Q&A: Daniel Hulme, CEO, Satalia and Chief AI Officer, WPP
Satalia is an enterprise AI solutions and strategy provider, deploying world-class talent and proprietary IP to solve industries’ hardest efficiency problems across marketing, logistics and the future of work.
Can you tell us a bit about your background?
I wear a few different hats, both in academia and in industry. I’m the founder and CEO of Satalia, and the Chief AI Officer of WPP, where I’m focused on defining, identifying, curating and promoting AI capability and new opportunities for the benefit of the wider group and society. Through these roles, I’ve become a frequent speaker and a recognised expert on the topic of AI, the singularity and the future of work, and have done countless keynotes for household brands, publications and think tanks – educating 000’s of executives in the process.
My entire background is in AI, including my Masters and Doctorate from UCL, where I am now the Computer Science Entrepreneur in Residence. I am also a lecturer for LSE’s Marshall Institute, focused on using AI to solve business and social problems.
What does your company do? / What is your USP?
Satalia is an enterprise AI solutions and strategy provider. We deploy our world-class talent and proprietary IP to solve industries’ hardest efficiency problems across marketing, logistics and the future of work. Our solutions – used and celebrated by the likes of Tesco, DFS, PwC, BT and DS Smith – include vehicle routing, workforce optimisation, field service optimisation, manufacturing optimisation, demand forecasting and digital twins, amongst others.
Our team is motivated to solve the challenging, business-critical problems that very few others can solve. How do we price our products? How do we allocate our people? How do we optimise our supply chain? We’re not selling widgets or dashboards – we’re building solutions that are core to our clients’ operations, driving significant commercial, customer and employee value.
What’s special about the platform and your approach?
Our approach is very much problem-led instead of solution led. We’re not trying to use a generic product to solve a very complex, nuanced problem – we know that doesn’t work for clients. Instead, we work with established teams and domain experts to understand the problem in as much depth as possible before figuring out how we solve it, using our existing IP, a bespoke solution, or a combination of both. This means we can solve 100% of the problem and drive long-term value for the business and its stakeholders.
We’re also unique in that we have a combination of capabilities in data science and optimisation, or predictive and prescriptive analytics. What does this all mean? It means, for example, that instead of just predicting footfall traffic for a store (data science), we can use those insights to produce optimised staff schedules that better meet demand, improving conversion rates and the customer experience. Most companies don’t have data science problems, they have decision making problems, and we’re one of the few companies with the capabilities to solve them.
How does a product/service implementation actually look like and how do you measure success?
Having identified a problem to solve – which we do by prioritising against objectives, available data, reusability and other key criteria – we start by scoping an initial MVP, which helps to de-risk any future investments. This could be as simple as proving (or disproving) a hypothesis using data. A retailer for example, might think that weather has an impact on footfall in their stores. We can test this assumption and others to identify what factors (weather, marketing spend, time of day etc.) impact footfall, which we use to build a demand forecasting model. For some projects, we’ll start by building and testing a prototype on a subset of the business. If building a staff scheduling tool for in-store employees, we’ll start by deploying in a single store, review the impact on a predefined metric, say conversion rate, and then iterate from there. Ultimately, we use an agile process but are flexible on the details depending on the clients’ needs.
How are retailers using your systems to gain competitive advantage and what does best practice look like? Can you share a case study with us?
Retailers use our solutions to differentiate across any part of the value chain, be that be faster deliveries, better stocked stores, more relevant communications, improved customer service or lower prices.
But it’s slightly more nuanced than that. The best practice is to recognise that AI is not a silver bullet or a short-term fix. But something that, if done properly, and applied to the core of your business, can provide hard-to-replicate moats and advantages far more significant than any off-the-shelf product could. The best retailers understand this and augment their existing domain experts with specialised partners to solve their toughest challenges.
DFS are a great example and have been very open-minded about using AI to essentially ‘solve’ end-to-end retail. We’re working with them on a number of projects, from supplier reliability forecasting to in-store workforce optimisation, to middle and last-mile delivery optimisation. And the benefits speak for themselves – 18% improvement in in-store conversion rates, 8% increase in NPS and 18% reduction in fuel consumption.
What challenges and opportunities do you see in UK retail for 2023 / What challenges are retailers facing in 2022?
Retailers have long seen the supply chain as a cost-centre instead of a source of value creation.
But I think retailers – partly due to recent supply chain woes – are starting to realise that’s not the case and that a robust and agile supply chain will separate the winners from the losers in the coming years.
As well as continued investment in ‘front-end’ applications of AI, such as marketing, pricing, communications and service, I expect to see more investment in ‘back-end’ operational applications of AI. Things like vehicle routing, warehouse optimisation, inventory optimisation reduce costs whilst simultaneously improving the customer experience in terms of choice, price, convenience or reliability.
Our latest research with Wunderman Thompson provides baseline into AI in 2022, a baseline from which AI will undoubtedly grow and flourish. At Satalia, we’re excited to help support as many businesses on their AI journey as possible.
What is on the horizon for you as a company?
Hiring! We joined the WPP group last summer and have been working on several internal and external projects. We’re working towards becoming the go-to AI agency in the network, similar to what DeepMind is for Google. We’re past the ‘shiny new toy’ phase now, but the interest in our work doesn’t seem to have slowed down, so if you know anyone who wants to work in AI, send them our way!
If you want to learn more about Satalia, our work or AI in general, please access the following resources:
- Website
- A keynote by Daniel Hulme at a Google Event.
- Our latest research, in collaboration with Wunderman Thompson Commerce, seeks to uncover the truth behind what global business leads and consumers really think of and understand about AI, revealing what the future has in store, and – more importantly – the clear growth opportunities in its adoption. Wherever companies are in their AI journey, Satalia can help.
