Strategy & Planning

How to synthesize customer research interviews as Professional Services Founders

Strategy & PlanningFor Professional Services Founders3 apps10 steps~20 min to set up

After six client discovery calls in a week, you have six Otter.ai transcripts, three sets of handwritten notes, and a Notion doc someone half-filled in. Synthesizing that into themes — what clients actually want, what objections keep surfacing, what language they use to describe their own problems — takes a senior consultant two to three hours. That person is also on three active engagements. So the synthesis either happens late, happens shallow, or happens in your head on a Friday afternoon before the proposal is due Monday. You miss the pattern that would have sharpened the pitch, and the proposal reads like every other proposal you've sent.

Build this on Starch → See the apps used

Strategy & PlanningFor Professional Services Founders3 apps10 steps~20 min to set up

Outcome

What you'll set up

A structured synthesis app that pulls interview transcripts, extracts recurring themes and verbatim quotes, and organizes findings by client segment — so a 6-interview batch takes 20 minutes instead of 3 hours

A searchable research archive tied to your Notion knowledge base, so next quarter's discovery can draw on this quarter's findings without anyone having to remember where they put things

An automated brief-to-deck pipeline that turns your synthesized findings into a polished client-facing presentation, ready to customize before the proposal meeting

The Starch recipe

Apps, data, and prompts

The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.

Apps used

Knowledge Management Meeting Notes Presentation Agent

Data sources & config

Meeting Notes transcribes calls live; Starch connects directly to Notion (scheduled-sync provider) to store and retrieve past research; Starch connects directly to Gmail (scheduled-sync provider) so follow-up briefs can be emailed to clients; any recordings stored in Google Drive are reachable through Starch's integration catalog queried live when the synthesis app runs. Presentation Agent (currently in beta — request access to be notified at launch) builds the client-facing deck from the synthesis output.

Prompts to copy

Summarize these six discovery call transcripts. Extract the top 5 recurring themes, pull 2-3 verbatim quotes per theme, flag any objections that came up more than twice, and group findings by client type — founder-led vs. enterprise team.

Take the synthesized research themes from my Q1 discovery calls and build a 10-slide findings brief: one slide per major theme, each with a client quote, the pattern we observed, and one implication for how we should position the engagement.

Add these synthesis findings to my Notion knowledge base under 'Client Research / Q2 2026' and tag each theme with the relevant service line so I can search for them later.

Run these in Starch → or paste them into your favorite agent

Walkthrough

Step-by-step

1 Connect Notion and Gmail to Starch — both sync on a schedule so your existing research library and email threads are available without manual export.

2 Install the Meeting Notes app from the App Store. For future discovery calls, Meeting Notes transcribes in real time and generates a structured summary — decisions, highlights, open questions — automatically after the call.

3 For existing transcripts (Otter.ai exports, Zoom recordings, Word docs), paste them or upload them directly. Tell Starch: 'I have six discovery call transcripts from March. Synthesize the key themes, pull verbatim quotes for each, flag repeated objections, and note which clients said what.'

4 Starch groups findings by theme across all six interviews — not summarizing each call individually, but identifying the cross-call pattern. You see 'pricing transparency' came up in four of six calls, with the exact words clients used.

5 Tell Starch: 'Add these findings to my Notion research database under Q2 2026 / Discovery, tagged by service line.' Starch writes the structured entry directly into Notion so your team can search it.

6 For the proposal, tell Starch: 'Build a 10-slide findings brief for the Meridian Group engagement: cover the three themes most relevant to operations consulting, include their specific quotes, and end with our recommended engagement framing.' Presentation Agent (beta) assembles the deck.

7 Before the proposal meeting, pull related past research: 'Show me any discovery findings from the last 12 months where clients mentioned change management or stakeholder buy-in.' Starch searches your Notion archive and surfaces the relevant entries.

8 After the meeting, Meeting Notes captures the proposal review conversation. Action items are extracted and assigned — no separate follow-up email needed.

9 Set up a recurring synthesis prompt for end-of-quarter: 'Every quarter, pull all discovery call summaries tagged in Notion from the past 90 days, identify any new themes or shifts in client language, and email me a one-page brief.' Starch runs this on schedule.

10 Over time, your research archive compounds. New team members can search it instead of asking you. The next proposal can reference patterns from six prior engagements instead of one senior consultant's memory.

See this running on Starch

Connect your tools, describe what you want, and the agent builds it. Closed beta is free.

Try it on Starch →

Worked example

Meridian Group Proposal — March 2026

Sample numbers from a real run

Discovery calls conducted	6
Hours of transcript (raw)	9
Themes extracted by Starch	7
Verbatim quotes surfaced	22
Synthesis time (actual)	0.33
Senior consultant hours saved	2.5

You run six stakeholder interviews across Meridian Group's ops and finance teams in the first two weeks of March — 9 hours of recorded conversation. Normally your senior consultant would spend half a day reading transcripts and building a themes doc before you could write the proposal. Instead, you paste all six Otter.ai exports into Starch and type: 'Synthesize these transcripts. Top themes, verbatim quotes per theme, objections that came up more than once, and flag any tension between what ops and finance said about the same topic.' Twenty minutes later you have a structured brief: 7 themes, 22 quotes sorted by theme, and a callout that ops leadership said 'we don't have visibility' in four separate calls while finance said 'the data exists, nobody looks at it.' That tension becomes the opening slide of the proposal. You tell Starch: 'Build a 10-slide findings brief for this engagement. Theme one is the visibility gap — lead with the ops vs. finance contrast and use these two quotes.' The deck is ready to review in under 15 minutes. The proposal goes out Sunday night instead of Monday afternoon. You win the engagement. The findings are already in Notion for the next time a prospect mentions operational visibility.

Measurement

How you'll know it's working

Hours from last discovery call to proposal sent (target: under 48 hours)

Proposal win rate by engagement type (tracked in HubSpot, surfaced in Starch)

Percentage of proposals that include verbatim client language from discovery

Research reuse rate — how often past synthesis findings are cited in new proposals

Senior consultant time spent on synthesis tasks per month

Comparison

What this replaces

The other ways teams handle this today, and how the Starch version compares.

Otter.ai + manual Notion synthesis

Otter transcribes well but the synthesis — finding cross-call patterns, pulling quotes by theme, flagging objections — is still fully manual; you're doing the analysis yourself, Otter just saves you typing.

Dovetail

Purpose-built for UX research synthesis with strong tagging and affinity mapping, but it's another standalone tool to maintain, doesn't connect to your proposal workflow or Notion knowledge base, and is priced and designed for dedicated research teams rather than a 12-person consultancy where synthesis is one of thirty things the same person does.

ChatGPT / Claude (manual paste)

Works for one-off synthesis if you paste the transcripts yourself, but nothing is saved, searchable, or connected to your Notion archive or HubSpot pipeline — so the insight dies in a chat window and you start from zero next quarter.

Notion AI

Good at summarizing a single document or page you're already in, but won't synthesize across six separate transcript files, won't extract cross-call themes, and won't wire findings into a proposal or presentation without more manual steps.

On Starch RECOMMENDED

One platform — knowledge management, meeting notes, presentation agent all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.

Try it on Starch →

FAQ

Frequently asked questions

My transcripts are in Otter.ai, some are in Zoom, and a few are Word docs from an old-school client who records on a phone. Can Starch handle all of those?

Yes. Paste or upload transcript text directly into Starch regardless of source format. Starch isn't transcribing the audio — it's synthesizing the text. However you get the text out of Otter, Zoom, or Word, Starch can work with it. The Meeting Notes app handles real-time transcription for future calls.

Will the findings actually go into my Notion, or is this just a Starch-internal thing?

Starch connects directly to Notion (scheduled-sync provider), so when you tell Starch to save findings to a specific database or page, it writes there. Your team can access the research in Notion exactly as they normally would — they don't need to be Starch users.

Presentation Agent says 'currently in development / request beta access.' When is it actually available?

Presentation Agent is not live today. You can request beta access and Starch will notify you when it launches. In the meantime, Starch can produce a structured written brief — theme by theme, with quotes and implications — that you or your team can drop into Google Slides or PowerPoint in significantly less time than building from scratch.

Is this only useful for the proposal phase, or does it help with ongoing client work?

Both. During active engagements, Meeting Notes captures stakeholder interviews, workshops, and check-ins. You can ask Starch to pull every mention of a specific concern across all meeting summaries for a client — useful for mid-engagement reviews or when a new team member needs to get up to speed fast.

What if I have confidentiality concerns about storing client interview content?

Worth knowing: Starch is not SOC 2 Type II certified today. If your clients have strict data handling requirements in their contracts, check before connecting transcripts that include sensitive information. For engagements where that's a concern, you can still use Starch for the synthesis structure and prompting while keeping raw transcripts local — paste only the sections you're comfortable with.

My team uses HubSpot to track which proposals we sent. Can the synthesis link back to the deal?

Starch connects to HubSpot (scheduled-sync provider), so you can pull deal context — who you met with, what stage the opportunity is in, what the estimated value is — and include it in the synthesis brief or proposal. Closing the loop the other way (writing back a note to the HubSpot deal record) depends on how you've configured your HubSpot access, but the data is available to pull into your Starch apps.