Earlier this month Ryanair decided that being horrid to customers was not a great business strategy and declared it would be a bit nicer. This was pretty remarkable and has, indeed, been much remarked upon.

Yet even more remarkable was what caused chief executive Michael O’Leary to make this U-turn. It was not market research. It was not social networks. It certainly was not anything to do with management consultants, whom O’Leary once said he would shoot if they ever darkened his doorstep. Nor was it due to pressure from the board.

Instead, the trigger was people who periodically accost him in McDonald’s to moan about his airline while he sits trying to enjoy a meal with his kids. As he said to shareholders at last week’s annual meeting, he is sick and tired of it.

So never mind big data. When it comes to bringing about change, it is criticism delivered in person by random strangers that counts. O’Leary is unusual in many ways, but in this one I suspect he is just like the rest of us.

On the face of it, placing so much emphasis on such meetings is irrational. The people who bearded him were surely no more disaffected than the thousands who for years have been posting their hostilities online. In the brief time since I started writing this article, several dozen angry tweets have been written, including this one, which I rather like: "That £50-100 difference between Ryanair and BA is taken not from your wallet, but from your soul."

Such stuff is both endless and up there for the world to see, and yet turns out to be easier for executives to ignore than half a dozen cross customers they meet in person.

You might have expected that, as the virtual world has grown and data proliferated, the value of the real encounters would have shrunk, but the reverse seems to be happening. The more bewildering the virtual world, the more we fall back on “real” evidence, no matter how subjective, presented by strangers under our own noses.

Chance meetings
It is not just O'Leary who puts disproportionate weight on chance meetings while he is eating his supper. Richard Dawkins recently told the [London] Times that he realised atheists such as him had won the battle over God because, at the dinner parties he goes to, he no longer meets anyone religious. When even scientists trust the anecdotal evidence of the dinner party more than data, you know something pretty fundamental has happened.

Just now I bumped into a fellow columnist. I was about to explain to him my theory about the inflated trust we place in face-to-face encounters but before I could open my mouth, he started boasting. He told me he had just got back from the US where he had been stopped twice by perfect strangers, once in a bookshop and once in an airport, both of whom told him they loved his writing. There was no need for me to ask his view of my theory: he was providing me with living proof of it.

This man gets a vast amount of adulation on email and Twitter, but by comparison to the real thing, they do not touch him at all.

Yet if praise delivered in person by strangers is powerful, criticism delivered in the same way is even more so. I can remember exactly the dinner party I went to about 15 years ago when a fellow guest who I had never met before looked me in the eye and said that she thought my columns were fatuous. I can remember the food, what I was wearing – everything.

This sort of thing is memorable partly because it is so unusual. Most people do not relish being nasty in person: we have all been brought up to be polite to strangers, especially if we are breaking bread with them.

Internet trolls
By contrast, on the internet our upbringing is non-existent. No one seems to think there is anything wrong with being gratuitously horrible – so long as we cannot be seen. So the dinner party/McDonald's test may not be unscientific after all. The person who approaches an off-duty chief executive to complain is not just another internet troll. They are someone who really means it and really wants an answer.

The executive is then put on the spot, in a way that almost never happens in the course of a normal day in the office. At a dinner party or at McDonald’s there is no PR person at hand to draft a sanitised reply. There are no underlings to delegate the tricky question to. There is no time to think it through: a convincing answer needs to be given in public then and there.

This means that the test set by the angry fellow diner is invaluable. If the chief executive cannot muster a good defence, it suggests that O’Leary’s response is the only honourable one. A U-turn is signalled. – (Copyright The Financial Times Limited 2013)