Essential for All Retailers: AI-Powered Workforce Management

Suvarna Sundarakrishnan, Director of Engineering – Predictive Analytics, shares her expertise on the use of AI for retailers addressing the current disruption.

Artificial Intelligence & Machine Learning Adds Value in Disruptive Times

These are uniquely trying times for retailers seeking to serve their customers via precise workforce schedules and optimized labor patterns. During the COVID-19 crisis, the creation of accurate schedules for those open retailers—pharmacies, and grocery and hardware stores among them—is more complex than ever, as the press notes. And now-closed non-essential retailers face immediate store-level scheduling challenges.

Essentials: Grocery, Pharmacy Stores

In past weeks, there have been unprecedented customer traffic spikes at pharmacies and grocery stores in particular. Customer activity and store workload are high throughout the day; while simultaneously, the crisis has caused some stores major reductions in associate availability.

The reasons vary: sometimes the employees may themselves, unfortunately, be sick or at-risk; or they may lack child or dependent care; or the store business volume is outstripping existing associate resources. The result: a limited staff pool spread thinner and thinner with changing shift demands.

Add to this:

• Adjusted timing: Sometimes stores are open longer hours—or with limited, specified hours for particular shoppers, such as the elderly
• On-boarding and training new employees are essential and continue, despite ongoing COVID-19 response challenges
• Product inventory levels are frequently inadequate, because of disrupted or stressed supply chains and abnormal customer demand—employees must continuously watch for stock-outs and ensure rapid replenishment

This overall situation is rife with challenges, which artificial intelligence (AI) and machine learning (ML) technologies can address.

Workforce Forecasting and Scheduling Blind Spots

Traditionally, workforce forecasting and scheduling solutions and processes use prior-year data and trends to predict upcoming workforce needs. Accordingly, the forecasting algorithms embedded in typical industry solutions will provide inadequate guidance, since the needs of today’s stores are vastly different compared to the same period last year.

In addition, some retailers are already coping with predictability pay requirements, while facing new or altered regulations. This forces businesses to juggle compliance efforts, while scaling their operations up or down, depending on their individual COVID-19 responses.

Incurring Pandemic-Related Workforce Costs

As we discussed above, during the pandemic, prior-year data is largely inadequate for generating an accurate forecast. And combining this data with old models and systems—which are unable to factor in new paradigms—results in unreliable forecasts and substandard scheduling.

Besides the reliance on irregular or unique data sets, managers must also cope with a limited labor pool, which can include irreplaceable pharmacists and others. This can result in:

• Overtime costs (due to unavailable associates)
• Labor law violations and subsequent non-compliance fines
• Unmanageable workforce shortages, spikes in customer demand, and the like

AI-Powered Workforce Forecasting

Retail demands a highly precise and realistic alignment of employees with demand. Fortunately, in these (and similar) circumstances, retailers can use AI-powered workforce scheduling. This enables nimble adaptation to new COVID-19—or future—workforce challenges.

Using AI, the system can perform drill-downs and make connections; it can reveal such things as otherwise impossible-to-see customer traffic patterns. These models can factor in the varying availability of different types of employees: cleaners, cashiers, stockers, and so on.

The solution can then forecast customer traffic by time, type, and other factors for a comprehensive view of future demand. And AI and machine learning-based models are fully configurable, so that business managers can quickly adjust forecasting models to support their own needs.

AI in Retail Action

Here is a hypothetical example: Managers mine a three- to five-week period, from, say March 2020, and extract data points that are weighed more heavily than long-term history. This process provides a more accurate prediction during the pandemic. Any uptick or downtick in the trend that is connected to the COVID-19 situation should be included—potentially using a growth factor feature.

The benefits include:

• More accurate predictions of customer demand and associate availability
• Optimized labor costs
• Reduced overtime costs
• Improved customer experience
• Greater scheduling equity
• Higher employee morale
• Continuous operational improvement in critical business areas

Ultimately, we live in an age of disruption and unpredictability. Despite all challenges, the AI-powered forecasting enables more intelligent operations throughout a retail organization, with potentially huge cost savings. The end result is a nimble execution template that supports a great customer experience, even during and after the COVID-19 crisis.

To learn about Reflexis AI-powered Workforce Management solutions, please join us for an informative webinar: “Solve Labor Scheduling Challenges with AI-Powered Workforce Management” on Thursday, April 30, 2020, from 2:00 PM-3:00 PM EDT.