Concepts & methods

Personal CRM: keeping up with people, without the ick

Borrowing customer-relationship tooling for friendships sounds cold until the alternative is forgetting. A reference entry on what the category solves, the instrumentation line it must not cross, and what friendships actually need.

Aleem O'BalogunBy Aleem O'Balogun|Updated |4 min read

The short answer

A personal CRM applies customer-relationship discipline to your own people: one place holding who matters, the dates that matter to them, what you last talked about, and when you last spoke. It solves real failures (the missed birthday, the friendship six months adrift). The line it must not cross is treating people as pipeline: the tooling should serve attention, never replace it.

What it is, and where it comes from

The personal CRM imports a business pattern (the customer-relationship manager: contacts, history, follow-ups) into private life. The category has no single inventor; its reference open-source project is Monica (monicahq.com), and a steady stream of apps has worked the same ground. The underlying problem is better credited: Robin Dunbar's research programme on friendship maintenance finds that relationships decay measurably without contact, that emotional closeness falls within months of reduced interaction, and that people sustain layered circles of relationship (a handful of intimates, wider bands of friends and acquaintances) with sharply different maintenance costs. His famous 150 is best held as a heuristic rather than physics (later reanalysis puts enormous confidence intervals around it), but the layering and the decay are the practical points: closeness is upkeep, and upkeep is exactly what busy decades quietly stop paying.

What it actually solves

Three real failures, none of them character flaws. The missed date: birthdays and anniversaries arrive unnoticed until the day, when the thoughtful version of caring needed two weeks of lead time. The silent drift: nothing went wrong, and it has somehow been eight months, because no moment ever announced itself as the moment to call. And the mental load of being the one who remembers: in many households one person carries every date, every follow-up, and every “we should have them over” in working memory, indefinitely. Externalising this is not cold; it is the same move as writing anything down, applied to the part of life where forgetting costs the most.

Where it turns cold

The ick is real and worth locating precisely. It arrives when the grammar of pipeline crosses over: contact frequency as SLA, friends as rows with overdue flags, affection optimised like a funnel. Logging every conversation turns presence into data entry; scoring relationships invites managing the score. The honest test is directional: notes that help you show up better (the surgery date to ask about, the name of the new puppy) serve the friendship; metrics that grade the friendship serve the system. A personal CRM should feel like a good memory, not a sales dashboard, and the moment updating it competes with actually calling someone, the tool has failed at its one job.

How it maps to the notice loop

Relationships need the loop's notice verb more than they need records: the birthday surfacing with lead time, the friend you keep missing surfacing before the drift sets, the visit that needs booking surfacing while trains are cheap. Capture holds what matters about people; noticing turns dates and gaps into this week's prompts; the weekly shape gives one relationship thing a real slot, which is how dates stop being ambushes. The record is the means. The plan is the point. A perfectly maintained contact database with no protected evening in the calendar is the relationship version of a beautiful backlog: complete, current, and unlived.

Sources

The maintenance research as published, with the 150 figure held explicitly as heuristic; the category's reference project named as such.

  • Robin Dunbar's friendship-maintenance research programme, including Roberts, S. G. B. and Dunbar, R. I. M. (2011), Communication in social networks, on decay without contact; the layered-circles model. On the 150: Lindenfors, P. et al. (2021), Biology Letters, whose reanalysis puts wide confidence intervals around any single number.
  • Monica (monicahq.com), the category's open-source reference implementation.
← All guides