We get asked this almost every week, usually by a business owner who has just watched an AI tool write a blog post in 30 seconds. If it can do that, why pay a marketing team?
Here’s an honest answer. AI agents can replace a lot of marketing tasks. They can’t replace marketing thinking. And businesses that run AI agents for marketing without anyone senior steering them end up producing more of the wrong thing, faster.
What can AI agents actually do in marketing?
AI agents are very good at first drafts, repurposing, research and consistency, which is a large share of any marketing team’s week. It’s also where businesses are leaning on AI most. McKinsey’s global survey found marketing and sales is the function where generative AI is most used, with 55% of technology companies using it regularly in marketing and sales.
We run our own delivery on agents every day, with a person reviewing every piece. The jobs they do well:
- Turning a 15-minute interview with an engineer into a first-draft article.
- Repurposing one blog into LinkedIn posts, an email and a sales snippet.
- Drafting case studies from project notes.
- Summarising vendor announcements and competitor activity.
- Drafting service page FAQs and Marketplace listing copy.
That’s real time back. For a business with one marketer doing ten jobs, it’s the difference between keeping up and falling behind.
What can’t AI agents do?
AI agents can’t decide who you should target, what you should stand for, or which of the hundred things you could do will actually move revenue.
They also can’t:
- Choose your niche or your positioning.
- Know which vendor relationship matters most this quarter, or which client can’t be named.
- Judge whether a claim is true and safe to publish. They will invent a statistic with total confidence.
- Ask “why are we doing this at all?”, which is often the most valuable question in marketing.
- Build relationships with partner managers, clients and event organisers.
AI makes your marketing faster. It doesn’t make it right.
Why DIY AI marketing usually goes wrong
The DIY version usually looks like this. An owner or office manager signs up to an AI tool, asks it for a month of LinkedIn posts, and publishes them. The posts are fine. They’re also indistinguishable from every other IT business doing exactly the same thing, and your buyers use these tools too. Forrester found 94% of business buyers now use AI in their buying process, so they know generic AI writing when they see it.
It goes wrong in three ways. The voice is generic, because nobody trained the tool on yours. The message drifts, because there’s no strategy behind it, so you publish more of the wrong thing. And claims go out unchecked. That happens more than most people admit: in the University of Melbourne and KPMG global study on trust in AI, 66% of employees said they rely on AI output without evaluating its accuracy, and 56% said they’d made mistakes in their work because of AI. With technical buyers who check, that costs you credibility.
More content isn’t the goal. More of the right content is.
What does a marketing team built on AI agents look like?
The most effective setup has senior thinking at the start, AI agents doing the heavy lifting in the middle, and a person checking everything at the end.
| Job | Who does it |
|---|---|
| Strategy, positioning and priorities | A senior marketer or fractional CMO |
| The ideas and expertise | Your engineers and leaders, often in a short conversation |
| First drafts of articles, posts and case studies | AI agents |
| Editing, fact-checking and approval | Your marketer |
| Vendor and co-branding compliance | Agents set up with the rules, checked by a person |
| Relationships, events and partner managers | People |
| Reviewing results and deciding what changes | A senior marketer |
Notice where the people sit: at the start and at the end. That’s where judgement matters most, and it’s exactly the part the DIY version skips. It isn’t only small businesses, either. McKinsey’s 2025 State of AI survey found only 27% of organisations using generative AI review everything it produces before it’s used.
So should you replace your marketer with AI?
No. Give your marketer well-built AI agents and a clear strategy, and they’ll produce far more of what actually matters.
If you don’t have a marketer yet, the answer is similar. Rather than hiring a junior and hoping AI fills the gaps, pair senior strategy with agents. We compare the options in detail in our guide to choosing between an in-house marketer, a fractional CMO or an agency.
Frequently asked questions
What is an AI marketing agent?
An AI marketing agent is an AI tool set up for one marketing job, such as case studies or LinkedIn posts, with your instructions, voice, examples and rules built in. It turns a general tool into a specialist.
Can AI write our blog posts?
It can write the first draft. The ideas, the opinion and the final check still need to come from your business, or the post will read like everyone else’s.
Will AI agents make our marketing sound generic?
Only if they’re set up generically. Agents built on your own content and voice, with a person editing every draft, sound like your business.
Is it safe to use AI agents with client information?
Agents should be set up so client and personal information isn’t needed and isn’t entered. The OAIC’s guidance on commercially available AI products is a good reference, and an AI use policy for marketing is the place to start.
Put AI to work properly
If your team is experimenting with AI but it isn’t adding up to better marketing, our AI Marketing service builds agents around your voice and trains your team to use them well. Book a call and we’ll look at where AI would save you the most time.
