What to Actually Track in Your Personal CRM (And What to Ignore)
Most personal CRM articles focus on which tool to use. This one focuses on what to actually put in it — because an empty or poorly-structured CRM is almost worse than no system at all.
A blank database creates the illusion of infrastructure without any of the function. You open it, feel vaguely virtuous, and close it. Nothing changes.
Here’s what to track, what to skip, and the logic behind both.
The One Signal That Matters Most
Before anything else: date of last meaningful contact.
Not last email. Not the LinkedIn comment. Meaningful contact — a real conversation, a call, a coffee, a substantive exchange that moved the relationship forward.
This single data point is more predictive of relationship health than anything else you could track. Relationships don’t decay evenly. They decay through gaps. A relationship you haven’t had a real conversation with in 12 months is at risk, regardless of how many times you’ve liked their posts.
If your personal CRM does nothing else, track this.
What to Capture After Every Interaction
Log the following within 24 hours of a meaningful conversation, while it’s still clear in your head:
Date and channel. When and how — phone, in-person, video, text thread. Channel matters because it signals effort level and relationship register.
What you discussed. Two to three sentences max. Not a transcript. The topics, the context, anything that surprised you or revealed something new about where they are in their life or work.
Their current situation. Job changes, projects they’re working on, challenges they mentioned. This is the data that makes your next outreach feel personal rather than generic.
Action items. Anything you said you’d do, send, or follow up on. Promises not kept are relationship debt.
A reason to reconnect. The best time to record the next conversation hook is right after the last one. “They’re launching something in Q3” is a natural check-in reason in three months.
That’s it. Five fields. If you’re logging more than this per interaction, you’re building a filing system, not a relationship maintenance habit.
What to Ignore
Phone numbers and email addresses. Your contacts app and email client already have these. Don’t duplicate information that’s maintained elsewhere — it just creates sync problems.
Elaborate tag taxonomies. “Investor / advisor / warm / tier-1 / bay-area / met-at-conf-2024” — you will not maintain this. Tags work when they’re sparse and action-triggering. One or two tags per person, max.
Social media follower counts, job titles as primary data. These change constantly and matter less than the interaction history you’ve built.
Every touchpoint. A retweet is not a data point. A one-word reply to a group text is not a data point. Capture meaningful contact, not ambient digital proximity.
The Right Size for Your Network
Most people approach their personal CRM by trying to import every LinkedIn connection. That’s the wrong starting point.
Robin Dunbar’s research on social network limits is consistent: humans can maintain approximately 150 stable relationships, with a meaningful inner circle of 15 to 50 people. Beyond that, the cognitive and time cost of maintenance exceeds what most people can sustain.
Your personal CRM should map to this reality. The 150-person network — not your full contact list of 2,000+ — is the right scope. Within that, segment roughly:
- Inner circle (15-20): People you’re in regular contact with. These relationships mostly take care of themselves.
- Active network (50-75): Relationships that matter but require intentional maintenance. This is where a personal CRM earns its keep.
- Extended network (75-150): Important relationships that need at minimum one meaningful touchpoint per year.
If you’re tracking 500 people, you’re not tracking anyone.
Relationship Direction: The Undertracked Signal
One signal most people never log: whether contact is mutual or one-directional.
A relationship where you’re always the one initiating is a different relationship than one where contact flows both ways. It’s not necessarily worse — some valuable relationships are asymmetric — but the pattern matters.
Track who reached out last. Over time, patterns in this data surface relationships that are coasting on your effort and relationships where the other person is genuinely invested. That’s information worth having.
How Relationship Intelligence Changes the Calculus
The limitation of a manual CRM is that you have to know what to look at. If you’re disciplined about logging interactions, you still have to open the tool and scan for who’s overdue.
Relationship intelligence automates that inference layer. Instead of you reviewing 150 records to find who you haven’t spoken to in 90 days, the tool surfaces it. It weights by relationship strength — a monthly call with someone creates a different baseline than an annual email — and flags deviation from your normal pattern.
The connection recession has made this more important, not less. The ambient contact that used to maintain professional relationships — running into people in hallways, at industry events, in shared offices — is largely gone for remote and distributed workers. Without it, relationship maintenance has to be intentional. Systems that surface who to reach out to and when remove the planning overhead that stops most people from being intentional.
Starting Small Beats the Perfect System
The biggest risk with any personal CRM is over-engineering it before you have a habit.
Start with three fields: contact name, date of last meaningful contact, and one note from that conversation. Build the habit of logging for 30 days. That’s it.
Once logging feels automatic, add the interaction fields. Then consider segmenting your network. Then, if the system still has gaps, add complexity.
A simple system maintained beats a sophisticated system abandoned. Every time.
The personal CRM category tends to sell you on features. What actually matters is the data that changes your behavior — and mostly that’s knowing who you’re losing touch with before the gap becomes a problem.
Start there.