By Kim Leary
The short version
- Leaders are under real pressure to be seen leading on AI. Prompting it yourself feels like leading.
- The most useful example we’ve heard came from an MD who used AI to prepare for the expert, not to replace them.
- When the chatbot gets the final say, the cost lands on the work and on the team.
- Arguing about whether AI is good or bad goes nowhere. The problem is how it’s being used.
- Look at what the AI version gets wrong. It tells you which problem you have.
- Give the AI your strategy and your brand to work from, and give leadership a way to use it that helps instead of overrides.
You write the article. It goes to the MD. It comes back rewritten, with a note: this is what ChatGPT said. The proposal gets a website mock-up made in an afternoon, in the client’s branding, with half the details invented. The team away day becomes an AI session, and by lunchtime nobody wants to share their ideas because a chatbot has already “designed” it.
This is coming up in almost every conversation we have with marketers right now.
Why it happens
Leaders are under real pressure to be seen leading on AI. Prompting it yourself feels like leading.
And the output looks good. It’s fluent, it’s fast and it’s confident. What it isn’t is informed. It doesn’t know your buyers, your positioning or why you chose the words you chose. And when the person asking owns the business, the questions they ask carry their own assumptions, so the answers do too.
What it looks like when a leader uses AI well
The most useful example we’ve heard came from an MD who used AI to prepare for the expert, not to replace them.
He put a contract through ChatGPT and asked how much risk each condition carried. Then he went back to his lawyer, in his words, “in a strategic mindset”. The AI didn’t make the decision. It made him better prepared for the conversation with the person who knew the detail.
Marketing works the same way. A leader who uses AI to test a plan, question the numbers or prepare for a review with the team is helping. A leader who uses it to produce the finished piece and overrule the people who know the buyer isn’t.
What it’s costing you
When the chatbot gets the final say, the cost lands on the work and on the team.
What doesn’t fix it
Arguing about whether AI is good or bad goes nowhere. The problem is how it’s being used.
“Just ban it.” It won’t happen, and it shouldn’t. Used properly, it’s useful.
“Prove the AI version is worse.” You’ll win some and lose some, and it’ll still feel like a fight.
“Let it all go through AI.” Then nothing sounds like you, and nothing stands out.
“Another AI training day.” A day of demos without your brand or your strategy behind it is how the resentment started.
Where to start: the strategy or the setup
Look at what the AI version gets wrong. It tells you which problem you have.
It’s probably the strategy if:
- The AI version and yours are written for different buyers.
- Leadership and marketing would describe what you do in different ways.
- The arguments are about what to say, not how it’s said.
It’s probably the setup if:
- You agree on the message, but the output still comes out generic.
- Everyone prompts from scratch, in their own way, and what lands on your desk is nearly right and slightly different every time.
- The visuals look like everyone else’s.
If it’s mostly the first list, start with the brand strategy workshop. If it’s the second, AI tools and training is the right first step. If it’s both, the strategy comes first, because the tools need something to work from.
What fixes it
Give the AI your strategy and your brand to work from, and give leadership a way to use it that helps instead of overrides.
Start with what the AI should be working from. If nobody has agreed who you’re for and what you say to them, our brand strategy workshop puts the MD and the team in one room to agree it, so the work and the prompts have something to answer to. From £2,800 + VAT.
Then AI tools and training sets the tools up properly: prompts, templates and guardrails built from your brand, and half a day of training on your own work, with leadership in the room. From £2,800 + VAT. If the visuals keep coming out generic, brand assets and design systems gives the tools your real assets to work with.
Keep reading
- “I don’t like it”: why the loudest opinion keeps rewriting your marketing
- Why AI-generated graphics all look the same
- Content without an opinion is just more of the same
Questions about leadership using AI in marketing
Agree the strategy and messaging together first, then set the AI tools up to work from it. Feedback can then be checked against what was agreed, not against what a chatbot produced.
Because without your strategy, brand and audience to work from, AI gives you the most likely answer, which is the same answer it gives everyone else.
As a tool that works from the agreed strategy, alongside the team, not as a replacement for the thinking.
Yes, for the right jobs: testing a plan, questioning the numbers, preparing for a conversation with the team. It goes wrong when the output replaces the brief and overrules the people who know the buyer.
Give it the real facts to work from, and make sure someone who knows the client checks anything before it goes out. AI tools and training builds the prompts, templates and rules from your brand, so the starting point is what’s true about you.
That’s often where this starts: whoever missed the beginning of the process reaches for ChatGPT instead. Walk them through what was agreed before they prompt against it, and think about their role before the next session. There’s more on that in when the CEO shouldn’t be in the room.
Free resource
Why your positioning still isn’t working. AI can only work from a message you’ve agreed. Four quick tests to check yours is.
The one thing to do next
AI tools and training
Prompts, templates and guardrails built from your brand, and half a day of training on your own work, so AI helps the team instead of overruling it. From £2,800 + VAT.
