Internal Comms & Meetings

How to run a team retrospective as Small RevOps Teams

Internal Comms & MeetingsFor Small RevOps Teams4 apps10 steps~20 min to set up

Your retrospectives happen on a Friday afternoon when everyone's already mentally checked out. You paste Slack threads, HubSpot deal notes, and Apollo sequence data into a Google Doc, try to reconstruct what actually happened across a 30-rep org this quarter, and end up with a document nobody reads. Action items from last retro? Buried in someone's inbox. The CRO asks 'what did we decide about territory overlap last cycle?' and you're digging through 12 tabs. You're a 2-person team — there's no Chief of Staff to own this, and you can't afford to lose another 3 hours to a meeting that produces a Doc that decays immediately.

Build this on Starch → See the apps used

Internal Comms & MeetingsFor Small RevOps Teams4 apps10 steps~20 min to set up

Outcome

What you'll set up

A running retrospective log that pulls deal outcomes from HubSpot and Apollo sequences directly into your meeting notes — so 'what worked in the pipeline this quarter' has actual data behind it, not vibes

Auto-extracted action items from each retro session assigned to the right rep or RevOps owner, tracked in a task or project view so nothing falls through after the call ends

A searchable archive of every past retro decision — so when the CRO asks what you decided about enterprise territory splits in Q1, you find the exact paragraph in under 30 seconds

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 Task Manager Knowledge Management Project Management

Data sources & config

Starch connects directly to HubSpot (scheduled sync) and Apollo.io (scheduled sync) so deal outcomes, sequence performance, and pipeline snapshots are available as context inside your meeting notes and knowledge base. Google Calendar is synced on a schedule to auto-detect retro meetings and pre-populate the meeting context. Gmail is connected via scheduled sync to pull relevant email threads into the meeting prep. Salesforce and Pipedrive are reachable from Starch's integration catalog and queried live when your apps need pipeline data from them.

Prompts to copy

Transcribe today's RevOps retro and generate a summary that includes: key pipeline decisions, territory or quota changes we agreed on, and any process changes for how reps should log activities in HubSpot. Extract action items and assign them by name.

After the retro ends, create tasks for each action item with the assignee, a P1 or P2 priority level, and a due date two weeks out. Flag any that don't have a clear owner.

Save today's retrospective summary to our RevOps knowledge base under 'Quarterly Retrospectives / Q2 2026' and tag it with: territory, quota model, HubSpot hygiene, and pipeline review.

Search our past retros for every decision we've made about opportunity stage definitions — pull the relevant excerpts and show me which quarter each one came from.

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

Walkthrough

Step-by-step

1 Connect HubSpot via scheduled sync and Salesforce or Pipedrive from Starch's integration catalog — this gives your retro prep actual deal-stage data instead of asking reps what closed.

2 Start a Meeting Notes session before the retro begins. Tell Starch: 'This is our Q2 RevOps retrospective. Pre-load open and closed deal counts from HubSpot this quarter, top Apollo sequences by reply rate, and any opportunities that went stale without activity in the last 30 days.'

3 During the call, let Meeting Notes transcribe in real time. You run the meeting — you're not also typing a doc.

4 After the call, prompt Starch: 'Generate a summary with key decisions, what we're changing about how reps log HubSpot activities, and any territory model updates we agreed on.' Review and edit in under 5 minutes.

5 Prompt Starch to extract action items: 'Pull every action item from today's retro, assign each one to the person named in the transcript, set priority P1 or P2 based on urgency, and give each a two-week deadline unless we specified otherwise.'

6 Each action item lands in Task Manager or a Project Management board — your RevOps project, not a Slack thread that disappears. You can filter by owner and see who still has open items from last retro.

7 Save the retro summary to Knowledge Management: 'Archive this retro under RevOps / Retrospectives / Q2 2026 and tag it: quota model, territory, HubSpot hygiene, pipeline review.' It's searchable immediately.

8 Before next quarter's retro, prompt Starch: 'Pull all action items from our last three retrospectives and show me which ones are closed vs still open.' Start the meeting with accountability, not amnesia.

9 If the CRO asks about a past decision mid-quarter, prompt: 'Search our retro archive for anything we decided about enterprise territory overlap.' Starch returns the exact excerpt and which quarter it's from — no tab-hunting.

10 Over time, your Knowledge Management base becomes the institutional memory for how your RevOps team actually works — process decisions, quota rationale, HubSpot hygiene rules — rather than living in one person's head or a stale Confluence page.

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

Q2 2026 RevOps Retrospective — May 30, 2026

Sample numbers from a real run

Deals reviewed in retro (from HubSpot sync)	47
Stale opportunities flagged (no activity 30+ days)	14
Action items extracted by Starch	9
Action items with named owners assigned automatically	8
Past retro decisions searchable in Knowledge Management	3
Time from call end to archived summary	12

On May 30, your team runs the Q2 close retro with 8 people on the call. Before the meeting, Starch pre-loads HubSpot data showing 47 deals touched this quarter, 14 with no rep activity in 30+ days — that's the first agenda item. During the call, Meeting Notes transcribes everything. After, you prompt: 'Summarize this retro — what pipeline decisions did we make, what HubSpot logging rules changed, and what's the new rule on how enterprise opps get reassigned when a rep misses 60-day activity?' Starch returns a 400-word summary in under a minute. It extracts 9 action items — 8 with named owners pulled directly from who spoke in the transcript. One item ('fix the enterprise territory split in Salesforce by June 13') gets P1, rest are P2. The summary is archived in Knowledge Management in 12 minutes total. Two weeks later, when one rep claims the territory rule was different, you search 'enterprise territory' in Knowledge Management and pull the exact paragraph from the May 30 retro.

Measurement

How you'll know it's working

Action item completion rate: what percentage of retro action items are closed before the next retro

HubSpot data hygiene score: stale opportunities (no activity 30+ days) as a share of total open pipeline — tracked retro to retro

Time from retro call end to archived summary in Knowledge Management (target: under 15 minutes)

Retro decisions referenced in subsequent meetings — searchable retrieval events as a proxy for knowledge base usefulness

Open action item backlog by owner — how many items from past retros each RevOps team member or rep still has open

Comparison

What this replaces

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

Google Docs + Otter.ai

Otter transcribes but doesn't extract action items, assign them, or connect to your HubSpot pipeline data — you still manually copy-paste decisions into a Doc that nobody finds next quarter.

Notion + Notion AI

Notion is good for structured wikis but doesn't sync from HubSpot or Apollo, so your retro doc has no live pipeline context — you're still manually pulling deal data before every meeting.

Gong + Confluence

Gong is built for rep coaching on customer calls, not internal retros, and Confluence is where RevOps process docs go to die — no action item tracking, no cross-tool data connection.

Linear + Loom

Linear handles tasks well but you're paying for another tool, there's no automatic extraction from a meeting transcript, and nothing connects back to your HubSpot or Apollo data in the same surface.

On Starch RECOMMENDED

One platform — meeting notes, task manager, knowledge management 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

Does Starch actually pull HubSpot deal data into the meeting context, or do I have to paste it in manually?

Starch syncs your HubSpot data on a schedule — contacts, companies, deals, and owners. When you start a Meeting Notes session and tell Starch to pre-load pipeline context, it queries that synced data directly. You don't paste anything. You can ask for stale opportunities, close-rate by rep, or sequence performance from Apollo in the same breath.

We use Salesforce, not HubSpot. Does this still work?

Yes. Salesforce is reachable from Starch's integration catalog and queried live when your apps need pipeline data. It's not a scheduled sync like HubSpot, so the data is pulled on-demand rather than stored in Starch — but for retro prep and meeting context, that's plenty.

Is the retrospective transcript stored securely? We discuss unreleased quota numbers on these calls.

Starch is not SOC 2 Type II certified today — that's worth knowing if your company has strict compliance requirements. Transcripts and summaries live in your Starch workspace, not a public surface. If SOC 2 Type II is a hard requirement, that's an honest blocker for now.

What if our retro is on Zoom or Google Meet — does Meeting Notes integrate with those?

Zoom and Google Meet are reachable from Starch's integration catalog. You can connect them and query meeting data live. For real-time transcription, you'd run the Starch Meeting Notes session during the call. The workflow works whether your retro is in-person, on Zoom, or async.

We've tried knowledge bases before and nobody updates them. Why would this be different?

Because Starch writes the entry automatically after the retro — you don't ask anyone to 'update the wiki.' The summary is generated, reviewed in 5 minutes, and archived with one prompt. The stale-doc problem happens when updating the doc is a separate task. Here it's a byproduct of running the meeting.

Can we track whether people actually completed their retro action items before the next cycle?

Yes. Action items land in Task Manager with assignees, priority levels, and due dates. Before your next retro, prompt Starch: 'Show me all open action items from our last retrospective.' You get the list by owner. That's your first agenda item — no chasing people on Slack the day before.