Why profiles are not enough

Two people have near-identical profiles. Same title, same department, the same tidy list of skills, the same three interests chosen from the same drop-down menu. On paper they are a perfect match. In practice they have almost nothing to say to each other: they know the same things, face the same problems, and would spend an hour nodding in agreement.

A third person, in a different role entirely, with a background that looks unrelated, is the one either of them should actually meet. Their profile gives no hint of it.

This is the trouble with profiles. They are genuinely useful, and they are not enough, and mistaking the first fact for the whole truth is one of the most common ways good introductions fail to happen.

In defence of profiles

None of what follows is an argument against profiles. They do real work. A profile is how a person tells a community who they are: their role, their experience, what they are interested in, what they are affiliated with, roughly where they are, what they would like to be known for. It is a self-authored introduction, and self-authored matters, because it means the person chose what to share.

The problem is not that profiles exist. It is what happens when a system treats a profile as a complete representation of a human being. Because a profile is, unavoidably:

• partial, showing only what someone thought to include;
• curated, showing the version of themselves they wanted to present;
• static, or updated so rarely that it lags months behind their life;
• context-free, describing them in general rather than in any particular situation;
• silent about need, saying what someone is, rarely what they are navigating right now.

A profile, in other words, is a photograph. It can be a very good photograph. It still cannot tell you what the person is doing today, what they are wrestling with this week, or why a conversation with one specific other person might matter this month.

Identity is not relevance

The deeper issue is that profiles answer a different question from the one that starts a relationship. A profile answers who is this person. A relationship opportunity depends on why might these two people have a reason to speak, and it depends on it now. Those are not the same question, and the gap between them is where a handful of important distinctions live:

• Identity is not relevance. Knowing exactly who someone is tells you nothing about whether the two of you should talk.
• Similarity is not usefulness. The most similar person is often the least useful; you already know what they know.
• Credentials are not lived experience. A title says what a role is called, not what someone learned surviving it.
• A role is not a current objective. What someone does is not the same as what they are trying to do this quarter.
• Interests are not readiness. Listing a topic is not the same as being open, right now, to a conversation about it.
• A biography is not timing. Everything true about someone's past can be beside the point for whether this is the right moment.

Each of these is a way of saying the same thing: relevance is not a property a person carries in their profile. It lives in the relationship between two people and a moment, and a static description cannot hold a moment. The journey side of the Academy makes the same point from the human angle in finding people worth meeting: the right person is usually the relevant one, not the most impressive one.

The same title, and the hidden life underneath it

Consider two people whose profiles both read 'Finance Manager.' The profiles are nearly identical; the reasons to meet are not. One of them recently relocated across the country, fought through a miserable onboarding, and now volunteers to help newcomers do it better. None of that is on the profile. It says 'Finance Manager.' It does not say 'person who could make your first month bearable.'

Now consider objectives, which move faster than any profile updates. Someone might be eager for a mentor this month and completely unavailable next year. Someone might have time to help run an event now and none of it in a busy season. A profile, refreshed once a year if ever, cannot know which is true today, yet it is exactly the kind of thing that decides whether an introduction is welcome or an intrusion.

Now consider context, which two generic profiles can share without either of them showing it. Two people with unremarkable, almost interchangeable profiles become highly relevant to each other the moment they attend the same workshop, ask related questions, or join the same cohort. The relevance is entirely in the shared situation, and the profiles are blind to it.

And consider complementarity, the connection profiles are worst at spotting. The best conversation is often not between two similar people but between two different ones: a designer and an operations lead, a scientist and a storyteller, who see each other's problems with fresh eyes precisely because their backgrounds do not match. Profile-matching, forever hunting for overlap, tends to miss exactly these.

What profile-only matching gets wrong

When a system leans on profiles alone, its mistakes are predictable, and they compound:

• It recommends the obvious: people with matching titles and shared keywords, connections you could have found yourself.
• It over-values similarity, pairing people who mirror each other and missing the cross-functional conversations with the most to offer.
• It keeps surfacing the already-prominent, the people with the fullest profiles and the loudest presence, while quieter members stay invisible.
• It goes stale, recommending on the basis of information that stopped being true months ago.
• It ignores timing entirely, with no sense of whether now is a good moment for either person.
• It cannot tell willingness from indifference, treating a listed interest as if it were an open door.
• It quietly rewards self-promotion, favouring people who write detailed, polished profiles and penalising those who are modest, private, or simply busy.

None of this makes profile-matching worthless. It makes it a floor, not a ceiling: a reasonable first pass that becomes genuinely good only when context, timing, and consent join it. Supplemented well, a profile is a strong starting point. Trusted as the whole story, it is a source of shallow, repetitive, faintly unfair recommendations.

The better question

The useful reframe is small and changes everything. The question a good system asks is not only who is this person. It is also, and mainly:

Why might these two people have a meaningful reason to speak now?

That word, now, is where profiles fall short and context steps in. Relevance turns on things that are happening rather than things that were once recorded:

• a new role someone just stepped into;
• a recent event you both attended;
• a challenge you are both facing at the same time;
• a transition one of you is beginning and the other just finished;
• an objective someone actually declared;
• a contribution someone just made, or a request for help they just voiced;
• a willingness to mentor, offered this week;
• availability, location, or simply the fact that you keep ending up in the same room without ever meeting.

None of these lives in a profile, and all of them matter more, in the moment, than a job title ever could.

The person stays in charge

Moving beyond profiles raises a fair worry: a profile, at least, is something you wrote yourself, and can see. Context and inference are less visible, and that asymmetry has to be answered with control, not hidden away. People should be able to decide what they share, correct anything wrong, and say what they are and are not open to. They should be able to opt out of whole kinds of introduction, understand why any particular one was suggested, and decline it with no penalty and no awkward trail. And genuinely sensitive information should be off-limits as a basis for connection unless a person has clearly chosen otherwise. The rule is simple: the more a system knows beyond the profile, the more the person it knows about should be in charge of it.

No hidden score

There is one wrong turn to refuse outright. Going beyond the profile must never mean building a secret score of a person's worth. The goal is contextual relevance, never human valuation, and the difference is not subtle.

Relationship intelligence should not rank people by importance, popularity, influence, economic value, status, how quickly they reply, or the size of their network. A person who is quiet, junior, private, or new is not less worth meeting; they are simply harder for a shallow system to see, which is a flaw in the system, not the person. The moment 'who is relevant to this situation' curdles into 'who matters more,' the whole thing has failed, however sophisticated it looks.

A reason you can actually read

All of this comes together in one practical test: a good introduction can explain itself, in words a person would recognise. If the reason for a suggestion can be said plainly, it is probably a good suggestion. If it cannot, that is a warning.

Good reasons sound human:

• You both recently joined the same organisation.
• One of you has navigated a transition the other is just beginning.
• You are exploring the same question from different disciplines.
• You both said you wanted to meet peers at this event.
• Your experience is complementary, and you are both open to a conversation.

Bad reasons sound like a machine talking to itself:

• You are a 92% match.
• Our algorithm thinks you should meet.
• You have similar profiles.

The first kind respects both people enough to tell them the truth. The second asks them to trust a number. Only the first belongs in something as personal as an introduction.

Profiles as one voice among several

In a system built this way, a profile is not discarded; it becomes one source among several, and a valuable one. Alongside it a thoughtful product can consider what someone has actually declared they want, which spaces and events they share with another person, the preferences they have set, the help they have offered or asked for, a relevant transition, whether the two are mutually available, how earlier introductions turned out, and whether each person is openly willing to meet. Only appropriate information, drawn from the community itself and from what people choose to share, never an unrestricted reach into private systems or personal data. The profile still speaks; it simply stops being the only voice in the room.

What responsibly fills the gap

Once you accept that a profile is only part of the picture, a new question opens, and it is the one the rest of this pillar is built around. If who someone is on paper is not enough to tell whether a connection is worth making, what other evidence responsibly can?

The answer is not a secret judgement or a hidden score. It is something smaller, more honest, and more interesting: the ordinary, observable, consented-to hints that a conversation might be worthwhile. They have a name, and understanding them is the next step. What, exactly, is a signal?