When not to make an introduction

Introductions are a credit line

Every guide to introductions is about how to make them. Almost none are about when to refuse, which is odd, because refusing is the harder skill and the one that determines whether the rest of it works.

Judy Robinett puts the reason plainly in How to Be a Power Connector. Every recommendation she gives reflects on her, so she is careful about who she refers and which introductions she makes. That is not fussiness. It is an accurate description of the mechanics.

When you introduce two people, the person receiving the introduction is not evaluating the stranger. They have no way to. They are extending trust to you and accepting the stranger on that basis. So the introduction draws on a balance you hold with them, and a bad one debits it.

The balance is finite and you cannot see it. Which means the ability to say no is not a failure of generosity. It is the thing that keeps your yes worth anything.

Five times the answer is no

You cannot vouch for the work. Liking someone is not the same as being able to speak for how they operate. Robinett treats competence as an entry requirement rather than a bonus, and the test is simple: could you describe something specific this person did well, without being reminded? If not, you are not introducing them, you are forwarding them.

The ask is too vague to act on. "I'd love to meet interesting people in fintech" cannot be fulfilled. It gives you nobody to picture and nothing to write, and any introduction you make from it will be arbitrary. This is not really a refusal, though. The right reply is a question: who specifically, and what would you want from the conversation? Most vague asks become good ones under one round of questioning, and the ones that do not were never serious.

The receiving side has no reason to want it. Run the introduction from their seat. What do they get? If the honest answer is nothing, and you are relying on their goodwill toward you to carry it, you are spending exactly the balance described above. Do it rarely and knowingly, not as a default.

Someone else is the better introducer. An introduction carries the sender's credibility, so the right sender is whoever can say the most specific true thing. If you met the person twice and a colleague worked with them for two years, your introduction is weaker than theirs even if you are more senior. Saying so is more useful than proceeding.

The asker has a history of not following through. Robinett lists this as one of the common failures: people say they want something, an introduction is offered, and then nothing happens. That is not a small discourtesy. You spent your standing to open a door and the person walked away from it, and the cost lands on you rather than them. Once is circumstance. Twice is information, and it is reasonable to act on it.

How to decline

Fast and plain. The wording matters much less than the speed, because the real damage is not the no, it is the three weeks of silence while someone waits for an answer you were never going to give.

"I don't know them well enough to introduce you properly" is true, complete and unarguable, and it does not require you to critique anybody. If you can offer the next best thing, offer it: a different person, a warmer route, or the observation that they should approach directly and can say you suggested it.

What to avoid is the soft yes. "Let me see what I can do" and "I'll keep an eye out" feel kinder in the moment and are worse in every other way, because they leave the request open forever and make you someone whose answers have to be interpreted.

If you have said no to a particular pairing before, keep a note of it. Nothing erodes trust faster than being asked to reconsider something you already declined, except forgetting you declined it and agreeing the second time.

The test that tells you how you are doing

Robinett offers a good measure of a network's real health, and it is not size. It is responsiveness. Do people reply to you quickly? Do they act on your recommendations without needing to be convinced?

That number moves in response to how carefully you introduce. It falls quietly, because nobody tells you they have started taking your emails less seriously. They just reply a little slower, and eventually a request goes unanswered with no explanation available.

Which is the argument for the whole piece. Scott Gerber and Ryan Paugh make the same case in Superconnector from the other direction: the shift worth making is from volume to curation. A person who makes twenty introductions a year that all land is worth more to everyone, themselves included, than one who makes two hundred that mostly do not.

Doing this at volume

An advisor with several companies can have twenty introductions in flight, and at that point the failures stop being judgement calls and start being bookkeeping. You forget you declined this pairing in March. You lose track of which asks are still open. You introduce two people who already met last year, through you.

That last one matters more than it sounds, because a duplicate introduction tells the recipient you are not paying attention, which is precisely the thing your introductions are trading on.

NetworkOS keeps that record. It knows which pairs have already been introduced and which were dismissed, so it does not resurface them, and it tracks how much outreach each contact has had so the same few names are not asked for favours repeatedly. What it will not do is make the judgement in this article for you. It cannot tell whether you can vouch for someone, and it should not try.

The mechanics of a good introduction, once you have decided to make one, are in the guide to making a warm introduction. The other side of this one, how to ask so that people can say yes, is in how to ask for an introduction.

The short version

Say no when you cannot vouch for the work, when the ask is too vague to picture a person, when the receiving side gets nothing, when someone else can say something more specific than you can, or when the asker has already failed to follow through. Decline quickly and plainly, and never with a soft yes that leaves the request open. Every introduction spends a little of the credibility that makes the next one land, and the people who are trusted with the big introductions are the ones who declined the small bad ones.