How to make human capital management systems demand forecasting smarter with external data
Staffing optimization has been a notoriously difficult problem for businesses from restaurants to retail companies. It’s also one of the more costly problems, so the battle to reduce over and under allocation of staff is a top-of-mind challenge for many executives, especially during a recession.
Having too few employees scheduled causes slower service leading to a decrease in customer satisfaction, and a loss of revenue from customers who aren’t being served. Yet having too many staff members scheduled on a slow day leads to operational redundancies where you are paying for staff who don’t have enough work to do. Studies by MIT’s School of Management found staffing stores accurately can lead to a 10% increase in sales. In these chaotic and dynamic times, leading companies are using data and artificial intelligence to meet exactly the amount of demand they will experience.
Why is workforce planning so hard?
It sounds easy in theory, but optimizing staffing is a difficult and dynamic process that is inherently costly and time consuming to do at scale. When sudden disruptions such as severe weather occur, they can lead to a surprising decrease in demand and result in overstaffing. On the other hand, your customers could experience a perfect storm of demand due to multiple impactful events occurring near each other, which can cause a surprise rush for their business.
Demand for consumer packaged goods, travel and hospitality, and retail all fluctuate drastically based on external events. Whether it’s a severe storm, a university’s undergraduate students returning to class, a socially-distanced community event or a sports game (both at the stadium and broadcast), all of these elements impact demand. Therefore, it’s not enough to simply rely on historical sales and transactional data in order to predict future demand and your labor requirements. You need a better understanding of the environment your stores are operating in. Your forecasts are only as accurate as your data sources, so it’s crucial to integrate forecast-grade data into your models to make sure you’re accounting for all relevant demand causal factors.
You need more than just historical data to forecast accurately
Human Capital Management (HCM) and Human Resources Information Systems (HRIS) software programs help solve for coordinating employee schedules. Many of these systems include smart features that analyze previous sales data to make recommendations for staffing, but in a particularly anomalous year like 2020, relying purely on historical data isn’t going to cut it.
Since many businesses are finding themselves in a data deficit, they’re leveraging external data sets to help them discover previously unknown demand causal factors that affect their day-to-day operations. Many have identified that events – or their absence – are big factors that can influence demand. Workforce management tools like Legion and Lineup have been able to get ahead of the curve by adding events as a feature in their models. This provides a number of benefits for your customers and gives you an edge over your competitors.
By taking impactful events into account when making recommendations for labor forecasting, your models are more informed and your company is more resilient when demand anomalies arise that couldn’t be predicted from just historical sales data. Improving this process is a unique competitive advantage for HCM platforms as it enables your customers to leverage your platform as a one-stop-shop for all of their workforce needs, from recruiting employees to scheduling their hours. Furthermore, integrating high-quality data into your platform improves synchronicity accuracy for data across your products, which makes your entire product suite that much more appealing to your customers.
How HCM companies leverage intelligent event data for more accurate forecasts
Building resilience into HCM and HRIS forecasting models is crucial. If 2020 has taught us anything, it’s that businesses need to take real-time data into account when planning for future demand, instead of just relying on historical data. Forecasting models are typically built for accuracy, but by incorporating new sources of information that add context to why demand incrementally and decrementally shifts, you are left with more resilient models that can adjust to previously unprecedented fluctuations.
Integrating demand intelligence data into HCM prediction models gives you the advantage of offering your customers a solution that includes a single source of truth to accurately determine staffing schedules. This allows your clients to not only control labor costs but also increase revenue opportunities. For example, on a predicted slower day, managers could run promotions to offset lower foot traffic around a store. If you don’t already have a workforce optimization solution, events can be a great starting point to give you and your customers insights into when they might experience spikes in incremental or decremental demand.
What does this look like in practice?
For HCM companies that cater to airline or transportation businesses, it’s important to mitigate any potential risk to drivers by knowing when it might be unsafe to operate in certain locations. Having insights into where Disasters and Severe Weather like floods and hurricanes are occurring helps ensure that you’re not scheduling drivers and pilots in inclement weather.
HCM providers that cater to restaurants can use event data to better understand foot traffic patterns around their locations. For instance, a coffee shop near a local university typically sees an uptick in business during exam week, when students are studying and caffeinating before their finals. However, with many schools operating online in 2020, you can use Academic Events to track school closures and prepare for a slower exam week than in previous years.
Retailers in city centers can be susceptible to higher demand when events like Festivals and Expos drive more foot traffic around their stores, and that’s something HCM systems need to be able to plan for. Knowing the Aggregate Event Impact around key locations on a particular day can inform not only how many employees need to be scheduled on hand, but also what level of experience or expertise they would need depending on the volume of traffic predicted.
How HCM companies use PredictHQ’s data
PredictHQ’s intelligent event data works as an input to labor planning forecasting models to help you take events into account when making staff rostering recommendations for your clients. We track events in 18 different categories with global coverage. Events like Public Holidays and School Holidays influence foot traffic around key locations, while scheduled attended events like Sports games often drive up demand for food and beverage deliveries. We even provide a feed of unscheduled events that are disruptive for your customers, such as Severe Weather, Disasters, and Health Warnings that could put their employees at risk. This event data comes enriched with intelligence and rankings, so it takes the guesswork out of understanding the impact it will have on key stores or locations.
When you take PredictHQ’s event data into consideration for your staffing recommendations, you’ll be able to:
Use Beam to correlate impactful events – or their absence – with peaks and valleys in transactional sales data to understand exactly which events move the needle for customers, whether you cater to quick-serve restaurants, accommodation providers, or brick-and-mortar retailers.
Reduce errors in forecasting to create more reliable models by tracking events that trigger both incremental and decremental demand.
Even if an unscheduled impactful event occurs, you can receive Notifications in real time to make sure you’re adjusting your staffing plans accordingly.
Leverage our Ranking system to get a more accurate understanding of not only an event’s time and location, but also the relative impact that it will have on your customers. Not all events are created equal, so Rankings provide an estimate of the scale of the event, so that you’re not over or under-forecasting for them.
These are just a few examples of ways we help HCM system providers. If you’re interested in learning more about these use cases and how we can help you make smarter workforce planning decisions.
