We're using AI. Here's exactly how, and where we draw the line.

Our goal is to stay in the top 10% of AI users in marketing globally, no matter how steep that learning curve is.

The right question to ask...

If you've asked us what our policy on AI is, that's the right question. If you haven't asked us, you probably should have. The white-collar world doesn't work the way it did even six months ago, and the pace of change is only going to double, then double again. The marketers who will have the most impact aren't the ones who are waiting to see how it lands, but those who have thrown themselves head first into the brave new world.

We’ve already invested heavily in training and process to make sure our team is genuinely ahead of where most agencies are operating, and we have a long-term commitment to continuing this effort.

This is our honest account of what we believe, what we're doing, and where we think this is all going.

What this is not, however, is a finished article. Our views are going to change as the world changes, and as we learn in real time what our team members and brand partners need from us. As with everything we do, we want your feedback.

What we believe

AI makes great marketers better. The thinking, the craft, the judgment can't be automated in full, but it can be kick-started, provoked and made more robust with good AI coaching. Everything around the core thinking absolutely gets faster, from briefing to research and idea generation. The marketers on our team who were already excellent are now getting significantly faster. Our thinking is now stress-tested against ten different frames instead of just a Brand Hacker and their manager, so it's tighter.

Our first job is to turn our team of marketing experts into marketing superhumans. Human expertise that is smart, kind, high energy and a joy to work with, powered by tools doing jobs in minutes, that used to take hours.

In the future, we see a new model for how a brand partner might interact with us. Instead of our team offering days, you might be licensing our technology as well as our time.

For now, you should expect us to move fast, but there isn't yet a silver bullet. The focus has shifted from time spent on admin and research to time spent on truly identifying the key problem and sourcing the data that will inform the thinking. We are spending our time differently, and you will get more from our time than you would have done a year ago.

Where robots are making a difference already

It’s always day 1. Right now, here are some examples…

• We build brand persona's in AI so our LLM's know our brands as well (if not better) than we do (using brand guides, TOV, call transcripts etc) - This improves research, ideation and gives our team their own personal assistant on every brand.

• We use Claude to draft decks in seconds that would have taken hours - ensuring our time is spent on value not admin.

• We synthesise data and identify trends and patterns in research and insight projects so that we spend time on interpretation, not data analysis.

• Our team all use AI notetakers, so we never miss a detail, insight or follow-up task and to ensure context is well documented & we’re not spending time organising information.

• We supplement real-world content with AI-generated product imagery (using Nano Banana and Adobe Generative AI to provide more diversity and range in our paid media creative outputs.

What we won't do

We won't share work that hasn't been interrogated by a human. AI doesn't get to be the last pair of eyes on something before it reaches you.

We won't use AI image generation as a substitute for a talented content team with a camera. The gap between generated imagery and the kind of visual storytelling that actually builds a brand is still enormous. Real places, real people, real moments is where much of brand gets made.

We won't pretend that AI can replace what you get from a proper conversation with a customer. Sitting across from someone who genuinely loves your brand, who might make you laugh, who tells you something about why they buy that no survey would ever surface. A language model can synthesise what people have already said. It can't tell you what they haven't said yet.

We won't lose ourselves in an LLM and stop looking at the world around us. Inspiration comes from seeing how customers shop a supermarket, from watching other brands activate, from meeting other marketers. We won't stop being able to tell you what we think is really cutting through.

Join us on the journey

If something we send you feels like it came straight out of a chatbot, tell us. If you've got a tool you think we should be testing, tell us that too. If you want to understand how we're using AI on your specific account, just ask. We're genuinely happy to talk about it, because the best client relationships right now are the ones where both sides are learning out loud together.

If you want to hear more about our training programme, or you're interested in getting involved in how we're building this, we'd love to hear from you. This stuff is moving fast. Our views will change. Our tools will change. What won't change is that we'll tell you exactly where we are with it, and we'll always be a step ahead so you don't have to be.

(This was written with the help of Claude, but a cup of tea and 30 mins of voicenotes reflecting).