AI-driven personalisation is fast becoming a marketing necessity, enabling retailers to optimise and scale the relevancy of their campaigns with unrivalled accuracy. Engaged retailers are gaining a competitive advantage by accurately reaching out to customers with hyper-relevant marketing offers, driving cost-effective conversation at scale.
AI can better customise experiences, adapt pricing and tailor promotions to meet the unique preferences of individual customers. It takes marketing capabilities to a new level by rapidly processing marketing permutations at a scale that humans cannot manually process. This includes selecting and orchestrating an optimal combination of factors — right person, right time, right product, right promotion, right price and right channel.
Starbucks offers a prime example. They are leveraging AI-personalised offers distributed via their Starbucks Rewards app to drive conversion, achieving 8% y-o-y growth in member spending so far. To increase success, they ceased mass promotions like Happy Hour and Treat Receipt and are instead emphasising gamified specials tailored to each Starbucks Rewards member’s preferences.
Personalised pricing and channels
With the power to coordinate the timing, value and channel for price changes on an individual customer level, AI is enabling retailers to reduce cannibalisation between substitutable products. For example, discounting multiple related products at once eats excessively into the overall margin of that category — a strategy that is no longer necessary.
Instead, retailers can utilise AI to harness the halo effect or “halo economics” — leveraging the pull of one product promotion to drive adjacent purchases within that category at a higher margin. Achieving this requires personalisation, which means promoting the right item for a given customer. Simultaneous promotion of other related items is not cost-effective.
BCG’s recent report, The US$70 billion prize in personalized offers, highlights that leading retailers who have redirected investments to personalised pricing offers are seeing financial benefits across categories and moving from piloting to scaling. The fashion industry offers a key example, evolving beyond its reliance on mass sales events to clear inventory.
The industry is now rolling out revamped loyalty programmes using off-the-shelf technology to enable personalised, challenge-based rewards. A US$1 billion ($1.34 billion) value fashion brand generated US$25 million in incremental ebitda using these tactics. To win market share, it is also key for retail brands to personalise channels.
Business objective — expanding online conversion versus driving offline inventory clearance, for example — is key to channel strategy. New channels beyond e-commerce can be considered as part of wider retail efforts. These include super-apps and social commerce across video, gaming and streaming networks. These growing channels have distinct customer demographics with varied preferences.
It is impossible to maximise omnichannel personalisation with manual management. Imagine 10 product categories with 90 substitutable products, with 100 customers all viewing these products, creating 9,000 unique substitution effects. AI can excel in automating personalisation for such an equation and tasks where humans fall short. Importantly, the feedback loop allows AI to track campaign performance and continuously adjust to improve it.
See also: Anthropic CEO says mandatory safety tests needed for AI models
Wider opportunities for AI in retail
AI offers a range of powerful additional opportunities for retail. Reinforcement learning — the ability for AI to learn through trial and error — is powering personalisation programmes at scale, helping retailers experiment with new promotion strategies while exploring proven ones through an automated approach.
BCG’s The US$70 billion prize in personalized offers report found that redirecting 25% of mass promotion spending to personalised offers could increase return on investment (ROI) by 200%, leading to a top-line growth opportunity of more than US$70 billion annually. Generative AI (GenAI) is unlocking further value in retail, enabling rapid content generation to experiment with in personalised campaigns and services.
BCG’s Fabriq marketing platform leverages AI-driven personalisation, using predictive capability for product selection and experimentation along with GenAI to support campaign automation and high-volume content creation. In a recent BCG survey of chief marketing officers (CMOs), two-thirds (67%) of respondents said they were exploring GenAI for personalisation.
Getting started with AI in retail
So how can retailers start to build AI-driven personalisation programmes? Data quality and technology infrastructure are important foundations for success. Without this, AI can lack training inputs, precision and scalability. However, the most challenging and time-consuming part of the transformation is empowering people and processes to adapt to and adopt AI.
Implementing AI personalisation systems will change the way many teams operate — marketing and sales, supply chain logistics, product design, customer service and more. Teams need to understand the value the technology will create for them, how it will alter their ways of working, and be encouraged to fully embrace this technology.
Every journey to embrace AI is different, yet a single thread connects pioneers of AI at scale — the 10-20-70 principle. BCG’s experience supporting companies to adopt new technologies shows that successful companies typically dedicate 10% of their AI efforts to algorithms, 20% to data and technological backbone, and the lion’s share of 70% to business and people transformation.
The 10-20-70 rule is at the heart of our AI approach. By redefining human-AI interaction and business culture, we help organisations unlock AI’s full potential. A robust personalised marketing programme requires thoughtful, effective execution of each component. However, with the right commitment and strategy, retailers can unlock the power and value of AI-driven personalisation at scale.
Jason Moy, managing director and partner, Boston Consulting Group and Jon Sugihara, managing director and partner, BCG X. The authors would like to give special thanks to Stephanie Brownlee and Saul Cai for contributing their insights to this piece
