DEPT is a tech, marketing and consulting agency that describes itself as ‘50% tech and 50% marketing’ – the 4,000 strong agency employing 40 in-house AI experts.
FutureWeek sat down for a conversation with Ali McClintock, SVP of Growth & Head of UKI at DEPT, to find out how AI can be used to help brands with global audiences, why using AI for efficiency is skimming the service, and more.
What AI tools do you use at DEPT?
Internally, the way we use AI is changing everyday. About half of our 4000 people workforce are in engineering and experience – CX, design, deep engineering and coding.
We’re cloud agnostic, so we build AI powered applications with all of the major LLMs and machine learning models – Claude, OpenAI, Gemini, and open source models. Google and OpenAI are our clients, which helps us understand where they’re going as businesses so we can incorporate those roadmaps into our projects.
Everyday we work with the tools that you’ve heard of: OpenAI, the Adobe AI stack, Midjourney etc, but we also build our own internal tools. We house all of our AI solutions under a platform called ‘DEPT Orchestrate’ – which enables any of our teams to access both internal tools and media tools to help with planning, buying and optimising our media.
How are these tools being applied?
We use AI to create mock-ups for clients and for visualising ideas. But we also use AI for trend analysis and forecasting and addressing niche audiences for big brands who have markets in different countries.
We’ve built trend forecasters that know a brand’s key product moments, values, and understand what’s happening in a specific geographical location. That’s really informing our creative ideas and delivery. We also use it to help build customer experiences, whether that’s in websites or applications.
How are you using AI in client content?
We use AI in content generation, QC, and transcreation – which is the process of adapting a message from one language to another.
Transcreation has been a big theme for us. This tech takes a base language and then translates it into another language whilst overlaying your brand tone of voice and nuance. That used to be exclusive to copy or stills, but now we are seeing it move onto videos. So we are able to dub now, not just provide a voiceover, but change how their mouth moves to align with the new language.
With content generation, people are still nervous from a regulations perspective. For any partner that’s a regulated business, we also use AI to help us adhere to regulations and QC against those regulations.
How is AI changing the business models of brands?
Media has historically been focused on a return-on-investment model. Through AI, we are changing that model and putting value in the customer lifecycle. For example, if a customer returns to a shop or how likely they are to return half their items – because these are really damaging to a business’ profitability.
For that retailer, we have been able to impact their ‘bottom line’ by building a customer lifecycle that we optimise towards rather than return-on-investment. The outcome is that we can reduce their returns but also acquire valuable customers for them.
For another client that’s a travel agent, we have built an AI system that uses their data and knowledge to build itineraries based on customer profiles.
From that point of view, AI is less about driving efficiency and more about using it in the product or service to help them be better than their competitors.
Can you give some examples of how your work has been impactful for clients?
We used GenAI to make product ads for Footlocker more engaging and improved click through rate by over 30% with our tools.
One of our other clients, JustEat, has 90 different markets. AI allowed us to build performance assets at scale and personalise them to both the market and the restaurants which are available to them wherever they are based.
We have built tools for forecasting, optimisation, and personalisation. So for example, if you’re EasyJet, which is another one of our clients, we might advise them to make content around ‘the top five places to visit in Barcelona’, because now lots of people are searching for that on TikTok or ChatGPT rather than Google – brands are having to consider Generative Engine Optimisation (GEO).
Why are businesses primarily adopting AI for efficiencies over core business outputs?
I think people are scared, and using AI for efficiency is a way for them to better understand the technology. If you don’t fully understand something, you’re unlikely to trust it to solve core business challenges.
We’re seeing businesses that want a five-year AI plan before committing – but that’s impossible in this space. You don’t need a big strategy; the strategy is simply: ‘How do I learn?’ AI is evolving so quickly that you need to have skin in the game.
