"Would you brief a junior copywriter like that?"

Yesterday I dropped this line in my newsletter: "The robots aren't coming for our jobs. They're here to do our work. But only if we learn how to be better managers than our machines are assistants."

The response was immediate. A subscriber messaged me: "This hit SO hard. I've been trying to get ChatGPT to write product descriptions for months and they're still rubbish. What am I doing wrong?"

Here's the thing…. she isn't doing anything wrong with AI. She's just not taking on the role of ‘AI manager’. The same with most of us.

The Lightbulb Moment

First of all, yes AI has a far bigger role than writing product descriptions. 100%. However, for many of us this is the starting point. How we either build trust or dismiss AI.

Three months ago, I was working with a furniture ecommerce team who'd hired me because their AI experiments were, and I quote, "failing spectacularly." They'd spent weeks trying to get AI to write their category descriptions, and everything came out generic and soulless.

Sound familiar?

I sat in on one of their AI conversations. The marketing manager typed: "Write a description for our dining tables category."

That was it. No context. No brand voice. No customer insights. Just a vague instruction and then frustration when the output was bland.

I asked him: "Would you brief a junior copywriter like that?"

… ‘we’ve never worked with a copywriter, we do these ourselves’.

Okay. We have a starting point.

We're Managing Humans Wrong Too

Here's what I've learned from watching dozens of ecommerce teams struggle with AI: Most of us are actually terrible at managing people as well. Harsh but fair. We just don't notice because humans are incredibly good at filling in the gaps.

When you ask a team member to "write some product descriptions," they automatically:

• Consider your brand voice from previous work

• Think about your target customers

• Recall conversations about positioning

• Apply common sense about what works

Problem is that they’ll hate the task and likely, have better things to do.

AI doesn't do any of that. It's like managing the most literal, context-free intern you've ever met. And it won’t sulk.

The Three Manager Types I See

After six months of AI coaching with ecommerce teams, I've spotted three distinct manager types:

The Micromanager: Tries to control every word. "Write a 47-word product description using these exact keywords in this precise order..." They’re sooooo brand. Usually ends up taking longer than writing it themselves.

The Vague Delegator: Throws out loose instructions and expects magic. "Make our email campaigns better." Then wonders why the results are disappointing.

The Effective Briefer: Treats AI like a talented but junior team member who needs proper context and clear objectives.

Guess which one gets the best results?

What Good AI Management Actually Looks Like

Last month, I worked with a cycling gear brand who'd cracked this. Their secret? The ecommerce manager managed his AI exactly like he managed his best freelance copywriter.

Instead of: "Write product descriptions for our road bikes"

He'd brief: "You're writing for competitive cyclists aged 25-45 who know their kit. They care about performance data and technical specs, but they also want to feel like they're investing in something that'll give them an edge. Our brand voice is knowledgeable but not pretentious. Think knowledgeable bike shop owner, not engineering manual. Here's three examples of descriptions that worked well for us..."

The difference in output quality? Night and day.

The Skills Transfer Problem

The irony is that most ecommerce marketers already have the skills to be great AI managers. You just need to apply them consistently.

Think about how you'd brief a new team member for your most important campaign. You'd probably:

• Give them background on your customers

• Share examples of what good looks like

• Explain the business objectives

• Provide brand guidelines

• Set clear success metrics

Do the same with AI. Every time.

Where Most People Go Wrong

I've seen the same mistakes repeatedly:

Assuming AI Knows Your Business: Just because it can write doesn't mean it understands your customer's pain points or your competitive positioning.

Skipping the Examples: You'd never ask a new hire to write in your brand voice without showing them examples. Same rule applies.

Forgetting to Iterate: Good managers give feedback and refine. Most people try AI once, get mediocre results, and give up. “Bloody robots… bet we can get somebody better on Fiverr.” Smirks.

Not Setting Clear Objectives: "Write better copy" isn't a brief. "Write copy that reduces our FAQ tickets about shipping times" is.

The Uncomfortable Truth

Here's what I've realised after months of this work: AI isn't exposing our technology skills gap. It's exposing our management skills gap.

The ecommerce teams getting the best AI results aren't the most technical. They're the ones who were already good at briefing, managing and iterating with their human teams. Same for founders at startups who are wearing all the hats. You teach AI. You manage AI. You explain how they can do a better job… to AI.

If you're struggling with AI, ask yourself: Are you giving it the same quality of direction you'd give your best team member? Or are you treating it like a magic box that should just "know" what you want?

What This Means for You (And if You Have One… Your Team)

The skills that'll matter most in the next few years aren't about prompt engineering or understanding transformer models. They're about:

• Clear communication

• Effective briefing

• Iterative feedback

• Understanding your customers deeply enough to explain them to others

• Knowing what good looks like in your brand voice

Sound familiar? These are the same skills that make great marketing managers.

The robots aren't coming for our jobs. But they are about to reveal who among us actually knows how to manage effectively.

And honestly? That might be the most valuable feedback we've ever received.