HAMPTON, N.H. – A lot of the excitement and noise in IoT has centered on manufacturing, healthcare, smart city and consumer, leaving some of the other important verticals underserved. With that in mind, I was very interested in a Jan. 9 IBM Retail Update where IBM shared some of its retail strategy, which had aspects related to IoT and AI, ahead of its participation in the National Retail Federation show in New York from Jan. 15 to 17.

IBM underscored that retailers must reform now, with IoT and cognitive capabilities, to compete with innovators such as Amazon. Fast-followers will lag behind into marketplace uncertainty, including in engaging with the digital consumer; differentiating by providing customized, localized experiences and engagement; and becoming more agile through new operating and business models supported by IoT solutions and cognitive computing.

To help retailers answer this call to reform, IBM is focusing its retail solutions around three key imperatives: shopping and customer experience; merchandising and supply networks; and operations and innovation. These imperatives include products such as Watson Marketing, Watson Commerce, Watson Merchandising and Watson Supply Chain. Potential use cases addressed by these products include enhancing product selection and design, improving marketing campaigns, improving customer services, and optimizing talent management.

[VIDEO: Watch a video overview from IBM at: https://www.youtube.com/watch?v=zpvZYV-vyGA ]

IBM outlined some deeper examples of how its Watson cognitive products will allow retailers to leverage AI-assisted insights and optimization. One example is exception monitoring, where AI can monitor millions of transactions 24/7 for anomalies. If a product was selling very well but had a poor profit margin, AI could determine it was due to multiple discount codes or rewards being applied to the transaction when the anomalies should not have been allowed. The AI solution could send an alert or fix the problem immediately. It could also recommend new strategies to improve profitability for that item if it was not a one-off error.

Supply chain management

Cognitive machines can also quickly manage supply chains to improve fulfillment and can balance customer requirements with business requirements such as using data to optimize the fastest and cheapest shipping paths or shifting inventory between fulfillment centers based on regional demand.

IBM stressed in its retail update that consumers are more knowledgeable about products and want to get in and out of stores as quickly as possible. Retailers are struggling to provide services that facilitate customer return, cross-selling or upselling. IBM highlights that its cognitive solutions can assist and enhance the ability of store personnel to serve customers by tying together back- and front-end systems via a mobile device connected to an IoT solution. This solution will allow personnel to quickly and painlessly process returns and observe customer information such as previous purchases, wish lists, and discounts or promotions for which customers are eligible.

IBM also highlighted its Metro Pulse solution, announced in July 2016, which is a cognitive platform aimed to be a data scientist’s playground. The “as a Service” platform, which can be hosted on an IBM, third-party or customer cloud, ties together data from IBM and partner platforms, such as weather and foot traffic (e.g., The Weather Company), events (e.g., Ticketmaster), demographic and economic data (e.g., U.S. Census Bureau), social (e.g., Twitter), and transactional data (e.g., MasterCard).

“Hyperlocal intelligence”

IBM positions the platform as a source for retailers to gain “hyperlocal intelligence” to tune operations to customer traits and exploit opportunities based on neighborhood insights and product demand signals. Examples include a local sports game played on a warm day, which may drive higher foot traffic to stores nearby for apparel sales necessitating more staff be on hand, while rain may reduce foot traffic but increase demand for umbrellas or rain gear. Or, targeted marketing promotions could be tailored to neighborhood demographics and purchase patterns, as buying behaviors of young professionals differ from those of upper-class families.

IBM noted that one of the adoption difficulties for its entire Watson product is that customers, including retailers, have seen it as a “black box.” They understand it performs analytical functions but are not sure how this is accomplished or which products they need. The retail update shows IBM is taking steps to reduce complexity and provide concrete use cases. However, in IBM’s Retail Solutions Guide, the “selected IBM offerings” for each use case still lists nearly 33 solutions or products for each strategic imperative.

IBM should continue to streamline around each imperative and focus on selling the result rather than the components. For example, its marketing and engagement imperative should be a solution or application marketed to solve a particular pain point that can be easily plugged into the Watson IoT platform, allowing customers to start benefitting from it immediately.