How generative AI can automate customer marketing

Person completing a web survey on a laptop
(Image credit: Shutterstock)

Personalize, personalize, personalize. That’s been the message retail customer marketeers have been listening to on repeat for the past few years. Yet, despite the proven value of doing exactly that – personalizing customer communications and cross-channel experiences – it remains a headache for retailers’ today.

The reality is: producing several different versions of a message, let alone hundreds for every customer segment, requires a lot of time and strains the limited resources available to most marketers today. As a result, almost half of retailers have branded their personalization attempts ‘unsophisticated’, as revealed by recent Planning-inc research.

Retailer marketers need a nudge in a different direction. And over the past year, one specific technological advancement has caused shockwaves through almost every business sector: generative AI. In retail, it might just enable the mass personalization that marketers have been seeking.

Graham Burton

Graham Burton is the Chief Technology Officer of Planning-inc.

The technology roadblock

Personalization in customer marketing has become crucial to success. As competition has intensified, retailers need to do more to engage their customers. It has become harder to ‘cut through the noise’ and personalization can act as a key differentiator. Indeed, customers increasingly expect experiences and messages to reflect their interests and relationships with a brand and crucially, when done right, it works; for example we know that basket sizes can grow by at least 20% when marketing teams employ personalized recommendations.

However, personalizing content takes time and effort – a lot of it. As such, one of the biggest challenges for marketers wanting to create tailored content at-scale, ensuring each customer is served promotions or marketing messages contextualized to them, is a lack of resources.

Generative AI, based on Large Language Models, promises to change that. Many retailers already have what they need to exploit the technology: large troves of valuable customer data. By combining customer data with the use of generative AI, customer marketing departments can automate the generation of targeted messages in real-time, helping bridge the gap in resources. Marketers can use generative AI to adapt and shape these communications around an individual customer’s behavior, taking into consideration their preferences, purchasing history and current context.

In theory, they could use this data to present ‘perfect’ content to every one of their customers. And, through the use of automation, generative AI is critical in helping eliminate the resource barrier.

Say, for example, a customer clicks on a link through to a website page that contains a banner creative with an offer that has historically been generic; i.e. every customer sees the same offer, positioned in the same way. Generative AI could help tailor the offer on the banner to an individual. With the right integrations in place, the process would automatically trawl through an AI generated content library containing thousands, or even millions, of permutations to find the right message for that customer based on their current activity and historical behaviour, such as previous purchases and overall value to the business. Then, it becomes possible to regenerate the banner the customer sees in real-time, ensuring it's precisely relevant to them. The same process would be carried out for each customer; an impossible task without AI automation. This blends both copy and image creation in one fell swoop, delivering a highly relevant message that is proven to increase the likelihood of engagement.

The martech revolution

With the technological progress of LLMs, the role of customer marketing teams is due a shake up. Writing or developing content, which has traditionally been part of a customer marketer’s role, will evolve into reviewing content already developed by generative AI. This will save teams hours of time every week. The reviewer role will be crucial in training the AI, directing and checking content for accuracy, and preserving the brand identity and tone of voice. If implemented properly, the model will become smarter and learn over time to further reduce manual intervention.

It’s important to recognise, though, that only companies with fully joined-up data pathways will be able to exploit this new way of communicating with customers. To fully leverage mass personalisation enabled by generative AI, marketers must also contextualise customer data. To do this, they need a structured route between their companies’ single source of data and its destination to be effective. They should also select several priority use cases for the addition of LLMs into their processes and work outwards from there. This will help them avoid needlessly stacking new technology into their systems to no effect.

With generative AI models advancing rapidly, brands must be extremely mindful of the data they are inputting into open AI models. Commercially sensitive data, along with customer information, must be safeguarded. Not only so retailers can avoid giving away their competitive advantage, but to minimize the risk of data breaches.

Looking forward

Generative AI presents a transformative opportunity for retail customer marketers to finally unlock the full potential of mass personalization. Without AI, personalization at scale simply isn’t possible for marketers currently constrained by a lack of time and resource.

By leveraging troves of customer data, and harnessing the power of generative AI, personalised content tailored to each customer's preferences, purchasing history, and context becomes possible. Implementing processes further enhances the customer experience by dynamically adjusting content based on customer actions.

Commercial and data privacy risks mustn’t be overlooked. But, it’s undeniable that in the future AI will be capable of leveraging first-party data, revolutionizing customer marketing and delivering a seamless, highly personalized shopping experience for customers. Embracing generative AI represents the future of personalization in retail, and those retailers who adapt to this technological revolution are poised to gain a significant competitive advantage in the market.

We've featured the best CX tool.

Graham Burton is the Chief Technology Officer of Planning-inc. Working with clients such as Argos, M&S and Halfords, his work covers advanced customer segmentations, predictive modelling and award-winning recommendation engines.