Strategy & Planning

How to synthesize customer research interviews as DTC Brand Founders

Strategy & PlanningFor DTC Brand Founders4 apps10 steps~20 min to set up

You finish a 45-minute customer interview, flip back to Slack, and the insight is already half-forgotten. Your notes are scattered across Otter exports, a Notion doc nobody updates, and a voice memo you'll never re-listen to. You've run maybe 30 interviews over the past year — between your Shopify launch, a Meta creative test, and a packaging rebrand — and when you try to write your quarterly board update or brief a new agency, you're reconstructing themes from memory. You can't tell your investors which objections keep coming up, which UGC angles customers actually described, or whether 'shipping speed' or 'ingredient quality' is the real retention driver. The research happened. The synthesis never did.

Build this on Starch → See the apps used

Strategy & PlanningFor DTC Brand Founders4 apps10 steps~20 min to set up

Outcome

What you'll set up

A searchable archive of every customer interview, auto-tagged by theme, channel mention, product feedback, and sentiment — so next time you brief a creative agency you pull the file, not your memory

A weekly synthesis digest that surfaces recurring patterns across interviews (what customers keep saying about your unboxing, your subscription cancel flow, your hero SKU) without you rebuilding a spreadsheet every time

An auto-generated board-ready slide or summary pulling from your synthesized research, your Shopify order data, and your growth metrics — so your quarterly narrative has receipts

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

Meeting Notes Knowledge Management Growth Analyst Presentation Agent

Data sources & config

Meeting Notes captures and transcribes customer calls in real time. Summaries and tagged excerpts are stored in Knowledge Management, which connects to Notion via Starch's scheduled sync — your Notion databases refresh on a schedule so nothing is lost. Growth Analyst connects to PostHog from Starch's integration catalog, queried live when the weekly digest runs, and sends its output to Gmail (also a scheduled-sync connection). Presentation Agent pulls from your Knowledge Management archive and Growth Analyst digests to build slides. Slack is connected from Starch's integration catalog for automated summary delivery.

Prompts to copy

Transcribe this customer interview recording, extract every mention of shipping speed, packaging, and cancellation reasons, and tag them by sentiment. Add the summary to my Customer Research database under the tag 'subscription retention.'

Across all customer interviews tagged 'subscription retention' in the last 90 days, what are the top 5 recurring themes? Which ones appear in more than 3 separate conversations? Show me the exact quotes.

Every Friday, pull the three most common objections mentioned in new customer interviews this week, cross-reference them with this week's Growth Analyst digest from PostHog, and send me a summary Slack message.

Build me a 6-slide board update summarizing our customer research themes from Q1, our top acquisition channels from the Growth Analyst weekly digest, and what we're testing next quarter based on customer feedback.

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

Walkthrough

Step-by-step

1 Connect Google Calendar and Zoom (or Google Meet) from Starch's integration catalog so Meeting Notes can detect upcoming customer calls automatically and start transcription without you manually triggering it.

2 After each interview, Meeting Notes generates a transcript, a 3-paragraph summary, and extracted action items. Tell Starch: 'Tag this transcript by product line mentioned, purchase driver, and any cancellation or churn language, then save it to my Customer Research Notion database.'

3 Starch syncs your Notion databases on a schedule — every new tagged interview entry lands in your Knowledge Management archive, searchable by theme, date, customer segment, or specific phrase within 24 hours.

4 Set up a weekly automation: 'Every Sunday at 6pm, scan all customer interviews added this week, identify themes that appear in 2 or more conversations, and post a bullet summary to the #founder-inbox Slack channel.' This replaces the Sunday-night manual re-read.

5 When you're prepping a creative brief for your Meta ad agency, ask Knowledge Management: 'What exact phrases did customers use to describe why they first bought our hero SKU?' Get verbatims, not paraphrases.

6 Growth Analyst runs on a schedule and emails you a weekly digest from PostHog — signup trends, top referral sources, conversion rate by channel. Cross-reference this with your interview themes to see whether what customers say in interviews matches where they actually convert.

7 Before a board meeting or investor update, tell Presentation Agent: 'Build a 6-slide deck summarizing our top 3 customer research themes from Q1 2026, our best-performing acquisition channel from the Growth Analyst digest, and our two biggest product bets for Q2.' It returns a complete draft in minutes.

8 If a customer submits a refund request or cancellation that contains detailed feedback, tell Starch: 'Read this support email, extract any product or experience feedback, and log it to the Customer Research Notion database with the tag churn-signal.' This keeps your research corpus growing even between formal interview cycles.

9 When you're running a new packaging test or UGC creative batch, search your Knowledge Management archive for every interview where customers described the unboxing experience — pull exact quotes to brief your creative team without scheduling another research round.

10 Set a quarterly review automation: 'On the first Monday of each new quarter, pull all interview themes from the past 90 days, count frequency by tag, and draft a one-page executive summary I can paste into my board update doc.' Starch drafts it; you edit and send.

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

Q1 2026 Subscription Retention Research Synthesis

Sample numbers from a real run

Customer interviews completed (Jan–Mar 2026)	14
Interviews mentioning 'shipping delay' as cancel reason	6
Interviews mentioning 'flavor fatigue' as cancel reason	9
Interviews mentioning UGC or word-of-mouth as acquisition driver	11
PostHog data — subscription page conversion rate, Feb vs. Jan	-4
Net new subscribers, Q1 2026 (from Shopify via browser automation)	312

You ran 14 customer interviews between January and March 2026 — mostly Zoom calls with repeat buyers and recent churners. Without Starch, those 14 transcripts lived in Otter, scattered across your Downloads folder. With Meeting Notes running, every call is transcribed and auto-tagged. When you ask Knowledge Management to synthesize Q1 retention themes, it surfaces two things you half-suspected but couldn't prove: 9 of 14 customers mentioned 'flavor fatigue' (you only had 3 SKUs in the subscription rotation), and 11 of 14 said they found you through a friend or Instagram post — not paid ads. The Growth Analyst weekly digest from PostHog shows your subscription landing page conversion dropped 4% in February, the same month you paused your influencer seeding program. Suddenly you have a hypothesis: the channel that actually acquires your best subscribers is also the channel you cut when CAC got tight. Presentation Agent builds your Q2 board slide in 8 minutes: 'Flavor fatigue is the #1 churn driver in customer interviews (9/14 mentions). Expanding the rotation from 3 SKUs to 6 is Q2 priority 1. Influencer/word-of-mouth is the top acquisition driver (11/14 interviews) — restarting seeding budget in April.' Your investors have the receipts. You didn't spend a Sunday rebuilding it.

Measurement

How you'll know it's working

Subscription retention rate by cohort (30/60/90-day) — because interview themes need to map to whether people actually stay

Top-mentioned cancel reason frequency across interviews (tracked in Knowledge Management, updated after each new interview batch)

Customer-reported acquisition channel vs. paid attribution — how often what customers say (word-of-mouth, friend referral) diverges from what Meta Ads Manager reports

Time from interview conducted to insight in board deck — the metric Starch directly compresses

Number of creative briefs or product decisions with a documented customer quote attached — proxy for whether research is actually being used

Comparison

What this replaces

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

Otter.ai + manual Notion synthesis

Otter transcribes but does nothing with the transcript — you still spend 30 minutes per interview pulling themes into Notion by hand, and your archive is only as good as the last time you felt like doing it.

Dovetail

Purpose-built for UX research synthesis and genuinely strong at tagging and theming, but it's a standalone research tool that doesn't connect to your Shopify data, PostHog metrics, or board slide workflow — so you're still assembling context manually at presentation time.

Notion AI

Works well if all your interviews already live in Notion, but it doesn't transcribe calls, doesn't pull in your growth metrics, and won't auto-generate a slide deck — you're still the integration layer between your research and your reporting.

Hiring a part-time research contractor

A good researcher will synthesize better than any AI on nuance, but at $50–80/hr they're expensive for 14 interviews a quarter, they go dark between sprints, and your findings still live in a doc that doesn't talk to your financials or board prep.

On Starch RECOMMENDED

One platform — meeting notes, knowledge management, growth analyst 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 customer interviews are a mix of Zoom recordings, voice memos on my phone, and some that are just my own rough notes. Can Starch handle all of those?

Meeting Notes works directly with Zoom and Google Meet calls — those get transcribed automatically. For voice memos or uploaded recordings, you can feed them to Starch and ask it to transcribe and tag them the same way. For calls that are just your rough notes, you can paste the notes in and tell Starch to structure them into the same format as a transcript summary. The output — tagged themes, quotes, and a structured entry in your Knowledge Management archive — is consistent regardless of input type.

Will my customer interview transcripts be stored somewhere secure? These are conversations with real customers.

Starch stores your data, but it's worth knowing we're not SOC 2 Type II certified yet — that's a real limitation to weigh if your customers are enterprise buyers with vendor security requirements. For most DTC founders doing qualitative interviews with consumers, this is less of a blocker. Don't share anything in Starch you'd be uncomfortable with on a non-certified SaaS platform, and check in with your own privacy policy language around how you handle interview data.

I don't use PostHog — I'm on Google Analytics. Can Growth Analyst still work for me?

Growth Analyst's pre-built template connects to PostHog. If you're on Google Analytics 4, you can connect it from Starch's integration catalog — the agent queries it live when your app runs. You'd describe what you want to Starch ('pull weekly sessions, conversion rate by source, and top landing pages from Google Analytics 4 and email me a digest every Friday') and Starch builds you a custom version. It won't be the same out-of-the-box experience as the Growth Analyst starter app, but the underlying capability is there.

I run a lot of informal DMs and voice notes with customers on Instagram — is that reachable?

Instagram's API doesn't give third-party apps access to DMs. But Starch can automate Instagram through your browser — no API needed — meaning it can navigate to your DMs, read conversations, and extract the content for you. This is browser automation, not a formal integration, so it works the way a human clicking through Instagram would. Set it up with a prompt like: 'Check my Instagram DMs weekly, extract any messages that mention product feedback or complaints, and log them to my Customer Research Notion database.'

Is Presentation Agent available right now? I have a board meeting in three weeks.

Presentation Agent is currently in development — you can request beta access to get notified when it launches. For your board meeting in three weeks, the practical path is to use Starch to synthesize your research themes and draft the narrative, then paste the output into Google Slides yourself. It's not as fast as a one-click deck, but the synthesis work — pulling themes from 14 interviews, cross-referencing growth data, writing the executive summary — that part Starch can do today.

How do I keep the Knowledge Management archive from getting stale after the initial setup?

Set a recurring automation in Starch: 'After every customer call logged by Meeting Notes, automatically tag the summary and add it to my Customer Research Notion database.' Because Starch syncs your Notion on a schedule, the archive stays current without you manually pushing updates. You can also set a quarterly prompt: 'Flag any customer research entries in Notion that are older than 6 months and haven't been referenced in a synthesis — I may want to archive or re-validate them.'