heard of it.

We know it.

Listener, Dunning, Kruger, two researchers, one called Dunning, one called Kruger.

And they identified that essentially people who are really, really smart about a topic assume they're not as smart as they actually are.

People who are quite dense about a topic, not as smart, incompetent, I imagine is the word they used.

Also, I don't think they're very good.

It's the people in the middle who sort of have this overconfidence.

But yeah, where does that fit in?

guess what it proves is that you've been through the years and you've lasted, right? So there's that obvious idea that you're resilient.

Yeah.

Well, as with all of these psychological principles, there is some real common sense behind it, right? If you are a company, if you're selling a product, if you're a store that has been around for 266 years, like in the case of Guinness, It is natural to assume that that company, that store, that product must be of extremely high quality to have survived all of that time.

Now, that is a heuristic.

That is a shortcut.

It's not necessarily true.

There are lots of things that have survived for an awful long time, not necessarily because they are good.

So everything is,

Gemini.

Gemini.

Yeah, exactly.

So nothing changed with like...

Uh, uh, uh, ChatGPT launched, uh, November 2022.

Nothing changed immediately, but, uh, in the years following, what they measured specifically was variance in language style.

And before the AI had launched, everybody spoke in this beautiful variety.

Yep.

Right.

And if we have said it publicly, and maybe we've written it down, we'll be even more likely to do it.

And there's this amazing study in the book, which I think Steve Martin, who's a co-author of Cialdini, ran himself, and he went to NHS general practices in the UK, so medical practices in the UK.

And NHS in particular have a real problem with people not showing up to their appointments, and this causes a backlog, increases the amount of people who need to sign up for treatments.

And they tried everything.

They tried fining people.

They tried incentivizing people to come.

Okay.

And so the data is going to pollute the result and therefore use your intuition and trust that more than anything.

Yes, one of the things that's so fascinating about marketing is that often the opposite of what the textbook tells you can gain better results than what the textbook suggested in the first place.

So I've just been writing about this amazing story in the book Influenced by Robert Cialdini.

And in the book, he goes undercover at high end restaurants to try and figure out what the best waiters do at restaurants.

And he discovers the best restaurant is a waiter called Vincent.

And Vincent's got this amazing tactic, which if the manager happened to hear Vincent, he would almost certainly be fired for doing it.

But Vincent's tactic is fascinating because what he does is he finds a table of definitely more than a couple of people, sort of decent group.

like how does credibility translate to social proof

So I guess my question is like, what? Beyond personalization, should software companies be thinking about when it comes to signaling this extra effort, delight, however you want to kind of frame it?

Yeah, I mean, the reason you feel that way about the personalization, Sylvie, is because the personalization you are seeing is scaled, automated, and effortless.

Us three, we'll all see this all the time.

I guarantee us three get the same email every single day.

which is an email from somebody at an agency that promises people that they can get them on every podcast in the world.

And that agency has a very specific prompt that it gives Claude.

And the prompt is, go and listen to the most recent episode of Nudge or the most recent episode of Talking Too Loud.

What does this make you think about? How does credibility translate to social proof?

Yeah.

Yeah.Really boring.

[chuckles] On purpose because they wanted to measure something.

So after the presentation was done, they told some groups of people that a presentation took eight minutes to prepare, so hard-- or no, sorry, 30, 34 minutes to prepare it was, so hardly any time.

But they told another group of people that same presentation took eight and a half hours to prepare.

And what they find is that people's ranking of the presentation had nothing to do with the content of the presentation, regardless of whether it was optical switches or electronic ink.

Their ranking was solely determined by what they perceived was the time spent on preparing the presentation.

I've got to know.

Do you spot them a lot?

The classic one, the really funny one is, because we've spoken about social proof already, there's this concept of negative social proof.

So social proof is very powerful.

We do follow the actions of other people.

But sometimes marketers get confused, and they share that lots of people are doing an action which the marketer would consider undesirable, and they put that in their campaign, and that actually influences the wrong behavior.

So let me give an example.

This was a, a study in universities.

Yeah.

Yeah, go

on.

So imagine you're there in March and the weather's absolutely stunning.

There's loads of people there.

You're walking down the timber.

Now it's the sun setting.

You're trying to decide what restaurant to go to.

Do you want to explain, tell us a little bit about what you did?