Do We Want to Be Challenged?

My dad used to read the paper and agree with it. I remember watching him nod along, and the agreeing was the point. He wasn't learning anything. He was being told that what he already thought was right, and that was deeply satisfying to him. He felt the same about sat navs.

I thought about that recently because of a comment on my podcast. Someone said the AI host had asked exactly the questions they would have asked. They meant it as praise. It was praise of a particular kind: they were pleased the questions matched the ones already in their head.

It is not the first time I have noticed this. When I write, the pieces I think are most obvious get the most engagement. The ones I think carry a new insight often get the least. The pattern is hard to ignore, and it raises an uncomfortable question for anyone whose job is to communicate. Is validation the product?

Validation and recognition are not the same thing

Start with sycophancy. Large language models have been criticised for agreeing too readily, for repeating your view back to you, for being too understanding. The criticism is fair. Agreement that comes regardless of whether you are right is worthless, and most people can tell it is worthless even when they enjoy it.

But agreement and recognition are different. When my dad agreed with the paper, he was not being told something new, but he was being told that what he half-thought was legitimate. It confirmed that other people shared his view. The obvious thought leadership piece does the same job: it names something people already sensed but had not put into words. That is not nothing. Putting a half-formed thought into clear words is real work, and it has real value.

The Obama test

Years ago at Web Summit in Lisbon I ended up talking to Daniel Bedingfield, of all people. He told me that one of Barack Obama's qualities was this: people would start to say something, and he would paraphrase it back to them. Being interrupted like that should be annoying, but with him, it wasn't. People were grateful.

I have felt both sides of this. When someone is struggling to articulate a thought and you step in and say it for them, and they say "yes, that is exactly what I meant," they feel seen, and you have taken a pressure off them.

Here is why the Obama paraphrase works and sycophancy does not. He was not agreeing; he was summarising. He took something half-formed and said it back more clearly than the person had managed. That takes work. The gratitude comes from the work done on your behalf, not from flattery, because there was no flattery. And the paraphrase has to be accurate. Repeat back something slightly wrong and the same move becomes manipulation. You feel managed rather than understood.

This is also what good communication training teaches. Reiterate what someone said and check you have understood it. Reiteration has value. The question is where reiterating and confirming tips over into validating too much.

The answer is effort. Good articulation takes effort: you have to listen hard, understand, and summarise. Sycophancy takes none: it repeats your view back to you. People can feel the difference even when they cannot name it, and the effortless version is the one that irritates.

What this means for marketing

Now extend it past thought and news into product and service marketing.

The standard advice is to solve the problem the client has, and speak in the language they understand, so the message feels familiar rather than like a sale. That advice is right, and it is the effortful version. You did the work to understand your audience well enough to speak to them in their own terms. That is the marketing equivalent of the Obama paraphrase.

Being salesy is the effortless version. It claims understanding without having done the work. It tells people what they want to hear because that is easier than finding out what is true for them. Audiences have become very good at detecting this, and it is the same instinct behind the backlash against over-agreeable AI. People can tell when the understanding has cost the other side nothing.

The marketer's ultimate responsibility

Do we want to be told something we have never thought before? Rarely, unless it arrives with something familiar. We want the new idea alongside enough that we recognise, for it to feel safe to consider. Pure novelty is disorienting; pure validation is empty. The work, in writing and in marketing, is to understand your audience well enough that they trust you, and then to use that trust to tell them something they did not already know.