Managing a shift-based workforce can be a challenge, but with the right tools in place, it doesn't have to be.
In this Q&A interview, Deputy's Senior Solutions Expert Steve Woods talks you through how to forecast demand in your business and ensure you have the right number of staff on each shift.
First off, what is demand planning?
Demand planning is looking at supporting businesses to better allocate labor to changing levels of demand through their organization. So making sure that you've got essentially the right number of employees working the right roles at the right time.
If you imagine a hospitality business, they've got traditional peaks and troughs throughout the course of the day: breakfast, lunch, and dinner.
So demand planning is about ensuring you’ve got just the right level of labor to hand during those different periods of the day. And effective demand planning brings different forecast data to the platform, such as sales data and foot traffic. Deputy has also looked at doing really advanced things like analyzing local events and seasonality, time of year, whether it's school holidays, et cetera to forecast the total number of orders the hospitality business might have per hour per day into the future.
What are the major challenges in getting demand planning right?
The first challenge is getting a level of accuracy around the forecasting process so it's providing a valid level of predictability into what's coming down the track for that particular organization.
And then, when we've got the forecast in place, I think the next challenge is looking at how we build relevant labor models. So let's say that I've got a forecast that says I've got 100 orders for breakfast, 200 orders for lunch, 300 orders for dinner. If that's the case, then how do I gradually ramp up my labor throughout the course of the day? How do I know how many employees I should have working for the number of orders that are coming through the organization? How do I map the forecast that I've so meticulously created in part one to the actual labor requirements that I've got in the organization itself? So at this stage, we look at things like the number of employees you need working versus the number of, let's say, orders coming into the organization per hour per day.
And then we can set floors and ceilings around that as well. So you know, you might only have the capacity to scale to a maximum of five kitchen staff.
What makes an optimal demand plan forecast?
After we’ve gone through the process above, there's lots of different work we can do to create different labor models for different parts of an organization. When those labor models are run, they provide businesses with a forecast of their staffing levels.
So in this scenario, we've gone from a forecast on the number of orders coming into an organization to a plan for the number of employees that they need, working in different roles in different departments per hour, per day.
And the important thing to mention is that even when you have that, it's still not the end result. Because if I've got a forecast that says I should have five employees working at 9 am and eight employees working at 12 pm, then I can achieve that with a variety of different shift types, shift durations, and shift start and end times.
But what's the optimal shift pattern that's going to help me allocate my employees in the most effective way?
We work with clients to better understand the different types of shift lengths required and when shifts should start. We combine this with forecasts on required labor levels to create an optimized shift pattern. The final stage is then automatically allocating the right employee to the right shift. To do this, we look at their qualifications, availability, whether or not they're on leave, and any other custom requirements that might influence the shifts you want your employees to work.
What’s the level of accuracy for shift pattern creation?
The more precisely we can forecast, the more accurate everything becomes further downstream in the process. Which is why, when we work with clients, we always start with forecasting. It’s key to generate accurate forecasts for core demand drivers (e.g. sales data) that take into account variables such as seasonality, weather, or local events. Pattern creation accuracy depends on the parameters that a customer puts in.
What role does the sector play in accurate forecasting?
The forecast we use is very dependent on the sector. For a care home, we're forecasting occupancy levels, and for a hospitality organization, we tend to forecast things like the number of orders coming in or foot traffic.
What are the most common misconceptions managers have about automated demand planning?
So when we talk about demand planning in general, there can be a bit of confusion in the market.
One common misconception is assuming technologies such as AI and machine learning are there to take over the role of labor planning entirely. AI is incredibly good at prediction, and the more inputs we provide, the more accurate those predictions can be. What it lacks, however, is a manager’s ability to rationalize, adjust, and course correct to changing events on the ground. It can't reason like a human can.
What’s a more accurate way of seeing this use of technology?
Here at Deputy, we use the term assistive intelligence. So it's about businesses not handing over everything to an automated process because there are still things that managers on the ground are going to need to do. Instead, it’s about this relationship between automating the more transactional parts of the process, but then relying on the institutional knowledge that managers have got to refine and tweak the recommendations that the platform's making. So together with our partners, Deputy is automating 70% of the manual tedious work, but we're still relying on the vast experience of managers to complete the process.
To summarize, what’s the best way managers can effectively manage their workforce through changing levels of demand?
Having the right solutions in place that can marry up changing levels of demand to different labor requirements and availability is key. Labor is the highest cost for many businesses. Without real clear visibility on what's driving that cost, it becomes very difficult to plan for changing levels of demand and increasingly expensive to both sustain and scale your operations.
If you're interested in efficiently reducing labor costs and ensuring optimal coverage across your schedule, check out how Trek achieved this with Deputy.