For consultants after the interviews
Interview Synthesis From Your Transcripts
Drop your interview transcripts or paste your notes. Get each theme with how many people raised it, and the quotes that prove it.
Free to start. No card needed.
| Theme | Raised by | In their words |
|---|---|---|
| Orders get lost between sales and the depots | 7 of 12 · every level | “We find out it’s sold when the truck is already late.” Depot manager, #4 |
| Leaders think the new system is live | Leaders 3 of 3 · staff 0 of 9 | “We still print the sheet.” Supervisor, #11 |
| Overtime is how targets get met | 4 of 5 supervisors | “Saturday is the plan, not the backup.” Supervisor, #9 |
Held back: pricing came up only after the interviewer raised it.
What you'll get
Themes from your interviews, with counts and quotes
- Every theme with how many people raised it, out of how many
- Quotes word for word, each with the interview it came from
- Where groups disagree, and what to ask in the next round
Before and after
Your interviews in. Findings you can defend.
Twelve Teams transcripts and three pages of typed notes from a logistics client: three executives, four depot managers and five shift supervisors.
1. Orders get lost between sales and the depots · 7 of 12, every level “We find out it’s sold when the truck is already late.” Depot manager, interview 4 “Sales promise Tuesday and nobody asks us.” Supervisor, interview 8 Executives and depot staff agree on this one. 2. Leaders think the new system is live · leaders 3 of 3, staff 0 of 9 “It’s rolled out everywhere.” Executive, interview 1 “We still print the sheet.” Supervisor, interview 11 Ask next round: which depots log orders in the new system today?
Counts and quotes side by side, so you can tell a pattern from one loud voice.
Why it works
What is interview synthesis, and how do you do it?
Interview synthesis turns a stack of interview notes into a short list of findings you can stand behind. River reads every transcript before it writes anything. Each theme then shows how many people raised it out of everyone you spoke to, quotes their words with the interview they came from, and notes where groups disagree. You get findings you can check line by line, instead of a summary that hides who said what and how often.
Counts matter because seven of twelve people is a pattern and one loud voice repeated is not. River separates what people raised on their own from what they agreed with after the interviewer suggested it, since a nod to a leading question is weak evidence. It flags themes that come from one team only and keeps every quote word for word. In a Harvard Business School study, consultants using AI did worse on a task built from interview notes and spreadsheets, so every point here shows its source.
Use it after a round of stakeholder interviews, customer calls or expert calls, while the conversations are still fresh. Drop up to ten transcripts from Zoom, Teams or Otter, or paste the notes you typed during each conversation. When the themes are clear, the storyline builder turns them into the answer first and a slide-by-slide outline. For a single call, the meeting minutes generator gives you the decisions and actions instead of themes.
What lands in your doc
What's in your interview synthesis
- A summary of how many interviews were read, from which groups, and the findings that matter most
- Each theme stated as a claim, with how many people raised it out of how many
- Quotes kept word for word, each tagged with the speaker's role and the interview number
- Where groups disagree, such as leaders and frontline staff, quoted side by side at equal length
- Themes held back because the interviewer suggested them, each with a neutral question to test it
- A theme table you can paste into a deck, plus the topics few people were asked about
How it works
From transcripts to findings
Drop your interviews
Drop up to ten transcripts or notes files, or paste your notes from each conversation.
River reads everything
It reads every interview in full before writing, noting who said what and who raised it first.
Get your synthesis
Themes land in your doc with counts, quotes and the interview each quote came from.
Take it further
Turn the themes into a storyline, a findings memo, or questions for your next round of interviews.
How to synthesize interviews by hand
Start by coding the transcripts. Nielsen Norman Group describes thematic analysis as tagging individual observations and quotations with codes so that significant themes can surface. Read each transcript closely and give every distinct point a short label, with the speaker's role and where it appears. Do this for every interview before you group anything, because the themes you expect after three interviews are rarely the ones that the twelfth interview actually supports.
Then group the labels into themes and name each theme as a claim, such as "orders get lost between sales and the depots", rather than a topic like "handoffs". Count how many people raised each one, and check how many were actually asked about it. Look at where each theme came from: if most mentions followed your own suggestion, or all of them come from one team, treat the theme as a question to test in the next round rather than a finding.
Last, pick the quotes. Keep them word for word, mark any trim, and attribute them by role rather than by name unless people agreed to be named. Watch out for small roles too, because the only finance director is identifiable however you label the quote. Share the strongest and most surprising themes with your client sponsor before the wider readout, so nothing lands on them for the first time in front of their own team.
Questions consultants ask
Common questions
What is interview synthesis?
Interview synthesis is the step after the interviews, when you turn what people said into a short list of findings. Good synthesis groups what you heard into themes, counts how many people raised each one, and keeps their exact words as evidence. It is the base for the recommendations you take to the client.
How many interviews can it read at once?
Drop up to ten files at a time, as Word, PDF or text, or paste notes from as many conversations as you like. The synthesis opens with how many interviews it read and which groups they came from. For a bigger programme, add the next batch in the same space and ask River to update the themes.
Will it keep my interviewees anonymous?
Quotes carry a role and an interview number, not a name, unless your notes say names are fine. River also flags any theme where a role is so small that a reader could guess who said it, such as the only finance director. You decide what goes to the client, and nothing leaves your space unless you share it.
How is this different from asking ChatGPT to summarize transcripts?
A summary tells you what was discussed. A synthesis tells you what holds across the interviews and how sure you can be. River counts how many people raised each theme, quotes them word for word with the interview they came from, and separates what people raised themselves from what they agreed to when asked. See River and ChatGPT side by side.
What should I do with the themes?
Test the strongest themes with your client sponsor before a wider readout, so nothing lands as a surprise. Then turn them into a storyline: the answer first, then the findings that support it. The storyline builder does that from the same space, or you can ask River for a findings memo in your own format.
Is my client's material private?
Files you drop in stay in your own River space unless you share them. Under River's privacy policy, nothing you upload trains AI models, and its AI providers are bound by their agreements with River not to train on it either. Check your client agreement for any rules about where its information can go.
Your next step
More for your project
Walk into the readout with evidence
Drop your transcripts or paste your notes. Every theme shows who raised it, so you can defend it in the room.