AI Won't Replace Client Services, But It Should Change How You Work

AI is still making a lot of agency teams feel slightly stuck.

Not because they don't see the potential. but because the practical application still feels messy.

• Where do you use it?
• What's safe to put into it?
• How do you make sure the quality is good?
• And how do you stop everyone in the team using it in completely different ways?

That's exactly why I invited Jules Love from Spark AI to run a Masterclass for the Client Services Collective on making AI work in Client Services.

And one of the biggest takeaways for me was this:

AI isn't really about replacing the human side of client services. It's about creating more space for it.

Because client services teams are already stretched. Lower budgets, higher expectations, more admin, more channels, more context switching.

And when you're constantly reacting, it becomes very hard to step back and think strategically about the relationship, the commercial opportunity, or what the client actually needs next.

Jules talked about using AI across the client lifecycle, from onboarding and brief analysis through to QBRs, status reporting, account planning and spotting growth opportunities.

But the important point was that AI only becomes truly useful when you give it enough context.

That means capturing the things that often sit in people's heads:

• Client background
• Stakeholder dynamics
• Meeting notes
• Project context
• Previous proposals
• What good looks like inside your agency

Otherwise, you just get generic outputs and generic outputs don't help anyone.

A practical example: briefs

One practical example Jules shared was around briefs.

Instead of simply pasting a vague client email into ChatGPT and asking for help, he showed how agencies can build a simple AI assistant that analyses the brief in a consistent way.

It could pull out:

• What the client is really asking for
• What's missing or unclear
• The five most important questions to ask
• The assumptions that need testing
• What the team should be ready to discuss on the call

That's where AI starts to become useful.

Not as a shortcut for thinking but as a tool that helps your team think better, more consistently and with more confidence.

The data problem

Another point that really stood out was around data.

If your case studies, client notes, project outcomes and performance insights are hidden in messy folders, old decks or people's memories, AI can't do much with them.

BUT the action isn't to go back and fix ten years of archived work.

That will feel overwhelming and probably stop you starting.

The action is to start from today.

Capture current project learnings in a format AI can actually read. Record debriefs. Save transcripts. Create simple forms at the end of projects. Build the habit into your workflow now.

So here's my challenge for you this week

Pick one repeated client services task that takes up too much time.

It might be preparing for a QBR, analysing a brief, writing follow-ups, creating an account plan or pulling together case studies.

Then ask:

• Could we build a more consistent way of doing this with AI?

Not to remove the thinking.

But to give your team more time for the thinking that actually matters.

Because AI won't replace strong client services.

But client services teams who learn how to use AI well will have a serious advantage.

Jo x