AI won’t replace relationships. But it should change how agencies invest in them.

Firstly, Happy New Year. And secondly, yes this is another AI post. I promise not every newsletter will be - but it’s clearly a key topic right now so it is something that feels right to explore - albeit with the aim of focusing on a more practical analysis rather than just throwing more tools out there.

We all know that AI is starting to play a much bigger role inside agencies particularly across analysis, reporting, planning, documentation and the operational work that tends to sit underneath delivery and absorb time without ever really being questioned.

Now we’re seeing more impact on production itself - as AI continues to evolve with creative and development capabilities, doing more, better and faster.

The risk now is in how agencies respond.

If AI simply makes agencies quicker, cheaper and more comfortable working at a distance then very little actually changes. Delivery becomes more efficient but the experience of working with the agency doesn’t meaningfully improve and efficiency on its own has never been enough to sustain strong relationships.

AI isn’t going to replace relationships. Indeed, I believe they will become the vital currency to growth and retention. The more important question is whether agencies use AI to create space for better ones or whether they allow it to reinforce distance.

Distance has become normal. Presence now stands out.

Around this topic is also how we work as teams. Remote working has its place, brings many benefits and it isn’t going away - but many agencies have gradually lost sight of what’s harder to replicate when they spend less time with clients in person.

Staying in touch is straightforward with email and Slack, work continues to move along and it’s easy to feel close to a client when you speak regularly - but that isn’t the same as really understanding how their business operates day to day or what’s influencing decisions behind the scenes. It’s that competitive edge - that closeness to their business. 

Spending time in a client’s office changes that. You see how teams interact, where decisions get stuck, what gets debated repeatedly and what tends to get avoided. That context rarely appears in scheduled video meetings, but it can shape outcomes more than most formal inputs.

That’s where effort becomes visible. And that effort matters.

I think of it as an effort index - the time and attention that clients can see being invested in their world. 

Over time that builds confidence and trust and it becomes far harder to remove an agency that feels embedded rather than external. And this is a key signal for retention and advocacy.

AI should free up time. What happens next is the important part.

Used properly, AI should reduce the amount of time spent on work that gets in the way of thinking and judgement, particularly the tasks that exist largely to keep the machine running rather than to improve outcomes.

The risk is letting that reclaimed time disappear straight back into ‘busyness’, utilisation or margin improvement. If that happens the operating model changes, but the relationship model doesn’t.

That extra time needs to be used deliberately.

More time understanding the client’s business properly, more time thinking about what matters next (rather than what’s already happened) and more time spent with clients in the situations and places where decisions are actually shaped.

This is where the thinking behind Unreasonable Hospitality, a book by Will Guidara, the former co-owner and general manager of Eleven Madison Park, one of the world’s most highly regarded restaurants. His core idea is simple but powerful - exceptional results don’t come from obsessing over process or efficiency alone, but from an uncompromising focus on people, relationships and the experience you create around the work. 

I think this is particularly relevant for agencies - not as a prompt for grand gestures or over-servicing - but as a reminder that consistent presence and thoughtful attention compound over time. 

Clients rarely remember the mechanics of delivery - but they remember how it felt to work with you when things were uncertain or uncomfortable.

This is still the hardest thing to copy.

AI will continue to make execution easier to replicate and over time more agencies will be able to deliver broadly similar outputs at a comparable level of quality.

What doesn’t replicate in the same way is judgement, context and trust. Those come from proximity - from spending enough time together to understand how a business really works and from being present often enough to see the second and third-order effects of decisions.

When AI removes friction, it should create space for exactly this kind of work. Less time managing processes and more time thinking, questioning and being involved.

That tends to lead to stronger retention, more open commercial conversations and relationships that grow rather than reset every year.

What this means in practice.

This is the conversation I’m having more and more with agencies and there’s a clear sequence to getting it right.

The first step is recognising where this is heading and doing so with urgency. Not in a theoretical sense, but in a way that accepts that value is shifting quickly and that delay makes change harder later on.

The second step is moving quickly on the operational side by putting a clear AI roadmap in place, focusing on removing friction from day-to-day work and giving someone senior in the business the ownership of driving that forward and protecting it as an organisational priority - now.

The third step (and the one agencies often struggle with most) is being explicit about what happens once time starts to free up. If people are no longer spending hours writing reports or managing administration, there needs to be a clear expectation that this time is reinvested rather than absorbed.

That means rethinking how effort is measured. What does good look like now? How do you recognise time spent with clients, deeper conversations, earlier ideas and better judgement? How do you guide people towards that work rather than allowing them to default back to what feels comfortable?

Unless that becomes visible and measurable, behaviour doesn’t really change.

Why this matters.

Spending more time in the right client conversations tends to strengthen trust, surface opportunities earlier and naturally lead to discussions about new projects and expanded ways of helping.

That’s how relationships grow in value, as well as duration, and it’s before you even consider how the same thinking applies to new business where presence, insight and context matter just as much.

AI will continue to level the playing field on delivery. The difference will come from how agencies show up, how involved they are and how well they understand the world their clients are operating in.

That advantage very much remains human.

High level: use AI to create time and re-invest this meaningfully with more time with clients - not just on calls, but with them - in real life.