Glossary
What is relationship intelligence?
Relationship intelligence is the practice of working out who your organisation actually knows, and how well, by reading the communication that already happened rather than asking people to record it.
It is derived, not entered
The distinguishing feature is where the data comes from. A traditional CRM knows what people typed into it. A relationship intelligence system reads the email and calendar traffic the firm generates anyway and derives the relationship from that.
This matters because the entered version is always incomplete and always flatters the person entering it. Nobody logs the meeting that went badly, and nobody logs the four-year correspondence with a contact who never became a deal, which is exactly the relationship worth knowing about.
What a strength score is made of
Recency: how long since the last real exchange. Frequency: how often over the relationship life. Reciprocity: whether they reply, and how fast, which is the component that separates a relationship from a mailing list. Depth: how many people on each side are involved, because a single-thread relationship dies when either person leaves.
A score built from those four is a rough number and should be presented as one. Its job is to sort a list, not to be quoted at anyone.
What it is used for
Warm introductions: before approaching a company, find out whether anyone in the firm already knows someone there, and how well. In a partnership or an agency this is the highest-yielding question that gets asked least, because asking it usually means emailing everyone.
Decay detection: relationships that were strong and have gone quiet. This is the list nobody maintains and where the next project usually comes from.
Continuity: when someone leaves, knowing which relationships they held so those can be transferred rather than discovered later by an unhappy client.
How relationship intelligence works in a CRM
- Automatic capture as the input. Email and calendar sync at the organisation level. A system fed by manual logging cannot compute any of this.
- A visible reason under the score. Last exchange, reply rate, who else is involved. A bare number nobody can interrogate gets ignored.
- Cross-firm search. "Who here knows anyone at this company" answered from the data, not by an all-hands email.
- Privacy boundaries that are set on purpose. This reads real correspondence. Who can see whose relationships is a decision to make deliberately, before switching it on.
How we do it in Lumenbase
Relationship strength is the signal Lumenbase leads with. It is computed from captured email, meetings, and LinkedIn activity rather than from logged notes, and the Feed uses it to rank who needs attention, with the reason shown next to the name.