AI can polish a bad pitch but it can’t make a journalist give a s**t

PR suddenly got much faster

A lot of the day-to-day mechanics of PR have become faster over the past couple of years.

Research that might once have taken an hour can now take a few minutes. You can summarise a reporter’s last twenty articles, pull out recurring themes, draft ten versions of a pitch, turn an interview transcript into comments, rewrite a subject line, build a briefing document and come up with three alternative angles before you’ve finished your first cup of coffee.

That is genuinely useful.

AI has made some of the more repetitive parts of communications dramatically easier, and pretending otherwise would be disingenuous. There are plenty of things I do now with AI that I would have spent far longer doing a few years ago, and there are probably plenty more that we’ll all be doing differently again in another twelve months.

The slightly awkward part is that everybody else has access to the same tools.

More polished does not mean more interesting

If PR teams can use AI to research and write faster, journalists are inevitably going to receive more of everything, more pitches, more contributed articles, more expert commentary and more emails that have clearly been personalised enough to look like somebody spent twenty minutes researching the recipient, even if in reality a model did most of the work in twenty seconds.

The result is not necessarily more terrible PR. In some ways, it is more dangerous than that… it is more competent-looking PR.

An email can now be cleanly written, reasonably personalised, grammatically perfect and completely irrelevant at the same time. A piece of contributed content, known in the business as ‘thought leadership’ can contain no original thought and have not been close to any leadership. A CEO comment can have polish and apparent sophistication while saying almost exactly what five other CEOs have already said that morning.

AI can remove a lot of the obvious signs of a bad pitch. No matter what though, it can’t make an uninteresting story interesting, and that distinction matters a lot more than you think.

Journalists don’t have an attention surplus

The media industry hasn’t suddenly created thousands of additional reporters because PR teams became more productive. Quite the opposite.

Newsrooms remain stretched, journalists are covering broader beats, and many are receiving an enormous volume of inbound every day. Giving communications teams the ability to produce more material doesn’t somehow create more space on the other side to receive it, review it and judge it. It just increases the competition for that space.

This is why I’m not convinced by the idea that AI somehow makes earned media easier. It makes parts of the process easier, absolutely. It can make a good team faster and a bad team more prolific.

But convincing somebody that your story deserves their attention is still an entirely different problem.

A journalist still has to look at what you are offering and decide that it matters to their audience, that the timing makes sense and that there is enough substance behind the idea to justify their time on it.

No model can make that decision for them.

When everyone can write, judgement becomes the differentiator

For years, a lot of communications work was partly constrained by production.
Could you research quickly enough? Could you draft the piece? Could you turn around the commentary in an hour? Could you prepare the CEO before the journalist came back to you?

Those things still matter, but they’re becoming less scarce. Judgement is not.

Knowing that a perfectly reasonable story is still not strong enough for a particular journalist is judgement. Knowing that the interesting part of an announcement isn’t the announcement itself is judgement. Knowing that a CEO should probably say nothing rather than add another generic opinion to an already crowded debate is judgement.

So is understanding that the strongest line in a conversation is sometimes the one the client barely noticed, or that the proof point buried on page 17 of a deck is actually more interesting than the headline the company has been using for six months.

AI can help you get to those decisions faster. It can challenge them, give you alternatives and help pressure-test the thinking.

But the decision still matters more than the output.

The volume trap

There is an obvious temptation here.

If producing ten pitches is easier than producing two, why not send ten? If we can create five contributed article ideas in an afternoon, why not pursue all five? If every executive can have a constant stream of social posts, commentary and thought leadership, why leave any empty space?

Because more communication does not automatically create more reputation. Sometimes it just creates more noise.

There’s a point where speed becomes useful, and another where it becomes an excuse not to make choices. The ability to create more material can make it harder to ask the uncomfortable question, is this actually worth anybody’s attention?

That question has become more important, not less.

AI should make good PR better

I don’t think the answer is to become nostalgic about how much longer everything used to take.

I’m quite happy not spending half my day manually going through search results or turning interview notes into something readable. There’s no particular virtue in doing repetitive work slowly.

The opportunity is to use that saved time on the things that remain difficult. Understanding the market better, finding the argument that is actually different, spending more time talking to clients, and ultimately knowing what journalists are interested in before asking them for something.

AI should give good communications teams more time for judgement. If it only gives them the ability to send more emails, we’ve officially missed the point.