How to run a retrospective or post-mortem as Small Customer Success Teams
Your team runs a post-mortem after every messy churn or rough onboarding — and then the notes sit in a Google Doc nobody rereads. The actual lessons (why account X churned three months after a rocky kickoff, why the same onboarding question keeps hitting Intercom) never make it into a process change. You're pulling context from HubSpot deal notes, Intercom ticket threads, Gmail threads, and one person's memory. There's no structured format, no action item tracking, no way to search what you decided last quarter. Three people covering 250 accounts can't afford to repeat the same churn patterns because the retrospective process is a shared doc that goes stale.
What you'll set up
Apps, data, and prompts
The combination of Starch apps, the data sources they pull from, and the prompts you use to drive them.
Starch syncs your HubSpot data on a schedule (contacts, companies, deals, owners) and syncs your Gmail on a schedule (message threads, labels) so the post-mortem app can pull account history without manual copy-paste. Connect Intercom from Starch's integration catalog — the agent queries it live when building the retrospective context. Google Calendar is synced on a schedule to pull meeting history for the account. Meeting Notes captures the retrospective call itself; Knowledge Management stores and indexes all completed retrospectives for future search.
Step-by-step
See this running on Starch
Connect your tools, describe what you want, and the agent builds it. Closed beta is free.
Acme Corp churn retrospective — March 2026
| Acme Corp (SMB, $18,000 ARR) | 18,000 |
| Churned month 7 of 12-month contract | 0 |
| 3 Intercom tickets flagged 'confused about reporting' in months 3-4 | 0 |
| Last QBR held 5 months prior | 0 |
| Action item from prior post-mortem ('schedule mid-year check-in for accounts > $15k') — never assigned | 0 |
You trigger the post-mortem app for Acme Corp the day after they churn. Starch pulls the HubSpot deal timeline (7 months, 2 expansion conversations that went cold), the last 14 Gmail threads (last reply was 6 weeks ago), 3 Intercom tickets all tagged 'reporting confusion' in months 3 and 4, and 4 Calendar meetings — the last one 5 months back. The post-mortem app populates the template: the root cause section flags that there was no structured follow-up after the onboarding confusion tickets were resolved, and the 'signals we missed' section highlights the 6-week email silence before renewal. Meeting Notes from your retrospective call adds the team's live discussion: your colleague mentions they saw the same 'reporting confusion' pattern in two other accounts that month but never flagged it formally. That observation becomes a task: audit all Intercom tickets tagged 'reporting confusion' in the past 90 days, P1, assigned to the team lead, due in 5 days. Knowledge Management indexes the whole retrospective. Three months later, when a similar account in the same segment hits month 4, you search 'reporting confusion onboarding' and find the Acme post-mortem immediately — you schedule a proactive check-in call before the silence starts.
How you'll know it's working
What this replaces
The other ways teams handle this today, and how the Starch version compares.
One platform — knowledge management, meeting notes, task manager all running on connected data. Setup in plain English; numbers stay current via scheduled syncs and live agent queries.
Try it on Starch →Frequently asked questions
Does Starch actually read the content of our Gmail threads with customers, or just metadata?
We use Intercom for customer support tickets. Can Starch pull that in?
Can the post-mortem app work for onboarding stumbles, not just churns?
Is the Knowledge Management app good enough to replace our shared Notion wiki?
What if someone forgets to open Meeting Notes before the retrospective call starts?
We're not SOC 2 certified — does that matter for a customer success team?
Related guides for Small Customer Success Teams
A strategic account plan is a documented, living view of a specific customer or prospect — their business goals, the stakeholders who matter, the gaps your product fills, the risks to the relationship, and the actions your team is taking.
Read guide →A customer knowledge base is the document — or collection of documents — that answers the questions your customers ask repeatedly.
Read guide →Lifecycle email flows are the automated message sequences that go out when someone signs up, goes quiet, upgrades, churns, or hits any other meaningful moment in their relationship with your product or service.
Read guide →A product roadmap is how you turn a backlog of ideas, customer requests, and strategic bets into a prioritized sequence of work your team can actually execute against.
Read guide →Run a Retrospective or Post-Mortem for other operators
The AI stack built for the founder's office.
Read guide →The AI stack built for small RevOps teams.
Read guide →The AI stack built for small marketing teams.
Read guide →The AI stack built for small IT and ITOps teams.
Read guide →Ready to run run a retrospective or post-mortem on Starch?
Request closed-beta access. Everything is free during beta.