A typical qualitative study generates fifteen-plus hours of transcript before anyone has learned anything useful. The synthesis phase, reading, tagging, clustering, writing up, is where the research budget quietly disappears. Dovetail is an AI customer intelligence platform that absorbs that work, turning raw transcripts and ongoing feedback into structured insight reports automatically.
A typical user research study, fifteen interviews at one hour each, generates fifteen hours of transcript before anyone has learned anything actionable. Turning that into something useful, reading all of it, tagging significant moments, grouping those tags into recurring themes, writing a summary, and building a presentation that makes the insight legible to a product team or a CMO, is where the research budget actually goes. ZipRecruiter's July 2026 data puts the average UX researcher salary at $54.38 per hour in the United States, which means the synthesis phase alone, typically 20 to 40 hours on a mid-complexity study, costs between $1,080 and $2,160 before any other project expenses. For teams that commission outside firms, full-service qualitative projects run $15,000 to $30,000 for a 20-person interview study. The conversations are the cheap part. The analysis is where the weeks go.
Dovetail is an AI customer intelligence platform built to absorb that synthesis work. You bring the raw material: interview transcripts, call recordings, survey responses, support tickets, product reviews. Dovetail's AI reads it, applies tags to significant moments, clusters related tags into themes automatically, and generates structured summaries at the project and workspace level. The work that took a researcher three weeks to produce manually arrives the next morning.
What the synthesis workflow actually costs
The manual synthesis process is not glamorous work, and it is not fast. A researcher working through 15 interview transcripts will typically spend two to three hours reading each one carefully before tagging begins. Tagging itself, selecting quotes, labeling them by topic or sentiment, creating new codes for patterns that don't fit existing categories, adds another six to ten hours. Grouping tags into themes, finding the structure in what looked like 300 unrelated data points, takes another three to five hours of pattern recognition across the whole dataset. Writing the insight report, with summaries, evidence quotes, and implications for each theme, takes the remaining time.
Across four qualitative studies per year, a team pays between $4,320 and $8,640 in UX researcher time on synthesis alone, not counting participant recruitment, moderation, transcription, or the project management time between. Research firms embed all of that into their project fees, which is why a 20-person study lands at $15,000 to $30,000 even when the interviews themselves only require 20 hours of total moderation.
What Dovetail does with the data
The platform's AI Analysis feature handles the tagging and clustering steps directly. You upload transcripts or connect a recording source, and the AI applies codes to text highlights across the dataset, then groups related codes into themes without manual sorting. Summaries are generated at the theme level and the project level, with the supporting quotes embedded as evidence rather than leaving the reader to find them.
Dovetail also offers what it calls Channels: live connections to ongoing feedback sources like support ticket queues, app store reviews, and survey pipelines that automatically classify new data as it arrives. Instead of commissioning a quarterly research project and waiting six weeks for a debrief, the intelligence stays current. An AI dashboard visualizes how themes are trending over time alongside business metrics. The shift is from research as a periodic event to research as continuous infrastructure.
Dovetail's pricing page shows a free plan at no cost with one project, one channel, and basic AI summaries, with no credit card required. The Professional tier runs $15 per user per month following a March 2026 pricing update from a prior $39, and covers unlimited projects, unlimited channels, advanced AI summaries, semantic search, and transcription. A 60-day unlimited trial is available. Enterprise pricing is custom and adds SSO, compliance controls, audit logging, and dedicated support.
The cost comparison with real numbers
For a two-person research team running four qualitative studies per year, synthesis work at $54 per hour across 30 hours per study comes to $6,480 annually in researcher time before any project overhead. A two-seat Professional plan at $15 per user per month costs $360 per year.
The point of comparison is not that Dovetail replaces the researcher entirely. It replaces the 30-hour synthesis task, freeing the same researcher to run more studies, work on more complex interpretation, or spend time on the parts of research that still require human judgment. A team that currently manages two qualitative projects per quarter because synthesis blocks the calendar could, in theory, manage twice that volume without adding headcount.
Who this is wrong for
Dovetail synthesizes research you have already collected. The platform has nothing to work with until there are transcripts, recordings, or structured feedback to analyze. Teams that haven't yet run the research, and need help designing studies, recruiting participants, moderating sessions, or transcribing audio, are looking at a different set of tools.
The platform is also the wrong choice when the interpretation of raw data requires domain expertise, not pattern recognition. Clinical research, compliance interviews, or any study where understanding what a participant meant matters as much as what they said benefits from an analyst who can apply professional judgment to ambiguous responses. Dovetail can identify that eight participants used the word "confusing" near mentions of the pricing page. It cannot tell you whether that confusion reflects a UI problem, a pricing architecture mismatch, or an expectation set incorrectly in the sales process. That reading requires context the platform doesn't hold.
Organizations with strict data residency requirements, HIPAA-regulated environments, or legal constraints on where customer voice data can be processed should review Dovetail's data practices before loading sensitive interview content into the platform.
What makes the problem structural
Research that never gets synthesized is just expensive audio files. Most organizations do not have a shortage of willingness to run qualitative studies. They have a shortage of bandwidth to turn transcripts into something actionable before the decision that motivated the research has already been made without it. The insight tends to arrive as justification for a conclusion already reached, rather than as input to one still forming. Closing the synthesis gap does not guarantee the insight lands in time to change anything. But it shifts the odds.