A recent study by the Capgemini Research Institute examines the multi-billion dollar opportunity that artificial intelligence (AI) offers for retail. AI deployments among large retailers have increased exponentially in recent years, but many use cases have yet to be tapped. With feedback from over 400 retailers, analysis of more than 40 real-world use cases, and secondary research of the top 250 retailers’ current state of AI deployment, the report not only investigates the status quo, but also offers a practical framework of use cases for companies that are just starting their AI journey.
Is your foundation set for AI?
Input from retailers lends some much-needed realism to the discussion about AI deployments. Retailers are slowly realizing that AI and machine learning (ML) are not silver bullets. In fact, without proper groundwork on information architecture and data quality, they can cause a big data mess. However, when data quality is ensured and use cases properly prioritized, tangible benefits and business growth occur.
Retailers in the apparel and footwear subsector are trailblazers in terms of AI penetration, followed by food and grocery retailers. eCommerce retailers, a driving force behind digital transformation overall, are predictably driving AI deployments as well, using AI for product searches, product recommendations, and to personalize the customer experience. Such deployments primarily rely on ML, chatbots, natural language generation, or image and video analytics.
AI use cases focused on improving customer experience
As a digital customer experience developer, I find it interesting that the majority (74%) of AI use-case deployments are customer facing, and only a quarter are operations focused. Retailers believe that AI will significantly improve customer satisfaction and reduce customer complaints. We all know Amazon is the pioneer in AI utilization, but there are others. For example, US-based apparel retailer, Guess Inc., partnered with Alibaba to deploy Fashion AI, a concept that meshes digital into brick-and-mortar stores with smart mirrors and a new kind of fitting room. French food and grocery giant Auchan combined gamification with AI-driven, personalized promotions. John Lewis (UK), Macy’s (US) and Zalando (Germany) all use image-recognition technology to offer product searches with the ability to find similar products based on image. And eBay estimates that AI and ML will drive incremental sales worth one billion dollars per quarter.
A multi-billion dollar prize at operations cannot be ignored
However, when it comes to the development of operational processes and how to connect them to customer-facing improvements, retailers are still missing a huge opportunity. The largest retailers are better poised to lead AI adoption since they have bigger investment budgets and higher purchase transaction volumes. It is easier to see the benefits and gain the full potential of machine learning faster with bigger transaction volumes since they can be scaled to multiple markets. Twenty-eight percent of retailers have deployed AI in their organization but only 1% of the initiatives reach full-scale deployment. Most are focused on sales and marketing concepts whereas AI offers opportunities across the value chain. Most initiatives aim to tackle complex use cases, while it would be smarter to start with the low-hanging fruit.
Retail value chain opportunities
The research provides some very practical guidance on how to get started with AI. It includes a framework of use cases across the retail value chain and example use cases by retail subsector, including:
- supply chain, AI-based optimized routing using ML
- multiple AI robots combined to collect and fulfill orders faster
- predicting shopping behavior patterns to prevent returns
- optimizing product selection to minimize inventory
- combining ML and image-recognition to improve in-store shelf product placements.
The top AI use cases have been categorized by retail subsector to help retailers identify opportunities. As for digital projects in general, the agile development approach is also a smart choice for AI initiatives. This means deploying MVPs (minimum viable products) first for fast roll-out and data-driven optimization. It is important to start with quick wins through simple use cases, so that it is easier to scale deployments and gain momentum for AI adoption. AI is not a panacea, but with clear focus, good ground work, and keeping the customer at the heart of these development initiatives retailers are facing a multi-billion dollar opportunity.
Interested? Contact me or read the full research paper here.