A stack diagram shows which tools are present. To understand how pipeline works, I also need to see the decisions and handoffs between them.
Five questions the system should answer
- What signal creates a record or changes its state?
- How does the team decide whether it is relevant?
- Who owns the next action, and by when?
- What happens when the normal path fails?
- Which feedback changes targeting, message, or process next week?
Apollo can send data to HubSpot and HubSpot can post to Slack while the operating gaps remain. An account may enter twice, reach no owner, or receive a reply that sales never sees. Those failures live between the tools.
Start with the state changes
I map the customer and account states before touching automation: identified, researched, qualified, engaged, responded, accepted, opportunity, disqualified, and recycled. For every transition, I define the evidence, the owner, and the next action.
Then connect the tools
Once the operating model is visible, I decide whether each step belongs in HubSpot, Salesforce, Clay, Apollo, n8n, the website, or a manual review. The business logic stays clear even when a tool changes.