Brand Managers and their counterpart Category Managers see their roles evolving in the digital era with the advent of machine vision and AI analytics. First of a four-part series

For consumer brands and the stores that bring their products to shoppers, “vision” has taken on new meaning in the digital era. Success at the shelf has never been more important for trading partners, and the pressure to monitor and optimize shelf conditions for shoppers has never been greater:

• Shopper expectations have evolved as a result of experience with online shopping
• In-store fulfilment of click-and-collect online orders puts greater stress on service levels and creates more out-of-
• Poor visibility of in-store conditions and shopper responses stand in stark contrast to the granular information flows from e-commerce

In the era of omnichannel retail, with convenient alternatives a click away, stores have to work harder and smarter to compete. The fundamentals of merchandising compliance have assumed a heightened importance in every dimension – space, assortment, promotion, price, and especially on-shelf availability.

Minimizing lost sales due to out-of-stocks remains a primary financial motivator for better practice in this area. Controlling excess inventory, labor and shrink also bring financial returns. At the same time, trading partners are seeking a clearer understanding of how shoppers see the shelves and how they respond.

Time, accuracy and scale: a modern view

The core principles of Category Management and Shopper Marketing that have been the foundations of CPG merchandising for decades now require re-examination in light of these changing business realities.

Frustrating limitations stem from an inability to observe shelf conditions, ensure implementation, and evaluate performance outcomes in a timely, accurate and comprehensive fashion.

The arrival of Artificial Intelligence promises a way to tackle this triad of time, accuracy and scale with far greater speed and precision across thousands of stores.

Highly-effective in-store implementation depends upon the ability to measure and confirm promotion execution, identify and correct gaps in planogram compliance, and monitor shelf conditions with high accuracy. This has been an elusive goal for many years.

For brands, keeping tabs on these activities at arm’s length across the entire store portfolio has meant accepting much compromise until very recently. Self-reporting from field merchandisers has helped, but the data is incomplete and subject to lag. Tracking POS data provides some read of performance outcomes a er the fact, but it has never been sufficiently diagnostic for managers to understand all the causes behind a result.

Reporting on discovered merchandising gaps long a er it’s too late to correct them is ultimately an unproductive undertaking. But new tools and practices fueled by better, faster, more accurate data make insights more accessible and contribute to speed of action and greater confidence.

AI vision brings new potential

When it comes to in store visibility, real change is happening now with the convergence of three technology trends: Rich data on shelf conditions are flooding in from a host of new and more affordable camera sensors. Robotics are maturing from science fiction to practical reality at retailers like Walmart, Giant Foods, and Schnucks, which have committed to deploy roving shelf-scanners in thousands of stores. AI technology which can rapidly analyze vast quantities of images has reached production readiness.

These advances are opening the door to a new era for in-store implementation.

Snap2Insight, a Portland OR based AI startup, is working with brands and retailers to analyze thousands of retail shelf images collected using employee smartphones, fixed cameras and retail robots. Its AI technology identifies products and promotions, compares it against what is expected, and drives actionable alerts, such as for out-of-stocks.

King identified signage consistency and sufficient display capacity at the store to support promotion plans as key areas of improvement. “A lot of times it doesn’t happen, and we can’t do anything about it because we don’t know. Now we can know, and we can make things happen the way they are supposed to.”

Three key guiding principles

Mastering the shelf in the new era of in-store visual sensing will require a degree of discipline and focus, as with any other new essential retail practice. Three guiding principles apply:

• Know “shelf truth” – Capture data on actual shelf conditions in real time and take timely action to maintain compliance and maintain optimal stock levels... Systematically track how your shelf has been executed, on a store-by-store basis. Apply these findings to improve on-shelf availability, compliance and promotional execution.
• Collaborate more effectively – Build merchandising plans and forecast more reliably with confidence in the same shared set of facts about shelf conditions. Trading partners can tackle shelf-execution and merchandising using objective data and shelf insights generated instantly from images that capture true in-store conditions.
• Win at the shelf – Stay on top of implementation. Use shelf-image analytics to deliver exceptional in-store experiences for your shoppers by revolutionizing shelf execution and visual merchandising. Stock the right items in the right quantities. Price and promote to maximize profitable take-away.