Amazon, the world’s largest online retailer, is now 30 years old. The retail landscape has certainly changed beyond recognition, with omnichannel customer experiences becoming the norm for most brands. Yet, mistakes still occur, deliveries get lost, and customers can spend hours navigating after-sales support. Technology advances, but so do customer expectations regarding personalisation, convenience and speed.
Is AI the missing piece of the retail puzzle?
Global engineering, technology, and consulting service provider, Expleo, looked to explore how retailers are using digital, data and artificial intelligence (AI) to keep pace through a panel discussion hosted recently in one of London’s oldest retail locations, Covent Garden. The session was co-hosted by AI-powered test automation platform, Leapwork, who brought together senior leaders, who drew on experience gained at Ocado, M&S, Selfridges, Burberry and MUJI among others, to uncover how technology and AI are powering seamless customer experiences.
Here are ten takeaways that explore how to deploy, adopt and evolve AI and automation while addressing barriers to successful delivery.
1. Are there now 5Ps of marketing?
The traditional 4Ps – product, price, place and promotion – are still highly relevant in retail marketing and there are opportunities to optimise each with data science and AI. However, the rise in AI adds another element: P for Personalisation.
Omnichannel customers who shop both online and in-store are the most valuable and contribute a higher percentage of sales. So, it is worth investing in digital tools and data to better understand their needs. AI can manage that information at scale, to deliver the next best message or offer to customers and reward loyalty in both the physical and digital spaces.
2. So, prepare for personalisation on steroids
In the next five years, we will see the rapid development of sophisticated conversational search. We could soon enter a world where most of a brand’s sales are off-platform, i.e., through a ChatGPT (or rival) interface. Few brands offer truly dynamic experiences, but they’re coming soon. The traditional marketing funnel is collapsing.
However, there’s a temptation to use AI to overdose on customer insight. Yes, knowing the buyer is important—that’s a hygiene factor. But hyper-personalisation could become annoying or even creepy, especially if most of your customers just want to get in, get served, and get out.
3. From hierarchical to decentralised
AI is driving a shift from traditional hierarchical structures in retail to more decentralised autonomous organisations. By adopting a consultancy model, these self-regulated workplaces reward employees based on the value they add rather than the hours they work. Younger workers, in particular, expect flexibility, automation, and digital integration in the workplace.
AI is also automating many of the tasks traditionally handled by managers, such as scheduling, forecasting, and operations. Could retail move towards Autonomous Decentralised Organisations (ADOs), where blockchain and AI reduce the need for a centralised management layer? Or will AI actually lead to greater centralised decision-making?
4. Listen to the voices in our heads
Headsets on the shop floor are not a new phenomenon; however, their sophistication is set to increase with advancements in AI. Staff can be directed to different areas of the store based on peak activity or specific task management needs. Additionally, conversational bots could help staff respond more accurately to customer enquiries about product provenance or availability.
AI improvements to logistics and operations could eventually eliminate the phrase ‘out of stock’ in retail. By speaking into their headsets, staff will be able to track down that must-have product or instantly reorder another batch.
5. AI is human augmentation, not human replacement
Brands must balance efficiency with experience. While AI can streamline processes, customers still value human connection. Additionally, there is a risk that blind dependence on AI may lead to homogenisation, making it harder to differentiate between brands.
Retailers should take deliberate steps, such as retraining, to ensure that AI functions as an assistant rather than as a boss. Brands that seek to replace their staff with iPads may also risk losing their customers.
6. CTOs and CIOs are the gatekeepers
Creating a service that revolutionises the customer experience sounds appealing, but realising this vision can be challenging on the backend. As brands embrace AI more and more, they will delegate systems to external vendors, such as cloud and AI providers, resulting in rising costs from these third parties year after year.
The first-user advantage with AI may prove limited, as this technology could take several years to mature. However, all retailers need to begin their AI journey now. In particular, educating teams about prompt engineering will help them understand the value and incremental cost implications of automation on SaaS and hosting costs.
7. AI is an expedition, not a race
AI adoption will be the focus of numerous trials and failures in the coming years. Data privacy will pose obstacles, and the cost of cybersecurity presents another potential challenge for companies. It is far better to take small, cautious steps—test and learn, monitor the benefits, and conduct regular audits and progress reports—than to rush ahead.
Companies can’t just drink the AI ‘Kool-Aid’ and expect everything to fall into place. There will be challenges – and in some cases, the costs of AI may outweigh the efficiencies. With the rapid pace of change in retail, it’s easy to get carried away. Implementing AI for its own sake could result in significant expense.
8. Believe in people
Companies will agonise over choosing the right technology to meet operational or customer experiential needs. Still, those decisions may fail without a high-quality team to develop and implement solutions.
Brands need to adapt their AI solutions to meet their shoppers’ habits. High-value products traditionally require frictionless concierge service, whereas high-volume and velocity products better lend themselves to self-service and automated personalisation. (Or maybe they’re not! Ask them to find out.)
Luxury brands in today’s cost-of-living crisis are also diversifying to offer more affordable products – lipstick, scarves and gift items – that appeal to customers who love the identity but not the price tag. AI can help the brand differentiate and reach out to the relevant customer at the right moment, especially online.
9. Look past the short term
The rapid turnover in tech innovation can result in a short-term mindset, but it’s important to lay down strong foundations that will support the next levels of inevitable modernisation.
AI is not just about implementing something and then walking away. It is continuous improvement and optimisation. The right people and processes allow you to expand the perimeters and widen the map. The golden sequence is people, process and technology, in that order.
10. AI can’t get a free pass on quality
AI brings plenty of fanfare, but fundamental customer expectation is still just as high as it’s ever been. Everyone’s trying to make everything as seamless and consumable as possible, but if it looks simple to the end user, it adds tons of complexity behind the scenes. Therefore, quality assurance is so important across all your in-house systems.
