There’s a huge potential for every generative AI company in every industry. I work largely in consumer retail and banking, and see a lot of opportunities every client should look into across the full value chain, for three reasons. First, this is really a big opportunity, and significantly different from anything we’ve seen before. Secondly, every organization should think this through right from value back. Thirdly, it needs risk-aware scalability built in right from the start.
In terms of opportunities, when it comes to customer interaction, for example, consumers can now interact with a chatbot that feels human. It can suggest items based on what they’ve bought before. How cool is that? But you can still request human assistance for things the machine can’t do. Design is another example, because even smaller brands can now afford to generate a significant number of designs that previously required hundreds of junior designers. When it comes to the supply chain, gen AI can identify more sustainable pathways as well as less sustainable ones, and forecast demand for different stores in different regions, reducing the amount of merchandise shipped to wrong locations. So there’s also a very significant impact on sustainability and CO2 reduction.
You also need to look at gen AI from a value-back perspective, which means understanding where the value lies, be it in an economic sense, a sustainability sense, or from a customer experience sense—which will probably translate into something economic. And although gen AI will add a lot of value, it’s crucial to bear in mind that in many cases, all that’s needed is regular AI. That’s a very important point we’re always trying to emphasize to our clients—understanding the value and understanding the best solution to capture it. Why? Because you want to maximise the return on the investment. For example, the supply chain sometimes requires nothing more than a normal AI approach to provide a lot of value, which you can then supercharge later on with gen AI.
Finally, scalability and risk-awareness scalability must be built in. Why? I’ve seen many clients who invested a lot of money in very fancy, shiny use cases, but never really scaled. They just created super prototypes that never delivered value. So you really need to think through what data you need, what text you need, and what you need in the beginning and later one when you want to scale. Consider that right from the start.
The other aspect is what I call being risk-aware. With gen AI, we’re all aware of the risks. They could be reputational, they could involve intellectual property, or they could result from biased results. Since you don’t want to risk destroying the brand, you need a solid risk-management framework encompassing both the gen AI algorithm and the data, of which gen AI requires massive amounts. So it’s critical to think about the data and creating an operating model that allows you to use it at scale, makes it accessible to everyone, and ensures its quality—while fostering a culture where everyone uses it in the best possible way.
