Your First Unframe Customer: Where Should You Start?

After my first two articles, I’ve had a few conversations with partners that quickly moved from why to how.

The logic makes sense: spend less time building AI plumbing, focus more of your effort on transformation, and use faster delivery to improve the economics of your AI business.

But then comes the practical question:

Where do I start with my first customer?

My answer is usually the same:

Don’t start by looking for an AI project. Start by looking for a business problem worth solving.

That sounds obvious. In practice, it changes quite a lot.

Don’t start with “Where can we use AI?”

Walk into an enterprise and ask where they could use AI, and you’ll probably leave with a very long list.

A copilot here. An agent there. A chatbot. Some document processing. A productivity use case somebody saw at a conference.

There’s nothing inherently wrong with those ideas.

But a long list of possible AI use cases isn’t an AI strategy.

And it isn’t necessarily a good starting point for a partner either.

The better conversations I see start with the business.

Where is the organisation losing time or money?

Which processes frustrate customers or employees?

Where are skilled people still doing repetitive work because the systems around them haven’t caught up?

Which business-critical processes have accumulated years of manual steps, exceptions and workarounds?

Now you’re looking for problems rather than technology.

That’s where partners have an advantage.

Start where you already have credibility

If you’re an SI or consultancy, you probably don’t need to go hunting for these problems.

You already know where many of them are.

Your teams are inside customers every day. They understand their systems and processes. They know which transformations are stuck, where the operational bottlenecks are and which problems keep appearing in steering committees.

More importantly, your customer already trusts you enough to discuss them.

That is a significant asset.

So I wouldn’t start an Unframe practice by creating a generic list of 50 AI use cases and sending it to every account team.

I’d start with the customers and industries you understand best.

Ask your account teams:

What is a meaningful problem this customer has wanted to solve, but where the traditional solution has been too slow, too expensive or too complex?

That is a much better place to begin.

The first use case needs to matter

There is a temptation with new technology to start small simply because small feels safer.

I understand the instinct.

But small and focused are not the same thing.

A first use case can have a narrow scope while still solving an important problem.

In fact, I think that’s the sweet spot.

You want something meaningful enough that the business cares about the outcome, but focused enough that you can get it into production quickly and demonstrate value.

Not a science experiment.

Not an AI demo looking for a business case.

And not a multi-year transformation before anyone sees a result.

A real problem. A clear owner. A measurable outcome.

That gives the partner and customer something concrete to build around.

Look for the process behind the problem

This is where the conversation becomes more interesting.

Imagine a customer tells you:

“We want an AI agent for customer service.”

That’s a solution statement.

Before jumping into it, I’d want to understand the process underneath.

Why are customers contacting support?

What information does the support team need?

Where does that information live?

Which decisions can be automated and which require human judgement?

Where are the handoffs?

What happens when something goes wrong?

What would actually constitute a better business outcome: lower cost, faster resolution, better customer experience, higher first-contact resolution, fewer escalations?

Once you understand that, the answer may indeed involve an agent.

But now you’re designing around the business problem rather than trying to find somewhere to deploy a technology.

That’s an important distinction.

Don’t confuse a fast start with a PoC

I’m also deliberately talking about getting something into production, not just building a proof of concept.

Enterprise AI has no shortage of PoCs.

The harder part is turning them into something people actually use, connected to the systems and processes the business actually runs on.

For the first customer engagement, I think this matters enormously.

A successful demo proves that technology can work.

A successful production use case proves that the business model can work.

For the customer.

For the partner.

And for us.

That’s what creates confidence to do the next one.

Design the first engagement with the second one in mind

This is probably the most important part.

The first use case shouldn’t be treated as an isolated project.

While you’re solving it, you’re learning a huge amount about the customer.

Their data.

Their systems.

Their processes.

Their governance requirements.

Their people.

Their appetite for change.

And, inevitably, you’ll uncover other problems.

That’s why I keep saying:

The first use case isn’t the destination. It’s the starting point.

A partner should already be thinking about that expansion path during the first engagement.

If we solve this problem successfully, what becomes possible next?

Which adjacent process could we address?

Can something we’ve built be reused?

Can the partner wrap a managed service around it?

Is there something here that could become repeatable IP for other customers in the same industry?

That’s how one project starts becoming a practice.

The partner’s advantage isn’t the platform

This is also why I don’t think partners need to become experts at pitching Unframe as another piece of technology.

Of course they need to understand what we do.

But their real advantage is everything they already know before Unframe enters the conversation.

The customer.

The industry.

The process.

The systems.

The organisational reality.

Unframe should help turn that knowledge into a production AI solution much faster.

So the motion isn’t:

“We have a new AI platform. Who can we sell it to?”

It’s:

“We know this customer has a problem worth solving. Can Unframe help us solve it in a fundamentally better way?”

That’s the partner motion I want us to build.

So, where should you start?

If you’re considering your first Unframe opportunity, I wouldn’t begin with your technology portfolio.

I’d get your account and industry teams in a room and identify a handful of customers where three things are true:

There is a meaningful business problem.

You understand that problem well enough to challenge how it works today.

The customer trusts you enough to have that conversation.

Then bring us in.

We can work together on whether it’s the right problem, what a production solution could look like, and how quickly we can prove value.

And if the first use case works, we don’t stop there.

We look for the next one.

Because the goal isn’t to sell another AI platform.

It’s to help you build an AI business that is more profitable, faster to deliver and easier to scale.

If you have a customer problem in mind and you’re wondering whether it could be a good first Unframe use case, send me a DM.

Let’s look at it together.