How to keep track of networking contacts: memory, spreadsheet, or CRM?
Memory works to 20 contacts, a spreadsheet to about 50, a personal CRM beyond. What to track at each stage — and what you should never write down.
Keeping track of networking contacts is a sizing problem: memory handles about 20 active relationships, a spreadsheet stretches to 50, and past that a personal CRM stops being a luxury. Just as important as the system is the restraint — track context and commitments, never surveillance.
The three stages, honestly compared
Almost everyone moves through the same progression, and each stage is genuinely right for a while. The failures come from staying one stage too long — usually because the current system fails quietly instead of loudly.
Memory + calendar
Free, frictionless, and legitimate below ~20 active contacts. Human memory is excellent at the relationships it touches weekly. It fails at exactly two things: long gaps (the quarterly contact evaporates) and promises (what you owe whom). When those two start slipping, the stage is over — no amount of trying harder extends it.
Spreadsheet
The right tool from ~20 to ~50 contacts. Costs nothing, bends to any workflow, and the act of building it forces useful decisions about who matters. Its weakness is structural: it cannot initiate. The system only works on the days you open it, and the first genuinely busy month usually ends those days. Two abandoned spreadsheets is a verdict, not bad luck.
Personal CRM
Earns its place past ~50 contacts, or at the second spreadsheet lapse. The difference is push versus pull: it resurfaces the right person on the right day instead of waiting to be checked. Costs money or setup time, and importing garbage produces organized garbage — it amplifies a habit, it doesn’t create one.
The honest caveat on stage three: a personal CRM amplifies an existing habit rather than installing one. If you’ve never logged a contact in your life, start with the spreadsheet — the networking tracker template is a ready-made CSV with the right columns — and graduate when it’s alive but straining. The full tool-choice question, including costs over three years, is covered in when you actually need a personal CRM.
What to track: context and commitments
The point of tracking is that your next message can be specific. Specific is what warmth looks like in text. That defines exactly what’s worth recording:
Where and how you met. “TechBBQ 2025, the panel queue” anchors every future interaction. Without it, contacts decay into names you’re vaguely sure you know.
What they care about right now. One line: the project, the search, the problem. This is the difference between “how’s things?” and “did the Series A close?” — the second gets answered.
Last contact and next touch. The two dates that make a follow-up cadence real instead of aspirational. A tracker without a next-touch date is a guest book.
Open loops. What you promised, what they promised. Kept small promises compound into trust faster than any other input; forgotten ones quietly end relationships. This column alone justifies the whole system.
What they chose to share. The marathon, the new baby, the sick parent, the move. Asking about these later is the single most relationship-warming move available — and the one memory reliably fumbles past twenty people.
Log it the same day. The half-life of conversational detail is brutal: by next Friday, the panel queue and the Series A have merged into fog.
One optional extra that punches above its weight: a give tag — one word for what you could plausibly offer this person. “Intro to Lena”, “feedback on pricing”, “my conference notes”. When you review weekly, the give tags turn “I should reach out to someone” into “I owe Priya those notes” — a concrete, generous action instead of a vague intention. It’s the column that keeps the whole system pointed at giving rather than extracting.
What NOT to track
Restraint is half the system, and it’s the half nobody writes about. Three categories stay out.
Data they didn’t give you. Salary guesses, relationship gossip, the results of an enthusiastic background search. If the person would be surprised you hold it, you shouldn’t hold it — surprise data poisons real trust the day it leaks into a conversation, and it always eventually does.
Sensitive categories by default. Health, religion, politics, sexuality. The exception is things explicitly shared in the relationship’s own context — a friend’s recovery you’re supporting them through belongs in your notes because remembering it is caring. The same fact harvested secondhand does not.
Judgments. “Arrogant — avoid.” “Wasn’t useful.” Verdict-notes feel efficient and act like poison: they pre-load every future interaction with a conclusion the person can’t appeal. Record behavior if you must (“didn’t follow through on the intro, twice”) and let future-you re-judge.
There’s a legal footnote here, and it’s reassuring: for genuinely private use, the GDPR’s household exemption puts your personal relationship notes out of scope. But the etiquette rule is stricter than the legal one, and the etiquette rule is the one to build on — store what was shared with you, the way you’d be fine showing.
A worked example: one contact, six months
Abstract advice about fields gets clearer with a single row followed through time. Here’s what tracking actually looks like in practice.
March, day of meeting. You meet Priya at a product meetup; she’s a design lead wrestling with a research-repository migration. That evening, 40 seconds of logging: Priya N. — ProductTank München, March. Design lead @ fintech. Pain: research repo migration. Owes: nothing. I owe: the article on repo taxonomy I mentioned. Next touch: this week. Tier: warm.
Same week. You send the article with one line of context. She replies warmly. Log: Sent taxonomy piece, she’s trying it with her team. Next touch: June.
June. The next-touch date surfaces her. Because the context is right there, the message writes itself: “Did the taxonomy approach survive contact with your team?” She answers with a war story and mentions she’s hiring a researcher. You know someone. One double-opt-in intro later, both sides owe you a small, warm debt. Log it; next touch September.
September. The reminder fires. The researcher got the job. Your note this time is pure congratulations — no ask, no agenda. Total investment across six months: maybe eight minutes of logging and three short messages.
Now multiply the counterfactual: without the system, the article never gets sent (forgotten by Friday), June never happens (no reminder), and Priya is a face you vaguely recall at next year’s meetup. The eight minutes weren’t bookkeeping. They were the relationship.
Getting started in one evening
-
Pick the stage that matches today
Count active contacts — people you’d genuinely reach out to this year — not address-book rows. Under 20: stay on memory, but add calendar reminders for the quarterly people. 20–50: spreadsheet. Past 50, or past your second abandoned spreadsheet: personal CRM.
-
Seed it with the living, not the archive
Add the 15–25 people who matter now, plus everyone from this month’s meetings. Skip the 400-contact import; a museum of dead rows is the fastest way to make the system feel like homework.
-
Fill only the seven fields
Name, where met, what they care about, last contact, next touch, open loops, tier. Anything more is maintenance debt. An empty notes field is fine; a stale one is misinformation.
-
Book the weekly 15 minutes
A recurring slot — Friday afternoon works — to scan upcoming touches, close loops, and log the week’s new people. This appointment, not the tool, is the system. Miss it twice and any tracker starts lying to you.
The endgame is unremarkable in the best way: you stop performing memory feats and start having them performed for you. We built Endearist around exactly this shape — the seven fields, the next-touch resurfacing, notes that stay in a local file you own — but the shape matters more than the brand. Pick the stage that fits, write down only what makes the relationship warmer, and let the system carry what your head was never going to.
FAQ
What's the best way to keep track of networking contacts?
The one that matches your network's size. Below roughly **20 active contacts**, memory plus calendar reminders genuinely works. From 20 to about 50, a **spreadsheet** with columns for context, last contact, and next touch is the sweet spot. Beyond 50 — or whenever the spreadsheet stops being opened — a **personal CRM** earns its place, because it resurfaces people instead of waiting to be checked. Most people's mistake is not bad tooling; it's running a 80-contact network on a 20-contact system.
When is a spreadsheet enough?
When three things are true: your active network fits in about **50 rows**, you actually enjoy (or at least tolerate) ten minutes of weekly upkeep, and you remember to open the file. A spreadsheet is free, flexible, and yours. Its weakness is structural: it can't tap you on the shoulder. The day you stop opening it — usually a busy month, two at most — the system silently stops existing. If that's happened to you twice already, the third spreadsheet won't fix it.
What columns should a networking tracker have?
Seven cover nearly everything: **name**, **where we met**, **what they care about** (one line), **last contact date**, **next touch date**, **what I owe / they owe** (open loops), and **tier** (inner / warm / outer). Resist the urge to add more — every additional column raises the per-contact maintenance cost, and unmaintained columns rot into misinformation. Our [networking tracker template](/en/templates/networking-tracker) ships exactly this shape as a ready CSV.
What information should I record about a contact?
Context that future-you will need: where you met, what they're working on, what you discussed, what you promised each other, and the **personal details they chose to share** — the marathon, the new baby, the move to Lisbon. These are what turn your next message from generic to genuinely warm. The test for every note: would the relationship be better served by remembering this? If yes, write it down the same day, because by Friday it's gone.
What should I NOT track about people?
Three categories. **Scraped or guessed data** — information they didn't give you and wouldn't expect you to hold (salary guesses, relationship gossip, anything from a background search). **Sensitive categories** — health, religion, politics, sexuality — unless they shared it and it's genuinely relevant to caring for the relationship. And **judgments** — 'arrogant, avoid' style notes that poison every future read. The glass-database rule covers all three: write every note as if the person might one day read it.
Is it legal to keep notes on people under GDPR?
For genuinely personal use, yes. The GDPR contains a **household exemption** (Art. 2(2)(c)): processing 'by a natural person in the course of a purely personal or household activity' is out of scope — your private address book and relationship notes qualify. The picture changes when the data serves commercial activity at scale, like a sales pipeline. A freelancer's personal network notes live in a gray zone best handled by the same rule that's good etiquette anyway: store only what was shared with you, and store it like you'd be fine showing it.
Isn't it creepy to keep notes on people?
Writing 'partner: Sam, kids: 2, training for Berlin marathon' isn't surveillance — it's what a good memory does, externalized. The creepiness test is about **content and intent**, not the act of writing: notes that help you show up warmer (remembering the job interview, asking about the dog) serve the other person. Notes that catalogue weaknesses or leverage serve you against them. Keep the first kind. People consistently feel honored, not monitored, when you remember what they told you.
When should I switch from a spreadsheet to a personal CRM?
At the moment of the second lapse. One abandoned spreadsheet is circumstance; two is a verdict on the model. The structural difference is **push versus pull**: a spreadsheet waits for you, a personal CRM resurfaces the right person on the right day. If your network is past ~50 active contacts, if open loops keep slipping, or if your 'last contact' column is mostly three months stale, the migration costs an hour and pays it back within the month. The deeper comparison is in our piece on [when you actually need a personal CRM](/en/blog/personal-crm-vs-contacts).
Can I just use LinkedIn to keep track of my network?
LinkedIn knows who you're connected to; it doesn't know **the state of the relationship**. No field holds what you discussed, what you promised, or when you last genuinely talked — and the connection list is sorted by LinkedIn's interests, not yours. It's also rented infrastructure: profiles vanish, accounts go dormant, the algorithm decides what resurfaces. Use LinkedIn as the discovery and update layer it's good at, and keep the relationship memory — notes, cadence, open loops — in a system you own.
How much time does keeping track actually take?
Less than the anxiety it replaces. The honest budget: **30–60 seconds per new contact** to log context the same day, and a **weekly 15-minute review** to scan upcoming touches and close open loops. That's roughly an hour a month for a 50-person network. The hidden saving is on the other side: no more reconstructing who someone was before replying, no more 'I know I promised them something' dread. Tracking is cheaper than remembering.
I have hundreds of unsorted contacts. Where do I start?
Don't import everything — that builds a museum, not a system. Start from zero and **add people as they become relevant**: today's meetings, this week's follow-ups, the 15–20 people you already know matter. A useful prompt: scroll your sent messages from the last three months; anyone you wrote to twice belongs in the system. The 400-contact backlog can stay where it is. A tracker with 30 living entries beats one with 400 dead ones.