If you are considering buying or running companies with about 50 or more hourly/shift employees, implementing workforce management/optimization (WFM/O) best practices immediately after acquisition can reduce costs, improve customer experience and improve employee satisfaction. This is especially effective where operations are standardized or can be standardized and/or provide services outside normal business hours (evenings, weekends, overnights). Often times I have found small to mid-size companies ignore WFM until it becomes an untamable monster because of the complexity involved or lack of planning time/resources. So, here is a structured breakdown of WFM activities that can be used to improve organizational efficiency and a source of competitive advantage.
WFM includes five distinct set of activities - Forecasting, Capacity Planning, Scheduling, Real-time management, Adherence measurement. Forecasting and capacity planning should be done keeping financial planning in mind (led by your CFO or wear your CFO hat while doing this). Scheduling and real-time management should be done keeping operations in mind (led by your COO or wear your COO had while doing this). The last one, adherence measurement, is a team effort which involves identifying where things did not go according to your WFM playbook, how much expense is incurred in each area that did not adhere to the playbook, and what areas need to be prioritized for better adherence.
Forecasting is a continuous activity that happens through the year in labor intensive organizations. It is best to have multiple forecasts that anticipates expected demand for labor planning for different time horizons. Annual demand forecast to budget for labor costs and to plan capacity, quarterly demand forecast to fine tune capacity plans, monthly/weekly forecast to schedule employees and daily/intra-day forecasting to fine tune your employee schedules. Keep in mind longer the forecasting horizon, less accurate your forecasts will be. So, it is better to forecast at a an aggregated level when the horizon is long (i.e. to plan for 2021 budget, your forecast should be at an aggregated at a monthly level, for the upcoming quarter your forecast should be at an aggregated at a weekly level, for the upcoming month it should be at a daily level and for the upcoming week or next few days it should be at an intraday level). If your tasks are multi-day tasks, you may not need a daily forecast. You can stop it at weekly level. If your tasks take only a few hours, you will have to forecast at an intra-day level.
Now the question is how to actually do it? What tools to use and what methodology to follow? That brings us to a key dependency. We need historic data and well thought out plans on new initiatives. To project baseline demand, use historical data and on top of that add additional demand based on the initiatives you have planned (adds demand) and headwinds you may face (reduces demand). You can use simple ratios (to extrapolate labor hours needed per unit of task) to complex models. WFM modeling is a vast topic by itself and can be covered in a few paragraphs here. But, feel free to ask any specific questions in the comments below or DM me if you’d like to chat.
Capacity Planning –
This is the part where companies plan the labor force needed through the year and plan for hiring, retention and employee churn by branch or by activity type. At budget level planning laying out capacity needs by month will get you to think about spikes you may have through the year (summer, back to school, holidays, winter, spring, etc.). Capacity plans can be used to come up with hiring/training plans depending on the amount of employee churn your business may face and spikes or troughs through the year. Quarterly refinement of capacity plan can be used to manage excess demand with extra time/staff augmentation and excess capacity with voluntary time off/up-training employees. Capacity plan directly impacts your budget. A good checks and balance system should be developed. What I have found useful is to look at the demand for any given week this year and see how much we spent to meet that demand. Then look at the anticipated demand next year and see how much you plan to spend to meet the demand for the same week next year. Of course, you will have to account for any cost saving initiative you have planned to implement. Keep in mind not to hire for the peak demand or average demand. This might leave your workforce underutilized (waste) or over worked (burnt out). Sweet spot is to hire for 70 to 95th percentile demand depending on the volatility of the demand and manage excess demand with extra time/staff augmentation and excess capacity with voluntary time off/up-training employees. This gives your employees opportunities to earn extra cash during heavy demand and time off during low demand. If effectively managed, this can improve employee morale and lower labor cost.
Scheduling processes determines the placement of employee activities. At a minimum, a plan with how many employees start their shifts at what time and how many finish their shifts at what time aligned with the daily/ intraday demand pattern will need to be developed. If your company allows for further standardization, planning for non-productive but necessary activities like breaks, lunch, trainings, team meetings, etc. should also be developed. In addition, this plan should also include non-productive activities like time offs, absences, etc. If you operate multiple branches within commutable distance, scheduling can help optimizing workforce by shifting staff between branches depending on the daily demand at each branch. If you have multiple activities cross training employees between tasks can also help switching employees between tasks. While creating schedules, accommodating inputs from individual employee will increases employee morale. On the flip side, frequent shifting of employees between locations or tasks will increase burnout.
Scheduling process is a constraints based optimization exercise. Spreadsheets are just enough if the number of constraints are reasonable. If not, there are open source industrial solvers that can do such optimization exercises. Again, WFM modeling is a vast topic by itself and can be covered in a few paragraphs here. But, feel free to ask any specific questions in the comments or DM me if you’d like to chat.
Real-time management (RTM) –
Actual demand is never exactly the same as what’s forecasted. Also, actual capacity on any given day is never the same as what was planned or scheduled. This is where real-time management comes into play. RTM helps keep a pulse on demand and capacity at a daily or weekly level and directs traffic based on fluctuations. RTM is a process that helps you keep your operations flexible and responsive to real-time surprises. Example of a few things that RTM can execute: If you see demand is higher than forecast on a given day/week, request overtime. On the other hand if demand is lower, then offer voluntary time off. RTM processes can help postponing or advancing planned auxiliary tasks like team meetings, maintenance, etc.
RTM is a rule based process. If your demand is up so much, take a set of actions that you have determined beforehand. Similarly, if your capacity is so much above the demand, take a set of actions that you have determined beforehand.
Adherence measurement –
This is where you measure how well your WFM playbook has worked. How accurate your forecasts have been? How well did you plan your capacity (hiring/retaining/employee churn)? How efficiently you scheduled your employees and managed them in the real-time? There are metrics for each one of activity in this value chain that can be customized for the kind of activities your business conducts. The key is to understand how much does ‘miss’ on each metric costs. Then it becomes easy to prioritize which metrics to focus on first for improvement.
Here are some high level metrics that can be customized for your business:
Forecasting – Mean absolute percentage error (MAPE)
Capacity planning and scheduling – Variance in planned capacity vs. forecasted demand
RTM – What part of the planning errors are mitigated
Finally, executing this well could save businesses a significant portion of OpEx, reduce missed sales or service opportunities and improve employee morale. Businesses need not have a separate team or personnel dedicated to this. Just partnership and communication between finance and ops on a well thought out set of policies, rules and actions will do.
Feel free to comment with questions/thoughts/challenges or DM me if you’d like to chat about any of the above or need additional details. Happy to help.
(briefly about my background: I’ve been running WFM operations and FP&A for a F500 company that has multiple subsidiaries with a labor force ranging from 50 to 7,000 shift/hourly employees)