Skip to main content

Updated Sep 30, 2026Verified against the product Sep 30, 2026

Open in ClaudeOpen in ChatGPT

Your First Week with FunnelStory

FunnelStory works in layers: your data becomes a context graph of each account, FunnelStory computes intelligence on it — predictions, needle movers, relationship risk — and agents act on that intelligence. This guide is a five-day plan that builds those layers in order, so by the end of the week your team is working from FunnelStory and your first agent is running. It's written for the admin setting up the workspace and the CS leader rolling it out; plan on a few hours a day.

Day 1: Connect your data and load your accounts​

Everything hangs off accounts, so start there.

  1. Connect your system of record — usually your CRM (Salesforce or HubSpot) or your data warehouse. See the Quick Start.
  2. Configure the account model so every customer, with its ARR, renewal date, and owners, appears in Accounts. Follow Configuring the account model, and finish with its verification checklist.
  3. Check a few accounts you know well. If their ARR, renewal date, or CSM is wrong here, everything built on top will be too.

If your CRM is HubSpot, you can configure the models with Claude instead of writing queries.

Day 2: Bring in how customers use the product and what they say​

Intelligence is only as good as the signals underneath it. Add the two kinds FunnelStory needs:

Conversations are what needle movers, the Account Dossier, and relationship maps are built from, so don't skip them. Each model refreshes on a schedule; give them a day to fill in.

Day 3: Turn on the intelligence​

  • Predictions — FunnelStory learns what churn and retention look like from your own history, using Revenue Tags that define a churned and a retained account. Your FunnelStory team sets these up with you; see Predictions.
  • Needle movers — open Needle Movers and skim what's been found. Turn on autosuggested action cards under Configuration → Features → Needle Movers so FunnelStory drafts a next step for each open risk.
  • Relationship risk — check the thresholds under Configuration → Features → Relationship Map against how your team works; see Relationship Risk Configuration.

Some features — the Dossier, Outcome Agents, and Plays among them — are enabled per workspace by FunnelStory. Ask your FunnelStory contact which are on for you.

Day 4: Bring your team in​

  1. Invite users with the right roles, and connect SSO if you use it. See Inviting users and Roles and permissions.
  2. Check assignments. CSMs should see their own book under My Book of Business in Focus Areas. If they don't, fix the owner mapping on the account model.
  3. Set up each CSM's daily tools — each person connects their own:
    • Calendar and Slack, for the pre-meeting briefing in Slack (Meeting prep);
    • Gmail, so customer email lands on accounts and Action Card emails send from their own address (Gmail and Calendar);
    • Claude with the FunnelStory skills, for briefs and decks on demand (Claude Skills for CSMs).
  4. Show them the daily routine — Focus Areas in the morning, the account page before a call. Prepare for a Customer Call is a good first read.

Day 5: Put an agent to work​

Pick one problem your team handles by hand every week, and give it to an agent.

  1. Fill in the worksheet in Design an Agent: Detect, Investigate, Act for that one use case — triaging customer requests or alerting on a champion leaving are good first choices.
  2. Build it as a draft with Claude, test it on real examples, and publish it in Assist mode, so a person reviews what it drafts.
  3. If Outcome Agents are enabled, open Outcome Agents and start a draft for your main product. The readiness check tells you whether there's enough renewal history to generate cohorts; if there is, follow Prevent Churn with a Retention Outcome Agent.

After the first week​

  • Week 2: review the agent's output with the CSMs who received it, and tighten its trigger and instructions.
  • Week 3–4: automate the leadership report you spend the most time on — see Turn Leadership Reporting into Agents.
  • Every month: check that new data sources and products are modeled, and that renewal and churn outcomes are being recorded, so predictions and Outcome Agents keep learning.
Last updated on