Dunbar's number is not your personal CRM contact limit

Dunbar's number is often reduced to a catchy rule: a person can maintain 150 relationships. That makes 150 sound like a target, a test or even a maximum number of contacts to keep in a personal CRM.

It is none of those things.

The useful idea is that relationships draw on limited time and attention, and they do so unequally. A handful of people may be woven into your daily life. Others matter in a specific professional, community or historical context. Your contact database can hold all of them without pretending that every name represents the same kind of relationship.

The number 150 is an estimate, not a quota

Robin Dunbar's work linked the size of human social networks to cognitive and time constraints. The resulting estimate of about 150 stable relationships became known as Dunbar's number.

But the exact figure is disputed. A 2021 paper in Biology Letters reanalysed the primate-data method behind the headline number. Different statistical methods produced widely varying estimates and broad confidence intervals. The researchers concluded that this method could not establish one precise cognitive limit for humans.

So, “150” shouldn't tell you how many friends you ought to have. It can't diagnose the health of your network, and it doesn't tell a consultant, founder or job seeker how many contacts belong in their system.

The definition is slippery too. A close friend, a former colleague, a neighbour and someone you met once at a conference can all appear in a contact list. That does not make them equivalent relationships.

The layers are more useful than the limit

Research associated with Dunbar's number describes social networks as a set of roughly nested layers, often summarised as 5, 15, 50 and 150 people. Inner layers tend to contain fewer, emotionally closer relationships. Outer layers contain more people and generally receive less attention.

Those figures are approximations, not four boxes you need to fill. Research using mobile-phone data found evidence of a layered pattern while also finding substantial variation in the middle layers. A 2025 survey of 906 adults likewise found wide differences in how people perceived the energy they gave to inner, middle and outer relationships.

That variation makes sense. Family structure, caregiving, relocation, personality, career stage and community involvement all affect where attention goes. Your network does not need to resemble anyone else's.

A less rigid version of the layers might look like this:

• Closest support: people you turn to when something truly matters.
• Close and active relationships: people involved in your current life, with trust and regular contact.
• Contextual relationships: people connected through work, a group, an event, a shared interest or a chapter of your life.
• Wider and historical network: people you know but do not actively maintain with the same intensity.

These descriptions aren't scores of a person's worth. They are a way to notice that different relationships already have different places in your life.

A database stores context, not closeness

A personal CRM can hold far more than 150 records. Storage is cheap. Attention isn't.

Technology can reduce the effort needed to remember a name, retrieve a note or see how two people are connected. It cannot provide the mutual care, trust or shared experience that makes a relationship close. Saving a note doesn't maintain a friendship. Setting a recurring reminder doesn't create a sincere reason to get in touch.

Online reach doesn't remove that constraint either. Dunbar's 2016 study of online and offline networks found similar ranges and inner-layer sizes despite the communication advantages of social platforms. That does not make online relationships less real. It suggests that easier communication doesn't make human attention unlimited.

It helps to keep four separate ideas apart:

• Closeness: how you currently experience the relationship.
• Recency: when a meaningful interaction last happened.
• Context: where the relationship came from and what connects you.
• Commitment: whether you have a real future action to take.

A trusted friend may need no reminder because contact happens naturally. A former colleague might need only enough context to make a future reconnection accurate. A new collaborator may need a one-off reminder for a promise you made, even though the relationship is not close.

One relationship-health score would flatten all of that.

Review your network without ranking people

Instead of forcing every contact into a numbered tier, open a few records from different parts of your network and ask four questions.

1. What is this relationship now?

Is it close, active, contextual, historical or unclear? Describe the present state without treating it as permanent.

2. What gives it meaning?

Look for shared experience, trust, an introduction, a company, a group, an event or mutual support. The origin often tells you more than a label does.

3. Does it need an action?

There may be a specific promise to keep, a deliberate stay-in-touch rhythm, a genuine reason to reconnect or nothing to do. “Nothing” is a valid answer.

4. What should your system preserve?

Keep the information your future self will need: current context, a dated activity, who introduced you, a reminder or simply the history. You don't need a schedule for every person you want to remember.

Give different relationships different system behaviour

The point of a relationship system is not to industrialise friendship. It is to support the attention you have chosen to give.

This is a practical interpretation, not a scientific mapping of Dunbar's layers.

It also leaves room for relationships to move. People change jobs, leave groups, become collaborators, drift apart or return to your life. Moving outward is not automatically a failure. Moving inward should not erase where the relationship began.

That is one reason weak ties can stay light and still matter. A quiet contact may still hold useful shared history. And when someone no longer belongs on your active map, archiving can preserve the relationship context without pretending it needs current attention.

Apply the idea in Kinnetly

Kinnetly gives you a private relationship graph rather than a flat list. People can stay connected to the companies, groups, events and introducers that explain how you know them. Former memberships can remain part of the history.

You decide whether Know, Like or Trust reflects your current view of a contact. Kinnetly does not infer closeness or assign Dunbar layers. You can record an activity, choose an optional follow-up rhythm, set a one-off reminder for a real commitment or leave the relationship unscheduled. Dormant-contact views help you review quiet relationships, but they do not decide that anything is wrong.

That distinction matters. The software can preserve context and surface the commitments you chose. It cannot choose who deserves your care.

Map the relationships you have, not the number a theory says you should have. Get started with Kinnetly.

Sources

• Lindenfors, Wartel and Lind, “'Dunbar's number' deconstructed”, Biology Letters (2021).
• Dunbar, “Do online social media cut through the constraints that limit the size of offline social networks?”, Royal Society Open Science (2016).
• Li et al., “Reflecting on Dunbar's numbers: Individual differences in energy allocation to personal relationships”, PLOS ONE (2025).
• Tamarit et al., “Cognitive resource allocation determines the organization of personal networks”, PNAS (2018).
• Mac Carron, Kaski and Dunbar, “Calling Dunbar's numbers”, Social Networks (2016).