage APIs
3 Canvas Lifecycle → Durable Artifacts Versioning with named snapshots, diffs, branches, and commit-to-repo. Embed/sync with Confluence/Notion; webhook/API for CI to store artifacts next to code. Product + Integrations 0–3 mo exports + snapshots; 3–6 mo diffs/branches + repo sync Export pipeline (PDF/SVG/Markdown), Partner app embed APIs, Webhook & public API
4 Facilitated Collaboration Suite Agenda lanes, speaker queue, ‘parking lot’, frame navigator, and hybrid-equity features (room-camera capture, companion app). Improve stylus/trackpad smoothing and handwriting legibility. Design + Client Eng 0–2 mo MVP rails; 2–5 mo advanced facilitation Realtime engine stability, Analytics for facilitation usage, Mobile/room device support
5 AI Summaries with Governance-first Controls Accurate, action-oriented summaries with owners/dates, citations to source, push-to-Jira/Docs, and tenant isolation. Pilot to prove minutes saved; include captioning/transcripts and exportable logs. AI/ML + Compliance Pilot in 8–12 weeks; GA with controls in 4–6 months Transcription/caption provider, Policy toggles (no-training, retention), Action export integrations, Evaluation harness for accuracy
6 Mobile Offline & Field Workflow Lightweight mobile with offline persistence, PDF markups pinned to coordinates, photo notes, and clean sync. Target Procore/PlanGrid-style flows and spotty-LTE resilience. Mobile Eng Design + prototype 0–3 mo; beta 3–6 mo; GA 6–9 mo Sync engine from Realtime initiative, Procore/Drive integrations, Field QA in low-connectivity environments

KPIs to Track

# KPI Definition Target Frequency
1 Time-to-collaborate Median time from first invite to all intended participants actively editing in-session < 60 seconds Weekly
2 Access Interruption Rate Sessions with permission/SSO errors causing >1 minute delay ÷ total sessions < 2% Weekly
3 Realtime Latency p95 Client-perceived time from edit to remote render across 8–10 collaborators < 150 ms Daily
4 Board→Action Conversion Sessions that create ≥1 action with owner/date pushed to Jira/Planner within 24h ≥ 40% Weekly
5 Durable Export Adoption Percent of boards exported via bundle (PDF/Markdown/SVG/CSV) with re-open rate in downstream system within 7 days ≥ 70% re-open Monthly
6 AI Minutes Saved Self-reported and log-derived minutes saved per user per week from AI summaries with citations ≥ 45 min/user/week in pilot Monthly

Risks & Mitigations

# Risk Mitigation Owner
1 Compliance/ATO scope expands, delaying enterprise adoption Phase controls (no-training, retention, audit logs) early; pursue partner-hosted compliant environments; engage third-party assessors Security/Compliance
2 Underestimated complexity of realtime performance overhaul Establish perf budgets, build synthetic load tests, ship behind feature flags with cohort rollouts Eng-Infra (Realtime)
3 AI hallucinations erode trust Require citations to source, constrain to meeting artifacts, add confidence markers, and enable easy human review before push AI/ML PM
4 Integration fragility with Jira/SharePoint/Drive Use certified connectors, contract for partner support, add end-to-end monitors and backoff/retry queues Integrations Eng
5 Perceived data lock-in Ship open export bundle, stable public APIs, and migration guides; publish data ownership policy Product
6 Scope creep in facilitation features reduces usability Pilot minimal rails, instrument usage, prune non-performing controls, maintain ‘calm defaults’ Product + Design

Timeline

0–1 mo: ship quick wins (Facilitator/Calm Mode, preflight checks, early exports, accessibility fixes).

1–3 mo: guest access revamp (time-bound tokens), export bundle GA, initial Jira/Planner push, perf instrumentation.

3–6 mo: realtime engine upgrade rollout, snapshots/diffs, AI pilot with citations, SharePoint/Drive autosave, facilitation suite v1.

6–9 mo: governance package (audit/retention/SCIM hardening), mobile offline beta, repo sync/embeds, partner-certified connectors.

9–12 mo: enterprise compliance packages, offline GA, expanded integrations and performance hardening.
Research Study Narrative

Objective & Context

This study explored how knowledge workers use digital whiteboards for remote brainstorming, what breaks their flow, and which features drive adoption and retention. Across 18 respondents, patterns were consistent: practical workflow failures-not drawing tool gaps-limit value; AI is attractive only when it removes grunt work with strong governance; and retention hinges on integrations, reliability (including offline), security/compliance, and team adoption.

What We Learned (Cross‑Question Evidence)

  • Access/permissions are the top flow killers. “Permissions hell” mid‑call stalls momentum (Jennifer Zapata). Time‑bound guest links, clear SSO, and simple license models matter more than new shapes.
  • Low latency is essential for co‑creation. Even 1–3s delay collapses collaboration into awkward turn‑taking (April Carney). Users value robust conflict resolution over bigger canvases.
  • Multi‑user chaos requires facilitation rails. “Lag + chaos cursors” and “Where are you?” pings force groups into single‑facilitator mode (Marsha Casanova). Features like presenter follow, temporary locks, spotlight pointers, and agenda lanes reduce cognitive load.
  • Boards “rot” without durable artifacts. Teams often re‑create work in Slides/Docs/Markdown or images in tickets. One respondent noted, “If a brainstorm cannot survive as a plain file in the repo, it is theater, not planning” (Kamrin Smith). Exports, versioning, diffs, and ticket integration make outcomes auditable.
  • AI must save time, not add noise. Highest value: accurate, action‑oriented summaries with owners/dates/citations and push to Jira/Docs. Willingness to pay is modest and conditional-e.g., $5–$8/seat if it reliably saves 45–60 minutes/week (Jennifer). Strong governance is non‑negotiable: no training on org data by default, role‑based controls, audit trails, and 508 compliance (April). Safety contexts demand zero‑tolerance for misses (Eric Reynolds). Accessibility (captions/transcripts) can justify spend (Caroline Whitaker).
  • Retention drivers: team adoption/network effects (“Habits beat feature lists”-Marsha), deep integrations and data portability (“Integration first”-April), reliability including offline (“Works in a basement with no bars”-Eric), and ease/speed/search (“keyboard‑friendly and predictable”-Kamrin). Price/templates are tiebreakers.
  • Amplifiers and edge cases: hybrid inequity, mouse vs stylus mismatch; multilingual/search fidelity (diacritics) impacting accessibility and trust.

Persona Nuances

  • Field/Mobile operators: Offline‑first, fast photo markup, simple roles, clean PDF/CSV exports (Eric).
  • Developers/IT: Human‑diffable plain‑file exports (Markdown/Mermaid/SVG), lightweight clients, APIs, repo/VS Code integration; skepticism toward opaque AI (Kamrin).
  • Public sector: FedRAMP/ATO, retention/eDiscovery, SharePoint as source of truth, 508 accessibility (April).
  • Education/bilingual: Simple onboarding, records‑friendly exports, privacy, and accurate diacritics/search (Marsha).
  • PMs/info workers: Low‑latency multi‑edit, two‑way Jira/Confluence/Drive, and measurable AI time savings (Jennifer).

Implications & Recommendations

  • Eliminate access friction: Zero‑friction guest tokens, session preflight and join‑health checks to prevent SSO/license stalls.
  • Performance & sync overhaul: Sub‑150ms p95 edits across 8–10 collaborators with resilient conflict handling and degraded/offline queues.
  • Facilitated collaboration suite: Presenter follow, calm mode (hide cursors), temporary locks, agenda lanes, frame navigator; improve stylus/trackpad parity.
  • Make canvases durable: One‑click export bundle (frames‑to‑PDF, Markdown decision log, CSV of stickies, SVG, Mermaid), named snapshots, diffs, and links to tickets/docs.
  • AI with governance‑first controls: Actionable summaries with citations, owners/dates, push to Jira/Docs; tenant isolation, no training on org data by default; high‑accuracy captions/transcripts.

Risks & Measurement Guardrails

  • Risks: compliance scope creep (mitigate with early audit/retention controls), realtime complexity (stage rollouts with perf budgets), AI errors eroding trust (require citations and human review), brittle integrations (certified connectors, end‑to‑end monitors), perceived lock‑in (open exports, stable APIs).
  • KPIs:
    • Time‑to‑collaborate: median invite → all editing < 60s
    • Access interruption rate: < 2% of sessions
    • Realtime latency p95: < 150ms
    • Board→Action conversion: ≥ 40% within 24h
    • Durable export adoption: ≥ 70% re‑open in downstream systems within 7 days

Next Steps

  1. Ship guest access revamp and preflight/join‑health checks; add presenter follow and calm mode.
  2. Launch export bundle (PDF/Markdown/SVG/CSV) and initial Board→Jira/Planner push.
  3. Instrument and address realtime hotspots; begin phased engine upgrade.
  4. Pilot AI summaries with citations and strict data controls; include captions/transcripts.
  5. Advance governance (audit logs, retention, SCIM/SSO hardening) and SharePoint/Drive autosave.

Success equals fewer mid‑meeting stalls, snappier co‑editing, visible actions tied to systems of record, and exports that withstand audits-exactly what respondents said they need to stick and pay.

Recommended Follow-up Questions Updated Jan 28, 2026
  1. How often do you use a digital whiteboard for each activity? Activities: Brainstorming; Diagramming/architecture mapping; Sprint planning; Retrospectives; Research synthesis/affinity mapping; Stakeholder reviews; Project kickoffs; Training/workshops.
    matrix Quantifies core use cases to prioritize templates, UX, and scenario-focused features.
  2. In a typical live whiteboard session, how many people are actively editing at the same time?
    numeric Sets scale assumptions for performance, cursor handling, and collision-prevention design.
  3. What percentage of your whiteboard sessions include external guests (outside your organization)? Please enter 0–100.
    numeric Informs licensing, default guest settings, and investment in frictionless external access.
  4. Which access and permission capabilities are most and least critical for your team? Items: One-click guest link (no account); SSO auto-join; Domain allowlist; Expiring view-only link; Time-limited edit access; Role templates; Waiting room/admit; Join as viewer then request edit; Per-object permissions; Access audit log.
    maxdiff Prioritizes access features that prevent mid-session drop-offs and reduce IT burden.
  5. Which facilitation features best reduce multi-user chaos during live sessions? Items: Follow presenter; Hide/spotlight cursors; Lock board/objects; Turn-taking mode; Timer; Private ideation with timed reveal; Bring everyone to me; Section-based permissions; Attention nudge; Agenda/timer templates.
    maxdiff Directs facilitation roadmap to improve focus and coordination in real time.
  6. How important are the following post-session outputs and governance capabilities? Items: Snapshot versioning; Full change history with authors/timestamps; Export to PDF; Export to slides; Export to images; Push tasks to trackers (e.g., Jira/Asana/Trello); Send summary/actions to Slack/Teams; Embed in wiki (e.g., Confluence/Notion); Link board items to issues/PRs; Retention policies; E-discovery export; API/webhooks for archival; Comment resolution status; Approval/sign-off workflow.
    matrix Shapes artifact lifecycle, compliance posture, and integration backlog priorities.
For matrix items, use 5-point scales (e.g., Never–Daily for frequency; Not important–Critical for importance).
Study Overview Updated Jan 28, 2026
Research question: Understand how knowledge workers use digital whiteboards, their collaboration frustrations, and which features matter most for remote brainstorming across access, AI, and switching criteria.
Research group: 6 remote knowledge workers (PM, government budget analyst, school counselor, construction manager, and developer/IT personas) spanning office, hybrid, and field contexts.
What they said: Biggest blockers are workflow failures-SSO/permission friction mid-session, latency/sync lag, multi-user cursor chaos, and canvases that don’t become durable, auditable artifacts-amplified by hybrid inequity, hardware mismatch, and compliance that pushes teams to “boring” formats like Slides/Docs/PDFs or repo-friendly text/diagram tools (e.g., Mermaid, VS Code Live Share, Procore/PlanGrid in the field).
AI sentiment: Wanted only when it eliminates grunt work via accurate, action-oriented summaries (decisions, owners, dates, links) under strict data governance; diagram generation and generic suggestions are low-value unless source-linked and near error-free, with modest, conditional willingness to pay (~$5–$8/seat or 5–10% uplift) after pilots prove minutes saved.

Main insights: Retention hinges on team adoption/network effects, integrations and data portability, reliability/performance (including offline/low-bandwidth), and admin/security/governance, with ease-of-use and search as the glue; price and templates are tiebreakers, while trust violations, performance hits, and forced redesigns trigger switching.
Product focus: Ship zero-friction access (guest tokens, SSO preflight), sub-150 ms realtime with calm facilitation controls, and a canvas lifecycle that yields durable artifacts (one-click exports to PDF/Markdown/SVG/Mermaid, versioning/diffs, board-to-ticket bridges to Jira/SharePoint/Drive).
Takeaways: Launch governance-first AI summaries with citations and push-to-systems, run short ROI pilots and bundle or price lightly; prioritize accessibility and multilingual fidelity, and for field/developer segments invest in offline-first mobile resilience and open, repo-friendly formats to win and keep teams.