It is one of the most confidently repeated pieces of advice on the internet, and no study has tested whether it works. [1] It is worth being precise about what is actually known.
The technique itself is simple. [2] The name is the whole instruction: be as interesting to provoke as a grey rock.
It is not the same as ignoring someone, and the difference matters. [3] Silence often reads as a reaction — a loud one. A flat "OK" does not.
Does it work?
The honest answer is that nobody has checked. Support for it is largely anecdotal, and the term traces back to a mental health blog post in 2012 rather than to a clinical literature. [4]
That is not the same as saying it does not work, and it is worth resisting the urge to hear it that way. The reasoning behind it borrows two ideas from the vocabulary of behavioural psychology — reinforcement, where a behaviour that gets what it wants becomes more likely, and extinction, where consistently removing that reward makes it less likely. This page does not cite a source for either, because it has none archived: they are named here as the shape of the argument, not as evidence for it. Applying them to this situation is a reasonable inference. It is still an inference.
This site is named after a similar piece of advice with a similar evidence problem, so the position here is the same one taken on don't feed the trolls: a plausible mechanism and a great deal of testimony is genuinely worth something, and it is not the same thing as a tested intervention. You are entitled to know which one you are being offered.
The part the advice usually leaves out
If grey rocking works the way its own logic says it does, then withdrawing the reward should make things worse before it makes them better. [5]
That has a consequence most versions of this advice never mention. Going flat and then relenting when the pressure increases is not a neutral outcome — it is worse than not starting, because the escalation is what got rewarded. [6] In practice that means grey rocking is not something to try for an afternoon to see how it goes.
It also has a hard limit, and this is the one to take seriously. Where abuse is ongoing or escalating, an expected increase in intensity is not an inconvenience to be waited out — it is a risk. [7] If that describes your situation, the useful move is not a communication technique. It is support and a plan.
What it will not do
It does not touch the reasons the behaviour exists, and the person may move to a different tactic or a different target. [8] A person who cannot get a reaction from you has not been changed by that; they have been redirected. If they have an audience, they may simply use it.
And it is genuinely difficult. [9] Advice that frames it as effortless is setting you up to read your own exhaustion as failure.
Yellow rock
Grey rock has one situation it handles badly: someone other than the difficult person is going to read what you wrote. A co-parenting app, a workplace grievance process, anything that may end up in front of a third party — in those settings a flat one-word reply can be picked up and described as cold, obstructive or uncooperative by someone who benefits from that reading. Yellow rock is grey rock adapted for that audience. It keeps the same substance — nothing to react to, no emotional content, no detail beyond what the message needs — and adds a surface of ordinary courtesy, so the message itself reads as reasonable to whoever sees it later, not just to the person it was sent to.
The distinction worth keeping: grey rock is written for the person replying to you, yellow rock is written for whoever reads the exchange afterwards. Same restraint, different reader.
None of that is backed by anything beyond this site's own description. There is no source for yellow rock specifically, and the archived source behind this page does not use the term at all. The only part of this with a source behind it is the point made at the top: there is no direct scientific research evaluating whether grey rock is effective. [1] The step from there to the variant is ours rather than anyone's finding — a named variation on a method nobody has studied is not better evidenced than the method it varies.
Where it earns its place
Its authors put the useful case plainly: it is for situations where contact cannot easily be avoided and the goal is to limit how much those interactions affect you — not to change the person. [10]
That is a narrower promise than the one usually made for it, and a more useful one. Grey rocking is a way of reducing what an unavoidable interaction takes out of you. It is not a way of winning, or of fixing someone, and the versions that promise either are the ones to distrust.
Online, the same logic applies with one advantage: you can usually leave. The reason to reach for a flat reply rather than no reply at all is that a visible non-answer, in front of an audience, often ends an exchange that silence would have prolonged.