What this actually is
A CRM nobody trusts is worse than no CRM, because the team keeps a private spreadsheet and the reporting quietly describes a company that does not exist. This module rebuilds the parts that decide whether a lead becomes a customer, and it does it in a way that survives the next reorganisation.
What gets built
- Lifecycle stages that mean something. Each stage has an entry condition anyone can check, an owner, and an exit. If two people can disagree about which stage a record is in, the stage is not defined yet.
- Properties with a purpose. Every field either drives a workflow, appears in a report, or gets deleted. Most CRMs carry hundreds of fields nobody has read in two years.
- Routing and ownership. By language, territory and segment expertise, so people receive work they can actually do. With a rule for what happens when the owner is away, which is where most routing quietly breaks.
- A handover both sides trust. An explicit definition of what marketing hands over, what sales accepts, and what happens to the ones sales rejects. Rejection with a reason, feeding back into scoring.
- Reactivation. The deals that went quiet are the cheapest pipeline in the building. A sequence that reopens them on a trigger rather than on a calendar reminder.
- Reporting that survives a reorg. Built on stages and properties rather than on individual people, so it still works when the team changes.
The test for a lifecycle stage. Show the definition to one person in sales and one in marketing, separately, with ten real records. If they sort them the same way, the stage is defined. If not, you have a label rather than a stage, and every report built on it is fiction.
Why HubSpot is our default
HubSpot
We are marketing people before we are CRM people, and HubSpot is the one platform where the top of the funnel, the automation in the middle and the contact record at the end are the same system. One place decides what a lead is, what happened to it, and who owns it now.
- No seam between marketing and sales tooling. No nightly sync quietly deciding which record is true.
- Lead source survives the whole journey. The channel that produced a customer is still attached to the closed deal.
- Automation and reporting read the same fields. The workflow that routes a lead and the report that counts it cannot disagree.
- You already own it. Everything we build sits in your portal, under your permissions, and stays there.
When it is not HubSpot
We build on whatever CRM you run. On larger accounts that usually means Salesforce, and we connect the marketing engine to it the same way. The engine does not change, only where the records live.
Where the AI part actually is
HubSpot exposes its CRM to Claude directly, so you can ask your own pipeline a question in plain language and get an answer out of your own records rather than a guess. Salesforce and Anthropic launched Claudeforce in August 2026 to do the same thing on that side. We build for both, and the useful part is the same either way: the CRM has to be clean first, or the model is confidently wrong about a database nobody trusts.
How it runs
| When | What happens |
|---|---|
| Week 3 | Audit of what exists, which fields are actually used, and where records currently get stuck. |
| Week 4 | Lifecycle stages and properties agreed and built. First workflow live. |
| Weeks 5 to 6 | Routing, ownership, handover and rejection paths. |
| Weeks 7 to 8 | Reactivation, plus the migration of historical records onto the new stages. |
When this is not the right module
If you have no pipeline yet because you have no leads, build lead sources first. A perfect lifecycle model with nothing flowing through it is an expensive diagram.