AI Is About to Change What Clients Hire Fractionals For

I run Gev Design, a fractional product and creative studio. Writing about the business of independent work and what I think is coming next.

I was having brunch with a friend the other day. As a parent, that happens about once in a million years if I’m lucky.

We were talking about AI and what it means for the work we do.

Here’s my bet:

Within a few years, the best fractional work will start with a question a company cannot afford to get wrong.

Should we build this?

Fix this?

Hire for this?

Sell this?

Or leave it alone?

Good operators already help clients make calls like these.

What will change is how far one experienced person can take the question before the client needs a bigger team or bigger budget.

Talk to customers. Find the real problem. Make something rough. Test it. Recommend what deserves more time and money.

I always like staying close to the work to see if the things we make actually changed anything.

That’s getting much easier.

So, what changes

Trying an idea before spending money on it isn’t new.

My prediction is that AI will let one experienced person get much further before a company needs to fund a whole project.

They can talk to customers, work out where the problem is, use AI to make a rough version of a possible fix, then watch people try it.

We’ll still need designers, devs, researchers, financial folks, ops, and other specialists.

But companies will have a clearer idea of what they need those people to do before bringing them in.

A company will pay someone to answer one important question before committing to a whole project.

That is what changes.

Not one person doing every job.

One person getting close enough to the problem to make the next decision easier.

The research points the same way. The International Labour Organization, a UN agency that studies work and employment, expects AI to change more jobs than it removes. People are still needed for much of the work.

The companies getting the most from AI are not just adding it to old processes. They’re changing how the work gets done.

What it looks like

Say a company says:

“We want to build an AI agent that chases customers for missing information.”

The first question isn’t:

“Can we build it?”

It’s:

“Will customers hate it?”

Give the agent one small job.

Ask for one missing document. Flag anything unclear to a person.

Then find out:

• Do customers reply?

• Does the team save time?

• Does it create more problems than it solves?

The work ends with a reco:

Build it. Change it. Or leave it human.

That is a better starting point than a six-month “AI agent project.”

The job gets clearer

The old way is often to divide the work immediately.

A researcher talks to people.

A designer makes screens.

A developer builds the thing.

A marketer tries to explain it.

A fractional leader tries to pull it together.

Each person may do good work. But every handoff means someone has to explain what they learned to the next person.

And when the first version proves the idea is wrong, the work travels backwards.

I think more companies will begin by asking one person or a small studio to get to the first real answer.

Then they bring in specialists with a clearer target.

A developer joins once there is something worth building.

A researcher joins when the question needs more than a handful of customer conversations.

A CFO joins when the idea needs a serious financial case.

The point isn’t to give one person five jobs.

It’s to involve the right people at the right time.

Before you scope it

Try this with the next request a client gives you.

1. Find the question behind the request.
“Redesign onboarding” becomes “What is stopping people from getting started?”

2. Separate facts from guesses.
What do customers actually do? What are you assuming?

3. Pick one thing to find out.
What would change the recommendation?

4. Make the smallest useful test.
Use AI to explore options or make something people can react to. Then get evidence from real people and real behaviour.

5. Agree on the next decision.
What would make you go ahead, change direction, bring in specialists, or stop?

Then offer a first step:

For [time and budget], I’ll investigate [question], test [assumption], and recommend what to do next, including what we still do not know.

That can be a small standalone project.

It can also be the beginning of a longer fractional relationship.

The stuff we can’t skip

A nice prototype isn’t proof.

A good demo isn’t proof.

A spreadsheet, a dashboard, a campaign, or an AI agent isn’t proof either.

You still need to understand the business. Talk to people dealing with the problem. Be honest about what the test tells you.

Some questions will take longer.

Some will need specialists from the start.

Some companies will do more of this internally.

More AI capability will not automatically mean more work for fractionals.

But I think it creates an opportunity to be easier to hire.

Not for a long list of services.

For one important decision.

The next brief might not say:

“Here’s what we need you to make.”

It might say:

“Here’s what we need to figure out before we waste six months.”

I’m also building Juggle for people running independent businesses. It turns client calls into notes and tasks, and helps you keep track of client relationships.

If you’re juggling several clients, conversations, and moving parts, I’d love you to try it.

Sign up for the Juggle beta →

Thanks for reading,
Gev