# Dovetail

> Dovetail organizes customer research and feedback with AI transcription, project summaries, searchable evidence, and channels for recurring signals.

Website: https://dovetail.com/

Dovetail is a customer-research workspace to evaluate when feedback needs to remain connected to its evidence. Consider it when your team repeatedly returns to the same research records and needs a consistent review process, rather than another one-off summary.

## Projects and recurring feedback

The vendor describes Projects with AI transcription, summaries, highlights or clips, tags, fields, and summary prompts. These are documented capabilities, not independently tested results. [Dovetail Projects](https://dovetail.com/product/ai-projects/)

Channels organizes recurring feedback such as tickets, reviews, surveys, and calls. Channels 2.0 is labeled Beta as of October 2026. [Dovetail Channels](https://dovetail.com/product/ai-channels/)

Choose the scope before you choose the surface. A bounded research packet and a continuing stream of support feedback need different source registers. In either case, keep the time window, export filters, and original record identifiers available to reviewers. A well-organized collection still needs an explanation of which customers and channels it covers.

## An illustrative workflow to evaluate

This is an **illustrative fictional example**, not a Dovetail trial. Suppose a team has an approved export of support tickets and survey comments about preparing weekly reviews. Begin by checking its identifiers, blanks, and duplicate records. Draft a codebook that distinguishes an export failure from difficulty finding a working export button.

Use a small reviewed set to evaluate the intended workflow. Check whether the reviewer can find the original comment, inspect its context, correct a label, and understand which codebook version was used. Record anything that the proposed setup cannot preserve. Then decide whether it supports the team's work before extending the process to a larger export.

Keep counting separate from interpretation. One person can submit several records; an anonymous comment cannot supply a known customer identity. The [customer-feedback guide](/guides/workflow/analyze-customer-feedback-with-ai) shows how to keep those counts visible and turn an observation into a bounded next check.

## Availability and review responsibility

Dovetail's current pricing page offers Free and custom Enterprise options; this directory classifies it as freemium. Check the chosen plan's terms. [Dovetail pricing](https://dovetail.com/pricing/)

Its AI-feature terms require user judgment of generated output. [Dovetail AI-feature terms](https://dovetail.com/legal/ai-features/)

Do not interpret the pricing category as entitlement to every described feature. Before bringing real records into a proposed setup, verify your organization's approval for those inputs and the arrangement you intend to use. This profile does not establish retention, contractual protections, or the correctness of generated labels.

## Fit and adjacent options

Use [Thematic Analysis](/glossary/thematic-analysis) to distinguish a label from an interpretation of meaning. The [Feedback Codebook Designer](/skills/product/feedback-codebook-designer) helps prepare label boundaries from supplied files in the existing Claude Code library; it does not connect to Dovetail.

[Fireflies](/tools/fireflies) serves a transcript-capture role, while Dovetail is framed around customer-research organization. A general assistant may suit a small ad hoc text review if the original records remain accessible. Evaluate these roles against the artifact you need, without assuming the products provide identical coverage or review controls.

---

_Source: https://agentscamp.com/tools/dovetail — Tool on AgentsCamp._