Dovetail
Data AnalysisUpdated October 21, 2026
Customer intelligence platform that transcribes interviews, tags research data and turns calls, tickets and surveys into cited AI insights.
We may earn a commission when you sign up through our links. It does not affect our reviews.
What is Dovetail?
Think of it as a shared memory for everything your customers say. Dovetail ingests interview recordings, usability sessions, sales calls, support tickets, app reviews and survey exports, transcribes and tags them, and lets anyone ask questions of that pile through an AI chat that cites the exact clip or ticket behind each answer. Projects handle deep qualitative work such as interviews and tests, while Channels classify high-volume feedback into ranked themes. Newer layers add Docs, scheduled Agents and Digital Twins, which are chat personas built only from your own customer data. The weak spot is access: since self-serve paid plans were retired, you either stay on a one-project Free plan or negotiate an Enterprise contract with credits. Pick it if research and product teams need one searchable evidence base, not a cheap transcription tool.
How to start using Dovetail?
Sign up for the Free plan with no card. You get one project, one channel and one dashboard, which is enough to see whether the transcription and tagging model fits your work. Inside the project, press Import or the plus button in the top right and pull files from your computer, Google Drive, Zoom or OneDrive. Video in mp4, mov, mpeg or avi and audio in mp3, m4a or wav is transcribed automatically, with the spoken language detected for you. Processing takes roughly 30% of the recording length, so a 60-minute interview is ready in about 20 minutes.
Your first real scenario should be a small study, such as five customer interviews on one topic. Let AI highlights mark key moments, then tag them with your own taxonomy. Highlights on audio and video become searchable clips, so a single quote can travel into a report or a highlight reel. If you run surveys, import a UTF-8 CSV with a header row; files are capped at 5,000 rows, and each respondent becomes one structured page.
The screen that matters most is Chat. Ask a question about one transcript, the whole project or, on Enterprise, a folder or the entire workspace, and the answer comes back with citations to the source data. From there you can turn the answer into a Doc such as a research brief or PRD draft. For recurring calls, set up calendar sync: Zoom works with Google or Outlook Calendar, Teams with Outlook, and Google Meet with Google Calendar, and recordings land in the project after the event ends.
Before sharing, check the AI work. Accept or edit suggested highlights, review the summary, and open the cited clips to confirm the quote says what the summary claims. Enterprise workspaces can redact PII in text, audio and video and restrict projects to specific people, which is worth configuring before stakeholders get access.
Who is it for?
UX researchers
Interview notes live in scattered docs and nobody finds last year’s findings. Dovetail turns each session into a transcript, tagged highlights and clips, and Chat lets stakeholders self-serve answers you already have.
Product managers
You need evidence for roadmap calls, not anecdotes from the loudest customer. Channels rank ideas by volume and, with Salesforce enrichment, by the ARR behind them, and one click pushes an idea to Jira, Linear or Claude with context attached.
Customer experience and VoC teams
Thousands of tickets, NPS comments and app reviews arrive every month. Sources such as Zendesk, Intercom, App Store, Google Play and Qualtrics feed a channel that classifies themes and sentiment continuously, so trends show up before they become escalations.
Designers
Concept feedback usually means recruiting another round of participants. A Digital Twin built from past interviews lets you pressure-test a direction against real customer evidence first, then save live sessions for the questions it cannot answer.
Research operations leads
Standardizing how dozens of people tag and report is hard. Global tags, fields, templates, folders and access controls on Enterprise keep projects comparable across teams.
Sales and customer success leaders
Gong calls hold objection patterns nobody has time to listen to. Bring them in as a source and ask Chat why deals stall on price, with answers tied to the exact calls.
Dovetail pricing
Dovetail now sells only two plans. Self-serve paid tiers, including Professional and Business, were retired and are closed to new subscribers, although existing customers keep them. Enterprise pricing has three parts billed on one invoice: the platform, the number of users, and a shared credit balance for data processing and AI. Each imported data point, such as a recording, ticket or CSV row, uses 1 credit, and applying redaction adds 1 more. A Chat response typically uses 1 to 10 credits, Deep research mode 5 to 30, and a scheduled agent run 5 to 50. Annual contracts are paid upfront, unused credits expire at the end of the contract, and there are no student, nonprofit or academic discounts.
One project, one channel and one dashboard, plus basic AI chat and summaries within that project. Version history, AI clustering, redaction and viewer roles are not included.
Custom. Pricing is quote-based and set by team size and credit volume. You get unlimited projects, channels, dashboards, Docs and Agents, advanced chat and summaries, semantic search summaries, AI clustering, global tags and templates, SSO, audit log, redaction, custom retention and a dedicated customer success manager. If you exceed your contract, overages are on by default at $0.15 per credit and $20 per additional user per month, prorated, and HIPAA is a paid add-on.
Dovetail key features
Projects for qualitative analysis
Upload interviews, usability tests, documents and survey CSVs, and Dovetail transcribes video and audio, extracts text from PDFs and slides via OCR, and suggests highlights. Tagged highlights become searchable clips you can stitch into reels. Stakeholders hear the customer directly instead of reading your paraphrase.
Chat with citations
Chat, powered by Claude according to Dovetail’s documentation, answers plain-language questions across a transcript, project, channel or workspace and links each claim to its source. You can turn an answer into a Doc or an agent without leaving the conversation.
Channels for high-volume feedback
Connect Intercom, Zendesk, Freshdesk, ServiceNow, Gong, Slack, App Store, Google Play, Pendo, Qualtrics or Snowflake, and incoming items are classified into themes and sentiment automatically. You can add custom topics on top of AI themes and see heat maps over time. Channels 2.0 is currently in open beta.
Agents and Docs
Agents run on a schedule, a webhook or an in-app event, watch your data and send write-ups to Slack, Microsoft Teams, email or a Doc. Docs draft VoC reports, PRDs and briefs from your evidence with embedded quotes. Weekly digests replace the readout meeting nobody can attend.
Digital Twins
A twin models a customer segment using only first-party calls, tickets, surveys and research in your workspace, never public web data. You can query it in Chat, Slack or Teams, including by voice, and every answer traces back to a real conversation. It cannot be built without existing data.
Integrations, API and MCP
Beyond the feedback sources, Dovetail connects to Jira, Linear, Productboard, Notion, Figma Make, Zoom, Teams and Google Meet, and syncs contacts with Salesforce and HubSpot. An API, CLI, Zapier source and an MCP server let Claude, ChatGPT or Microsoft Copilot read your Dovetail data.
Security and governance
Dovetail states SOC 2 Type II and ISO 27001 certification, GDPR compliance and data residency in the US, Europe or Australia, and says customer data is never used to train models. Enterprise adds PII redaction across text, audio and video, custom retention and audit logs that include MCP and API calls.
Dovetail Pros & Cons
Pros
Reviewers consistently praise accurate transcripts and the way tagged highlights turn hours of video into shareable clips that land well with executives. Having every interview, ticket and survey in one searchable place means past research stops disappearing when a researcher leaves. Chat answers come with citations, so you can check a claim against the actual quote rather than trusting a summary. The integration list covers the tools most product orgs already run, from Zoom and Gong to Zendesk and Snowflake, and the MCP server lets teams query research from Claude or ChatGPT. Security credentials and regional data residency make procurement easier in regulated companies. Start with one well-tagged project so stakeholders see the value before you connect noisy sources.
Cons
The biggest drawback is the pricing gap: with self-serve plans gone, there is nothing between a one-project Free plan and an annual Enterprise contract. Credits are shared across the whole workspace with no per-user or per-feature budgets, and Slack or Teams questions from non-users still draw from that balance. Several G2 reviewers report rising prices, a learning curve, and UI changes that disrupted repositories they had set up years ago. Opinions on AI accuracy vary, and splitting feedback by product inside one workspace takes tuning. Ask sales for a credit estimate based on your real data volumes and keep overage alerts on before rolling out Chat company-wide.
Use Dovetail if you want…
- Transcribe and tag customer interviews in one place.
- Turn hours of research video into shareable highlight reels.
- Ask questions of your research and get cited answers.
- Classify support tickets and app reviews into themes automatically.
- Rank product ideas by the revenue behind them.
- Send weekly feedback digests to Slack without manual work.
- Test concepts against Digital Twins built from real customers.
- Give stakeholders self-serve access to past research findings.
Dovetail alternatives
Dovetail Reviews
Product teams like that they can connect a data source and feed in raw feedback without building integrations, but splitting signals per product inside one workspace is hard and tuning the AI takes trial and error.
Researchers rate the transcripts as very accurate and find tagging plus AI chat useful for pulling quotes, though interviews still have to be recorded elsewhere and the recent pricing change makes cost a bigger factor.
Long-time users value having interviews and contextual inquiries in one repository, yet some complain that prices keep climbing and frequent UI changes broke the way they first organized their projects.
Clips and highlight reels are described as the most persuasive way to share findings with senior stakeholders, while one daily user wished search could be limited to a single project instead of the whole workspace.
Capterra reviewers call it a solid central home for notes, transcripts and tags, but mention the lack of per-recording access limits for sensitive calls and weaker transcription in some languages such as Hindi.
FAQ
Is there a free plan?
Yes. The Free plan costs $0 with no card and includes one project, one channel and one dashboard. If a workspace drops to Free with more than one project, it becomes read-only until you are within the limit.
Can I still buy a self-serve paid plan?
No. Professional and Business plans were retired for new subscribers, so the only upgrade from Free is an Enterprise plan through sales. Existing legacy customers keep their plans unless they cancel or switch.
How do Dovetail credits work?
Each imported data point uses 1 credit and redaction adds 1 more, while Chat, agents and Digital Twins use a variable amount depending on the work. Search, including semantic search, does not use credits, and failed or canceled actions are not charged.
Is my data used to train AI models?
Dovetail says customer data is never used to train its models and that processing runs on its own AWS infrastructure. It also states SOC 2 Type II and ISO 27001 certification and GDPR compliance.
What file types can I import?
Video in mp4, mov, mpeg and avi, audio in mp3, m4a and wav, documents such as PDF, Word, PowerPoint, Keynote and Pages, plus CSV survey files up to 5,000 rows. Zoom, Teams and Google Meet recordings can be imported automatically through calendar sync.
Does Dovetail support languages other than English?
Transcription detects the spoken language automatically, and translation of transcripts and summaries into 75 languages is available in beta on Enterprise. AI clustering works best in English but supports a dozen other languages.
Does Dovetail have an API or MCP server?
Yes. There is a Dovetail API, a CLI, a Zapier source and an MCP server that lets Claude, ChatGPT and Microsoft Copilot access your workspace data.
What happens if I cancel?
You keep access until the end of the paid period, with no refund for the remaining time. The workspace then enters a 30-day read-only window for exporting data before deletion under your retention policy.
Similar tools
HexCollaborative analytics workspace that mixes SQL, Python and AI agents in notebooks, then publishes results as data apps and plain-language answers.
WindwardMaritime AI platform that fuses AIS, satellite and ownership data into vessel risk scores, sanctions screening, alerts and container ETAs.
Delve AIAI market research platform that builds personas from analytics, CRM and social data, then lets you chat with them or run synthetic