What Is Dovetail?: And How Does it Compare to Sprig

Example H2

Introduction

Dovetail is a customer intelligence platform that collects qualitative customer evidence from interviews, sales calls, support tickets, and research studies, and uses AI to turn it into searchable themes, cited answers, and shareable insights.

Product managers most often use Dovetail as a research repository: the place to check what customers have already said before commissioning new work.

Dovetail analyzes evidence that other tools collect. Dovetail does not field surveys, recruit participants, or run interviews itself, so product teams typically pair it with a quantitative platform such as Sprig to measure how many customers a theme affects.

This guide covers:

Define Dovetail in its own current terms, then explain why the repository label still fits
Walk through how Dovetail ingests, analyzes, and distributes customer evidence
Show what Dovetail leaves to other tools, and name the tool that covers each job
Give product managers a diagnostic for sorting repository questions from measurement questions
Work one example end to end, from a repository theme to a sized survey result

Dovetail capabilities, ratings, and documentation cited here were verified on September 24, 2026 against Dovetail's site and developer documentation, Wikipedia, Sprig's site, G2, Capterra, and TrustRadius. No prices appear anywhere in this guide.

How Dovetail describes itself in 2026

Dovetail describes itself as an AI-first customer intelligence platform.

The homepage headline reads "Build with facts, not vibes," followed by "One platform for every customer signal. Dovetail's AI turns sales calls, support tickets, and research into answers your whole organization can act on."

The wording is recent, since Dovetail announced the customer intelligence positioning on October 8, 2025, expanding a product that Wikipedia describes as a tool for transcribing, coding, and tagging interviews and survey responses.

Why the repository label still fits

The repository has not gone away, and Dovetail still sells it as a named solution, headlined "You've answered this question before.

Your team is about to ask it again." G2 also continues to list Dovetail in its User Research Repositories category.

The customer intelligence platform is best read as the repository with more intake channels and more AI on top, rather than as a different product.

The core job is unchanged: store customer evidence once, then let anyone in the company find it again.

Who makes Dovetail

Dovetail was founded in 2017 in Sydney, Australia, by Benjamin Humphrey and Bradley Ayers, both formerly of Atlassian. Ayers left the company in January 2024.

The company raised a Series A of 63 million US dollars led by Accel in January 2022, according to Wikipedia's account of its funding history.

Humphrey, chief executive at the time of the October 2025 launch, framed the repositioning this way:

"Building great products has never been the job of one team. Sales, success, product, and design all bring critical signals, but too often those signals are fragmented or lost."

The quote names sales, success, product, and design as sources of customer evidence. Dovetail's homepage now addresses product managers, designers, researchers, customer experience, sales, and marketing teams.

How Dovetail works

Dovetail works in three stages: evidence comes in from other systems, AI and human analysts turn it into themes and insights, and those insights go back out to the tools where decisions happen.

Each stage is described below using Dovetail's own product names.

Getting customer evidence into Dovetail

Dovetail lists 38 native integrations on its homepage, plus an API, a Model Context Protocol (MCP) server, and a command-line tool.

Dovetail's July 14, 2026 launch added 18 more, including Qualtrics, Salesforce Service Cloud, Pendo, and PostHog.

Typical inputs include:

Interview recordings and transcripts
Sales call recordings
Support tickets
Survey exports
Research study files
Product feedback from integrated tools

Survey data arrives as a CSV file.

Dovetail's documentation says each row becomes a single respondent document, with one column chosen as the title and the rest imported as content.

Analyzing evidence inside Dovetail

Dovetail's analysis layer combines manual research practice with AI assistance. Researchers can still tag and highlight passages by hand alongside the AI features.

The named features, as of September 2026:

AI Chat
AI Docs
AI Agents
AI Dashboards (beta)
Channels 2.0 (closed beta)
Digital Twins
Search and Explore

AI Chat answers questions with citations back to the underlying evidence.

AI Docs drafts product requirement documents, voice of customer reports, and summaries with citations, according to Dovetail's October 2025 launch post. AI Docs output is a draft, and a product manager remains responsible for opening the cited evidence and confirming each claim before the document circulates.

Channels 2.0 classifies high-volume feedback into themes and generates what Dovetail calls "Ideas, backed by customer evidence." Digital Twins lets teams query AI replicas of customer segments built from their own data, and this guide does not evaluate it.

AI Agents run on events, schedules, or webhooks. Dovetail says agents can be set to require human approval before they take a write action, and teams that let agents edit a shared repository should turn that approval setting on.

Getting insights out of Dovetail

Rather than asking colleagues to log into another tool, Dovetail pushes evidence into the tools where product work happens. Its repository page names Jira, product briefs, and planning documents, and its July 2026 launch added MCP connectors for Linear, Slack, Notion, Canva, and Snowflake.

Dovetail's developer documentation lists roughly 50 MCP tools covering projects, data, docs, highlights, tags, channels, and files. Authentication uses OAuth 2.1 or a bearer token, and Dovetail lists built-in connections for ChatGPT, Claude, Figma, and Copilot.

Dovetail as a research repository for product managers

A research repository is a shared, searchable store of customer evidence that outlives the study that produced it.

For a product manager, Dovetail is most useful in that role: the first place to check before asking a research team, or a customer, the same question twice.

Historically, research findings lived in slide decks and shared drives.

A product manager starting a new initiative would commission a study that a colleague had already run a year earlier, because nobody could find the old one.

What product managers use Dovetail for

Dovetail's page for product managers promises to help teams "prioritize with confidence, write briefs grounded in evidence, and ship what customers actually need." Four jobs commonly follow from that promise:

Checking prior research before scoping a new feature
Pulling cited customer quotes into a product brief
Tracking recurring complaints across support and sales channels
Sharing evidence with engineering and design without sending raw recordings

Each of those jobs depends on the evidence already being in Dovetail, which is why its 38 native integrations generally matter more to a product team than any single AI feature.

What Dovetail claims for product teams

Dovetail's product manager page cites a 2.3x return on investment, 30 hours saved per user each week, 66 percent faster shipping from discovery to delivery, and payback in under six months.

Those figures are Dovetail's own, and the page does not publish the methodology or baseline behind them.

Treat them as vendor claims rather than benchmarks, and ask for the underlying study before relying on them. A product team evaluating Dovetail can also run a time-to-insight comparison on its own backlog during evaluation.

Where the repository model strains

Dovetail's G2 reviewers name the strain directly. Limited advanced features, complex tagging systems, and cross-project data management are among the most frequent complaints, while ease of use and AI synthesis draw the most frequent praise.

How product teams typically set up a Dovetail repository

A Dovetail repository generally works when a team makes three decisions before the first import: who can see what, how evidence arrives, and how it is tagged. Each decision maps to something Dovetail's own pages or reviewers describe.

Set access before the first import

Dovetail's repository page describes granular access controls that protect raw data while sharing insights. A product team should decide which roles can see recordings and personal details before inviting the wider company, rather than after a transcript circulates beyond the people who needed it.

Connect intake before inviting the team

Dovetail lists 38 native integrations covering sources such as sales calls, support tickets, and survey tools. Connecting the sources a product team relies on most often, before anyone else is invited, means the first search returns current evidence rather than an empty project.

Agree a short tag list up front

G2 reviewers cite complex tagging as one of Dovetail's most frequent complaints. Product teams that agree a short list of themes, product areas, and customer segments, and review it each quarter, typically avoid the drift those reviewers describe.

What Dovetail leaves to other tools

Dovetail does not collect new evidence on its own. Its July 2026 launch material describes a platform that analyzes existing customer data, and it does not field surveys, recruit participants, or conduct interviews.

The collection boundary is a design choice rather than a shortcoming, and it defines what a product manager pairs with Dovetail.

Dovetail does not field surveys

Dovetail imports survey results but does not write, target, or deliver a survey. A product manager who needs a new measurement uses a survey platform to field it and then brings the results into the repository.

The right tool is a survey platform with the distribution the question requires: in-product for current users, email for a customer list, or a panel for people outside the customer base.

Dovetail does not size a finding against a known population

Dovetail's survey documentation describes AI theme summaries and "a clear overview of quantitative results." It does not document crosstabs, quotas, sampling controls, or significance testing.

The right tool is a survey platform that targets a defined population and reports results by segment. A repository can show that a complaint recurs. It cannot say what share of active users it affects.

Dovetail does not recruit or moderate

Dovetail does not recruit participants or moderate interviews. The right tool is a recruiting service or an interview platform, with the recordings and transcripts sent to Dovetail afterward.

Dovetail lists Outset, an AI-moderated interview platform, among its integrations, so interview sessions can run in one tool while the evidence lands in Dovetail.

Repository question or measurement question?

A product manager should decide which kind of question is being asked before choosing a tool.

A repository question asks what customers have said and why. A measurement question asks how many, which segment, or whether something changed.

The table below routes seven common question shapes.

The pattern is consistent: repository questions are about meaning, and measurement questions are about incidence. The checkout example later in this guide runs one question through both, starting in the repository.

How to read the diagnostic

A question that starts with "why" or "what do customers say" usually belongs in the repository first. A question that starts with "how many," "which," or "did it change" needs new data collected from a defined population.

Two warning signs suggest a product manager is using the wrong instrument. The first is quoting a single vivid interview as if it described the whole user base.

The second is running a survey to discover a problem nobody has described yet, which tends to produce answer options that miss the real issue.

How a repository and a quantitative platform divide the work

Dovetail and a survey platform do different jobs, and a product team gets more from both by assigning each job explicitly. The repository holds the explanation, while the survey platform measures the scale.

Both Dovetail and Sprig connect to AI clients such as Claude and ChatGPT through MCP, but they expose different data. Dovetail exposes the repository, while Sprig exposes studies, responses, and themes from fielded surveys.

Where Sprig fits

Sprig is the quantitative counterpart to a repository like Dovetail. Its homepage headline reads "Enterprise surveys. Powered by agents," and it fields studies inside web and mobile products, by email, through shareable links, and to external research panels.

In-product targeting is the part that matters most to a product manager. A Sprig study can trigger on a product event and filter by user attributes, so a question about checkout reaches people who just used checkout rather than a general customer list.

Panels support more than 300 targeting attributes for audiences outside the customer base.

Sprig runs three named agents across the research lifecycle. The Design Agent drafts a programmed study from a research goal, the Field Agent adds real-time follow-up questions, and the Synthesize Agent turns responses into themes with supporting quotes.

Researchers remain responsible for approving every study and interpreting what the results mean.

Where Sprig stops

Sprig is not a repository. Rather than offering a company-wide store for interviews, sales calls, and support tickets, Sprig keeps its own study results and supports cross-study queries through MCP.

Sprig also does not publish its weighting approach, sample-size guidance, or significance-testing methodology. A product team that needs documented statistical inference should confirm those methods with Sprig before relying on a high-stakes number.

Worked example: sizing a checkout theme

The checkout example follows one question from the repository to a sized result and back. The company, figures, and theme are illustrative.

A product manager at a subscription software company opens Dovetail before planning the next quarter.

AI Chat, asked about checkout complaints, points to 14 support tickets and three interview transcripts. The product manager reads the cited sources and confirms the theme: customers are confused about when a plan upgrade takes effect.

Step 1: Read what the repository says

The repository answers the "why," and the quotes show customers expect an upgrade to apply immediately and are surprised by a prorated charge at the next billing date.

What the repository cannot answer is scale, since fourteen tickets could represent a widespread problem or a small, vocal group.

Step 2: Write the measurement question

The product manager turns the theme into a measurement question: what share of customers who upgraded in the last 30 days were unsure when the change took effect?

The survey uses the customers' own words from the repository. Rather than inventing answer options, the product manager writes them from the highlighted quotes, which is the main reason to consult the repository first.

Step 3: Field it to the right population

The study runs as an in-product survey, triggered after a customer views the billing page, filtered to accounts that upgraded in the last 30 days. It asks one closed question, one follow-up, and one open-text question.

Targeting on behavior is what makes the result usable, because a general email survey would mix customers who never upgraded with those who did.

Step 4: Read the result by segment

Suppose 400 eligible customers respond and 31 percent say they were unsure when the upgrade took effect.

The textbook margin of error at this sample size is roughly 4.5 points at 95 percent confidence, but that formula assumes a random sample. In-product respondents select themselves, so the figure describes the customers who answered rather than every customer who upgraded.

Split by plan, the share is higher on monthly plans than annual plans. That segment difference is the finding the repository could not produce.

Step 5: Send the result back to the repository

Rather than living in a slide deck, the survey results go into Dovetail through the integration, where they sit beside the original tickets and transcripts. The next product manager who asks about checkout confusion finds both the explanation and the size.

Connecting Sprig and Dovetail

Sprig sends survey responses to Dovetail through a Zapier connection. The integration moves responses in one direction, from Sprig into Dovetail, and it is not a native connector.

Setup takes five steps, according to Sprig's integration page:

Create a Zapier account
In the Sprig app, open Integrations, then Zapier, then Get API Keys
Identify the environment, development or production
Copy the environment ID for a later step
Complete the configuration using Zapier's Dovetail documentation

Rather than exporting and importing CSV files after every study, a team can route responses automatically as they arrive. Teams should still confirm which fields map into Dovetail during setup, since a Zapier workflow only carries what it is configured to send.

Common mistakes when pairing a repository with surveys

Three mistakes commonly undermine a program that combines Dovetail with a survey platform, and each has a correction.

Treating a recurring theme as a measured result

A theme that appears in many tickets feels like a large problem, but ticket volume typically reflects who contacts support rather than who is affected.

Rather than reporting ticket counts as incidence, size the theme with a survey to a defined population before it goes on a roadmap slide.

Writing survey options without reading the repository

Survey answer options drafted from a product manager's assumptions often miss the language customers actually use. Reading the highlighted quotes first, and writing options from them, generally produces options that respondents recognize.

Leaving survey results outside the repository

Survey results that stay in a separate dashboard are frequently invisible to the next person who searches the repository. Sending results into Dovetail after each study keeps the explanation and the measurement in the tool the next person searches.

Where Dovetail is the stronger choice

Dovetail is the stronger choice for any job centered on qualitative evidence that already exists. Two dimensions stand out, and the independent ratings follow.

Qualitative synthesis across sources

Dovetail brings transcripts, sales calls, support tickets, and research studies into one analysis layer. Sprig has no equivalent for recordings it did not collect.

Dovetail wins qualitative synthesis outright, and a product team with a large interview archive should generally start there.

Reusing past research

Dovetail is built for finding evidence again, months after a study closes. Its search, AI Chat, and cited answers are designed around the repository job.

Dovetail wins research reuse outright, since Sprig is not designed to be the place a company searches for past interviews.

Ratings on independent review sites

Dovetail holds a G2 rating of 4.5 out of 5 across 169 reviews in the User Research Repositories category, 4.6 out of 5 across 97 reviews on Capterra, and 8.7 out of 10 across 101 reviews on TrustRadius, all retrieved September 24, 2026.

Sprig holds 4.3 out of 5 across 199 reviews on G2 in the Survey Software category and 8.5 out of 10 across 10 reviews on TrustRadius, as of September 2026.

The G2 figures sit in different categories, so they are not a like-for-like comparison, and Sprig's 10 TrustRadius reviews are too few to compare against Dovetail's 101.

A checklist for evaluating Dovetail as a product repository

Before committing to Dovetail as a product team's repository, six checks generally surface the fit questions that a demo does not:

Confirm that the integrations your team depends on are generally available rather than in beta
Confirm how tagging, projects, and AI Chat behave on a repository the size of yours
Confirm which roles can see raw recordings, transcripts, and personal details in each project
Confirm how survey results arrive in Dovetail, and which fields survive the import
Confirm what the MCP server exposes and which administrative controls govern it
Confirm the methodology behind any return-on-investment figure presented during evaluation

The MCP check matters for teams with strict data policies. Dovetail's MCP server authenticates with OAuth 2.1 or a bearer token, and a self-hosted option is available.

The documentation page reviewed for this guide does not cover administrative controls, so teams connecting AI clients to customer recordings should ask Dovetail how access is scoped and revoked.

Bottom line

Dovetail is best for product teams that already generate qualitative evidence through interviews, sales calls, and support, and need one place to find it again. It is a repository and an analysis layer, not a data-collection tool.

If your team's gap is remembering and reusing what customers have said, Dovetail is the right starting point. If your team's gap is knowing how many customers a problem affects, and in which segment, a quantitative platform such as Sprig fills it.

Most product teams increasingly need both, with Sprig's Zapier integration sending measured results into Dovetail beside the evidence that prompted them.

Frequently asked questions

What is Dovetail used for?

Dovetail is used to store, analyze, and share qualitative customer evidence. Teams bring in interview transcripts, sales calls, support tickets, and survey exports, then use tagging, AI Chat, and AI Docs to find themes and cite them in product decisions. Product managers most often use it as a repository to check before starting new research.

Is Dovetail a research repository?

Yes, Dovetail began as a research repository and still sells one as a named solution, although the company now describes the whole product as an AI-first customer intelligence platform. G2 lists Dovetail in its User Research Repositories category, and the repository page targets product managers and designers as well as researchers.

Is Dovetail a survey tool?

No, Dovetail does not create or distribute surveys. It imports survey results, typically as a CSV file where each row becomes a respondent document, and summarizes them alongside other evidence. Fielding a survey to a defined population requires a separate survey platform, such as Sprig, whose results can then be sent into Dovetail.

Does Dovetail have an MCP server?

Yes, Dovetail publishes an official Model Context Protocol (MCP) server with roughly 50 tools covering projects, data, docs, highlights, and tags, authenticated by OAuth 2.1 or a bearer token. Dovetail lists built-in connections for ChatGPT, Claude, Figma, and Copilot, and other clients can connect to the hosted endpoint.

What is the difference between a research repository and a survey platform?

A research repository stores and organizes evidence that already exists, such as interviews, support tickets, and past studies. A survey platform collects new, structured answers from a defined population. Dovetail is a research repository and Sprig is a survey platform, so the two typically work in sequence rather than as substitutes.

How do you import survey data into Dovetail?

Survey data is typically imported into Dovetail as a CSV file. Each row becomes one respondent document, with one column chosen as the title and the remaining columns imported as content, and Dovetail's AI then summarizes themes across all respondents. Sprig responses can also arrive automatically through Zapier.

How much does Dovetail cost?

Dovetail publishes its plans and packaging on its own site, and this guide does not state prices. Capterra reviewers frequently describe Dovetail's pricing as high and mention seat minimums, so a product team should confirm per-seat terms, contributor limits, and which AI features each plan includes during evaluation.

What are the alternatives to Dovetail?

Dovetail's closest alternatives are other research repositories, which G2 groups in its User Research Repositories category, including tools such as Marvin. Sprig is a survey platform rather than a Dovetail alternative, and product teams typically use a survey platform alongside a repository instead of in place of one.

Who founded Dovetail?

Benjamin Humphrey and Bradley Ayers founded Dovetail in 2017 in Sydney, Australia. Both previously worked at Atlassian. Humphrey was chief executive at the October 2025 launch, and Ayers left the company in January 2024.

Can Dovetail replace a survey platform for product managers?

No, Dovetail cannot replace a survey platform, because it does not collect new data from a defined population. Product managers generally use Dovetail to understand why a problem happens and a survey platform to measure how many customers it affects.

How do Sprig and Dovetail work together?

Sprig sends survey responses into Dovetail through a Zapier connection, one way, so measured results sit beside interviews and tickets in the repository. A product manager can find a theme in Dovetail, size it with a Sprig in-product survey, and store the result back in Dovetail.

What do reviewers criticize about Dovetail?

Dovetail's G2 reviewers most often criticize complex tagging systems, difficulty managing data across projects, and limited advanced features. The same reviewers most frequently praise ease of use and the quality of its AI synthesis.