From a one-(wo)man show to a full-blown marketing team (in just six months)

Within six months, my job shifted from a solo operation to collaborating with a full marketing team.

Where I used to do everything myself—from strategy to content and campaigns—I now work daily with five colleagues, each with their own specialty.

Lisette works on our SEO strategy. Sophie dives into everything related to GEO. Kim writes content for TeamValue and Bizure. Charlotte guards the strategic direction. And soon, Nova will join us as our SEA and campaign specialist.

A complete team, then. Focused, agile, and with everyone bringing their own expertise to the table.

And no, I didn’t get any extra headcount (currently: 0.9 FTE). So… how?

Exactly. With AI.

Because all these colleagues are virtual. Created with a prompt, fueled by our strategy, and sensitive to nuance. They know TeamValue and Bizure inside and out, understand our tone of voice, have a feel for the market, and know what works.

Why am I telling you this?

I’m a marketer at TeamValue Group. And in this digital-first organization, innovative ideas are constantly flying around.

Developing new propositions, conducting market research, sharpening personas, setting up campaigns, and creating content that actually works? Yup. I do it all. By myself.

And to keep up with all the experts at TeamValue and Bizure, I basically need to double myself. Or better yet: quadruple myself.

I need people to spar with. People who are fast, can think critically, speak the same language, and provide the right input.

That’s why I work with AI.

Collaborating with AI makes me more productive. Much more productive. Sometimes even three times as fast.

But, and this is important, it also requires critical thinking. AI isn’t an all-knowing oracle. You have to keep checking, keep thinking. Because even AI can get it wrong.

And just as importantly: keep talking to your human colleagues. Hear what customers are saying. Share insights, ask questions, and listen closely. Because that’s where the real value is created.

From random prompts to a smart team

I know what you’re thinking: aren’t these just prompts? I’ve been using those forever.

True. But the difference is in how you use them.

Instead of random prompts with random names, I work with virtual colleagues. Carefully chosen roles with a clear division of labor.

Because when you think of your prompts as actual colleagues, you automatically sharpen your game.

What’s their expertise? What are they responsible for? What should they definitely stay away from?

That’s how you build structure. You give your ‘team’ boundaries. You know exactly who to go to for SEO, who helps with strategy, and who sets up your campaigns like a pro.

Missing someone in your virtual team? Then it’s time to hire a new colleague. Not just another prompt, but one with a clear role and responsibilities.

Building AI colleagues

I’ve learned what works and what doesn’t. That’s why I’m sharing how I build my AI colleagues step-by-step, so you don’t have to reinvent the wheel.

It starts with a solid briefing. These are the steps I use:

1. Assign a clear role

Start with the basics: who is this colleague? For example:

You are an SEO marketer at TeamValue. You are responsible for brainstorming, writing, and optimizing SEO content. You work according to our content strategy, help with keyword research, and keep our tone of voice, internal link structure, and metadata in mind.

2. Feed your prompt the right info

Does your AI colleague need a strategy to work from? Give it to them. Think about your SEO approach, tone of voice, preferred writing style, content examples, and things you absolutely don’t want to see. The more concrete you are, the better the result.

3. Set ground rules for structure

Do you want specific elements in your content, like headers, internal links, metadata, or a CTA? Be clear about it. It saves you from editing later and makes your output much more consistent.

4. Teach your prompt to ask questions

Add a instruction that your ‘colleague’ should always ask questions if something isn’t clear. This encourages the AI to think critically instead of just churning out content blindly.

5. Let AI learn from you

Happy with content you’ve created before? Have your prompt analyze it. Ask the AI to save the key style elements and apply them to future output. That way, every new SEO blog gets better at matching your specific style and standards.

6. Provide context for every task

Even if your prompt knows the strategy, tone of voice, and structure, it’s smart to briefly reiterate the context with every task. Think along the lines of: This blog is part of a campaign about X or This post is intended for target audience Y. This prevents the AI from going on autopilot and keeps your content relevant.

7. Start with an outline or test version

Have your prompt create an outline before you ask for the final product. This allows you to make adjustments early on and prevents the AI from heading in the wrong direction.

8. Use a feedback loop

Give the AI explicit feedback after every output: what went well, what could be better? Also, ask for a one-line summary of that feedback. This builds a mini-logbook that gets sharper the more you work together.

AI on your own turf

Be careful about what you share with your virtual colleague. It’s best not to drop personal data, revenue figures, or internal information into an AI chat. Many AI tools (temporarily) store your input to improve their models. And if you’re working according to ISO 27001—the standard for information security—that’s simply not allowed.

That’s why I’m excited that we’ll soon have our own AI environment at TeamValue. Built in Azure, completely locked down and secure. Including all the smarts of ChatGPT. Everything you put in stays within our own environment. Nothing is stored or shared outside of our control.

Soon, my virtual colleagues will be working in their own secure AI environment with the same smart support, but completely under our control.

A great step forward. And the best part: I get to keep building on this proposition so your organization can work with it safely, too. Naturally, with my virtual colleagues by my side.

You’re hiring!

Whether you're just starting out with AI or have been prompting for a while, the real difference lies in how you collaborate with it. Don't think of it as a magic trick; think of it as a new team member that needs proper onboarding. Provide direction, give feedback, and build something that actually moves the needle in your work.

Ready to hire your first AI colleague? Let’s go!

Want to learn more about building your own AI team members? Feel free to shoot me an email at [email protected]