So far, we have covered two of the four most common approaches to scheduling. The first was service-based scheduling, which is also known as interval-based scheduling. We discussed supermarket cashiers and baggers as examples of this where the activity is closely tied to the specific day and time interval when the work occurs. You cannot do the work ahead and excess service in the morning does not offset inadequate service in the afternoon.

The second approach was non-service or production scheduling. This is where data are captured at a daily or weekly level. Also, the timing of the work activity, as associated with the volume driver demand (e.g., cases stocked), must be defined rather than implicitly tied to historical data. Grocery stocking is a good example.

Our third scheduling approach is a composite of our first and second. Think about a supermarket deli operation, particularly during the evening hours. Much of the deli work is tied to servicing customers with the data (e.g., customers served, or pounds of product handled) tied to the exact timing of when those customers require service.

However, many know that deli is a very intensive department that requires multitasking. In between serving customers at the service counter, deli clerks perform fixed or variable production tasks. Some of those tasks are scheduled at specific times, while others are intentionally executed amid the flow of customers. That dynamic, the combination of interval-specific work along with non-interval specific production tasks, is what makes for this service and production scheduling approach.

I note this because many retail and specialty retail departments rely on this versatile blend of on-demand customer service tasks in accordance with production, cleaning or preparation tasks in the background. Some systems handle service tasks well, others handle service tasks or production departments only, but few systems handle both within the same department – and it makes a big difference.

If your system doesn’t do this exceptionally, the system may force you to break the department in two and complete separate planning for the “front of the house” (the service portion) and the “back of the house” (the production portion). Yet, it will fall short of the optimization you can get with a system that is designed to support both scheduling types within the same department.

As one learns with the basic tools of carpentry, home maintenance or auto repair, you want to use the right tool for the right job or you will probably bash your thumb, bruise your knuckles, or wish you bought the better toolbox when you did. In our next post, we will introduce the fourth most common scheduling approach. Then, we can move into more detailed discussions on specific system features.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

When we last discussed Service or Interval-based scheduling we discussed that requirements are clearly associated with activity occurring at specific time intervals (usually 15-minute intervals) through a specific day. Cashiering at a supermarket is a great example. There is no forgiveness for being over in the morning and being short-staffed in the evening. Correct placement at the interval level is critical for providing the associated service to customers.

Not all activities or departments in a retail business work that way. Having steaks cut and ready for customers when they need it, having grocery shelves filled and stocked, having displays look appealing all are necessary parts of retail work planning. However, did that steak need to be cut at 8:15 or could it have been cut at 6:45 or 9:00? Obviously, there are some boundaries on the window to support product freshness, but when it was cut is not nearly as important that it is in the case when it needs to be shopped. There is no direct correlation between the work and exactly when it needs to be done.

Now, I’m an advocate for product freshness and we work with stores to help shape best practices around time of day standards for freshness, merchandising, and inventory levels on hand. Logile’s Day Clock™ visuals have really helped organizations communicate and implement these merchandising expectations better than ever before. But no one will make that argument for a can of peas or a 4-pack of bathroom tissue. When to stock is optional, it’s not tied to the act of customer shopping in the same way that cashiering or bagging in store Front End operations are.

So, if I’ve made the point that some activities do not have a clear timing directly associated with them or implicit in the historical data, I also want to be clear that when you develop a work plan the timing of such activities need to be decided. You might have one store that stocks grocery loads overnight, and another that works the load through the day, and that’s fine. But each store needs to have the activity for grocery stocking defined if you want to create a schedule and an associated work plan to run your store at standard time meeting all built in service, safety, and other expectations. It needs to be defined. Failure to do so will almost always result in lost time, indecision, and unmet expectations. Applied to perishable production, it leads to freshness issues, shrink, and disappointed customers.

Since the interval data is not significant for these Production or Non-Service departments, data is usually captured either at the weekly, or preferably at the daily levels. If you are using POS data to capture the information, then a lead/lag factor might be needed to anticipate when the replenishment for those items sold needs to occur. If a store only gets two deliveries a week, the workload for stocking that load may get distributed over two days, not just one. And stores with limited shelf space may anticipate a certain amount of restocking the day or days after the initial stocking occurs.

Time windows are also typically used to take the time earned and to locate it into the workplan during the appropriate hours. Some activities have multiple time windows. A good work planning system will give you good options to profile the timing to fit your operations.

You will get more details about how this work in upcoming editions of Scheduling Insights as we discuss task-based scheduling, defining task instances, options to distribute daily demand within task windows, task dependencies (you need to unload the truck before you can stock the load) and the staffing parameters that convert raw demand into scheduling requirements.

For now, just keep in mind that different departments have different work planning and scheduling needs. It is important that you have a labor management system that meets the needs for all types of departments that you need to schedule for.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

Picking up from our last Scheduling Insights introducing the four basic types of scheduling, we will explore Service Scheduling which is sometimes also referred to as Interval-based Scheduling.

It’s an understatement to call this type of scheduling the most commonly understood and practiced type of scheduling. The whole idea of retail labor scheduling was built around scheduling jobs (or tasks) like Cashiering and Bagging. And it’s a very clean, clear process from capturing history in 15-minute intervals and using that history to forecast future needs for those same 15-minute intervals in future days.

Scheduling a Front-End cashier at a supermarket checkout lane is the perfect example of service scheduling. History is captured directly from the Point of Sale (POS) transaction logs and is summarized in 15-minute intervals. At the very least, the register captures the amount of sales performed by cashiers across all Front-End lanes in every 15-minute interval through the day. Better approaches also capture the workload drivers of the items processed and the transactions tendered. And the best of approaches will capture the detailed workload drivers of the item and tender processing so that you know how many items are scanned versus key entered, versus scaled in each 15-minute interval; and how many tenders were made in cash, credit with a signature capture, credit without a signature capture, debit, checks, electronic benefits transactions (EBT) and WIC program transactions in each 15-minute interval of any given day. Other actions that cashiers are required to handle are also captured in the interval data. Examples of those activities would include loyalty card interaction, coupon tendering, removal of security tags, and other miscellaneous activities.

The bottom line of service or interval-based scheduling is that history is captured and detailed in each 15-minute interval. Forecasting then utilizes this history to develop a workload for each 15-minute interval. Labor standards translate that workload into raw labor demand. Allowances and staffing parameters (which we will discuss in greater detail in upcoming sessions) then translate that raw labor demand into your scheduling requirements and workplan.

The important concept for Service Scheduling is that the forecast is detailed to the 15-minute interval and the work is tied to each specific interval. It doesn’t do you any good to overschedule cashiers in the morning if customers need to be cashiered in the evening. Two wrongs don’t make a right! Each interval stands on its own work requirements.

For most automated scheduling systems, this is very straightforward. Your forecast, hopefully as detailed as possible, is summarized in each 15-minute interval. Once you apply your standards to that interval data in each 15-minute bucket you have the engineered hours of work content or the raw demand hours required to perform the anticipated workload volumes.

Why are these not the final scheduling requirements or final staffing demand? That’s because there are several more staffing parameters that need to be applied before we finalize the requirements for scheduling. These are the most widely used parameters for Service Scheduling:

Sustainment – What happens when you have 4 intervals requiring 4 cashiers, followed by another interval calling for 7 cashiers, and then several more intervals of 4 cashiers. How can you work your schedule to go from 4 cashiers to 7 cashiers and back to 4 very quickly? Is that really what is required or is that a datum anomaly? Most systems handle this by looking both backward and forward from each interval to determine if the requirement is sustained over time. This can also be referred to as a smoothing factor as the result is that it smooths or normalizes the curve of interval requirements into something that can be better scheduled.

Rounding – Although most systems schedule on the 15-minute interval, what happens when your demand is for 2.3 cashiers? Is that 2 cashiers or 3 cashiers? Rounding, or various processes associated with rounding, determine the final whole number requirement for each interval.

Minimum or maximum requirements – It’s fine for the system to calculate that early morning demand calls for only 0.4 (four-tenths of one cashier) in the 7:30 time interval. But what if you require at least one dedicated cashier at that time? If you have a minimum staffing requirement, then anything less, whether it meets the normal rounding factor, must be increased to the minimum of 1. Suffice it to say that the same holds true on the maximum side. It’s hard to schedule 7 cashiers if you only have 5 registers to open.

The beauty of Service scheduling is the direct relationship from the forecast to the necessary placement of the hours. There is no interpretation. The forecasted workload you anticipate servicing at 3:45 is the direct basis of what you need to schedule at 3:45, barring minor adjustments for staffing parameters, like those I have mentioned above. This direct relationship between the forecast data, the workload and the scheduling requirements, the workplan, defines Service Scheduling.

What happens if that relationship does not exist at the 15-minute level? We will take up the next type of scheduling, Production or Non-Service Scheduling, in our next installment.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

In this edition of Scheduling Insights, we will be building on our discussion of the lifecycle of a labor schedule.

Given that you now have a good foundation in the terms and process steps from forecasting to demand planning to staff planning to scheduling, it’s a good time to discuss different types of scheduling. In this edition, I will lay out the basics for understanding the 4 most prevalent types of scheduling you will want to understand for retail scheduling, so we can discuss the needs and process for orchestrating good solutions for each of these very different types of scheduling challenges.

My focus here will be primarily on scheduling for retail, including all types of retailing from grocery to big box to specialty retailing. If you are scheduling for another type of business, hang in there, as the needs of retail cover a wide range of scheduling strategies for most other business segments too.

Let’s start by understanding the four most distinct types of workplan modeling and scheduling for retail that derive from the type of work you are scheduling. I will also provide an example of each.

1. Service or Interval-based scheduling (Cashiering in a Front-End checkout at a supermarket)
   This approach applies to a type of work where the labor demand is aligned directly with dynamic volume as is the case for cashiering or bagging in larger retail environments where that work is performed mostly by dedicated employees.
2. Production or Non-Service Scheduling (Stocking on an overnight stock crew at a big box store)
   This approach applies to volume work that can be quantified in total but the timing of when the work needs to be performed is not directly associated with interval forecast data. The business could elect to do this work when customers are present or might use overnight or low volume hours to perform this work. Timing for this work is flexible and must be configured within scheduling parameters.
3. Service and Production Scheduling (Clerking in a deli department at a supermarket)
   This approach is a hybrid mix of our first two approaches where some work is directly associated with customer or item traffic as captured within the interval data, but other work components may be defined and patterned into the schedule where you choose to do the work. Some production work may be used to fill in the valleys, or low points, in the service work. However, as volume gets larger this may require a greater separation of duties and specialization which allow the service and production functions to be staffed independently of one another.
4. Coverage or Workstation Fill Scheduling (Staffing a Manager on Duty in a large retail store)
   There are numerous jobs where a workstation must be filled for a defined time regardless of volume. In some businesses a Manager on Duty follows this strategy, although other business may choose to base that on volume. Receiving clerks, layaway counters, service desks, greeters, security clerks and bookkeepers are examples of jobs that can often be scheduled in this manner. This is a simple matter to define and to schedule, but you may want to be sure that your approach allows for flexible configuration of shifts for full and part time employees depending on the composition of your workforce. If you use a “fixed shift” approach rather than defining your requirements, then not all associates may have the ability to work the particular shift length you define.

These quick examples illustrate that different types of scheduling are required for different store functions or different departments.

Remember back when your dad or mom or buddy told you how important it was to use the right tool for the right job?  Well that insight into home maintenance also applies to scheduling. And, it’s a big reason why many organizations fail to bring the benefits of automated scheduling to all portions of their business (what we often refer to as “wall to wall scheduling”).

The fact is, most systems are designed to handle only one or two of these scheduling types. So, it is essential that you understand your business needs and use the right tools (processes and applications) to understand your forecast work volumes and translate those work volumes (e.g., cases, items, customers, fitting room occurrence, etc.) through effective demand planning and staff planning to create your final scheduling requirements.

Each of these scheduling approaches can be implemented at different levels – scheduling at the department, job code, or task level are all possibilities. We will get to that topic too.

In our next edition, we will take a deeper dive into the most universal type of scheduling in retail:  Service or Interval-based scheduling.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

In this edition of Scheduling Insights, we will be building on our discussion of scheduling terms.

We have outlined the process as a series of steps, as follows:

Forecasting
The use of history and other insights to predict key metrics (like sales) and workload volume drivers (like cases to stock, items to scan, cash transactions to tender, pounds and packages of deli products to serve, etc.)

Demand Planning
The use of labor standards applied to those workload volumes to calculate “raw” or “engineered” work content time.

Staff Planning
The translation of the “engineered time” into scheduling requirements including the timing of the tasks or activities by day of the week and time of day, and the modification of the raw requirements to include all the time needed with allowances, service requirements, minimum staffing requirements and other staffing parameters considered in preparation for effective scheduling.

Scheduling
The matching of specific resources (availability, skills, other qualifications) to the needs called for in your staffing plan.

Those are the major steps leading to the creation of a schedule. Many systems see this as the entirety of the process. Later in this series we will discuss that we see and provide functionality for additional steps, most specifically in working the schedule successfully through the week in progress (“Operations Management”) and post-scheduling analysis so that we create a learning loop for continuous improvement in the optimization of user skills and system management of the process for future weeks.

The Lifecycle of a Labor Schedule

In addition to these process steps, it is also helpful to think of the various states a schedule may go thorough because the whole idea of effective scheduling analysis is designed to minimize the variation you experience in each of these iterations. Here is a perspective on what those iterations might be, and I will use an example of an organization with centralized schedule writers supporting the stores in the creation of their schedules. Organizations not using central schedulers would remove the steps performed by the central scheduling team.

1. System Generated: The system generated forecasts, requirements and initial schedules
2. Centrally Reviewed: The central scheduler reviewed and edited forecasts
3. Store Reviewed: The stores reviewed and edited forecasts (usually this review involves both department and store management)
4. Published: The schedule is published at the posting deadline defined as company policy. This is when employees see their schedules and work assignments
5. Week in Progress Revised: Until the week ends various changes may be made. Shifts may get swapped, changed, deleted or new shifts may be added
6. Final schedule as Worked: This is recorded through the actual punches of employees in the Time and Attendance system

Understanding these iterations of a schedule’s lifecycle is helpful to understand the work needed at each step, the system functionality required and how to analyze the effectiveness of the entire process from the development of the forecast estimate of what will happen and what resources will be needed to the point of knowing what did happen and what resources were really earned (required) by the volumes achieved.

We will reflect of different types of scheduling in our next installment.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

If you want to understand the process of scheduling and to discern what is different between various system offerings, it is helpful to understand some of the basic terms that you are likely to encounter.

A good schedule starts with a forecast.  Simply put, a forecast is a prediction of what the business will require.  A sales forecast predicts the volume of sales your store will do during an upcoming specified period of time.  A weekly sales forecast predicts your sales for an upcoming week.  A daily sales forecast predicts sales for a given day.

Sales is not the only metric that is forecast.  In fact, more advanced forms of forecasting are about forecasting workload volumes rather than just sales.  Workload volumes are the volumes of things that create the work for your employees to do.  When you buy groceries, your cashier’s workload comes from the items you buy and the tender transactions he/she must make to process your order.  The sales are only a bi-product of those items that you choose to purchase.  The items are what defines the work that needs to be performed to ring up your order.

Defining workload volumes

Here is an example that I think helps to understand volume drivers and illustrate how having the best approach to forecasting workload volumes and using time standards can optimize your results.  Think about service work at your local grocery store’s Deli counter, and let’s compare three different scenarios:

1. One person buys one package containing 10 pounds of roast beef.
2. One person buys 10 packages each containing 1 pound of roast beef.
3. Ten different people each buy 1 package containing 1 pound of roast beef.

If we only forecast the sales, and Roast Beef is $9.99 per pound this week, then each of these scenarios totals out 10 pounds of roast beef and sales of $99.99.  But the work content between these scenarios is very different.  Put yourself in the place of the Deli clerk who must meet and greet each customer and take their order, then slice the roast beef to order, and make the package(s) as requested.  Each of these scenarios has the same sales forecast, but the workload volumes are really the customers, pounds, and packages that our Deli clerk must process.  People might refer to these as workload volumes or as forecast elements, or volume drivers.  But keep this example in mind because it’s a good one to remind you that all approaches to forecasting and defining workload volumes are not equal.

Understanding work content

When forecasts identify the true workload volume, then we can develop the clearest understanding of the work content that we are writing our schedule to satisfy.  Work volume might be items, pounds, cases, shirts, etc. but work content is about time – hours, minutes or seconds.  Once we forecast workload volumes we can calculate work content, or work demand, by using labor standards.

Labor standards can be very basic or very specific depending on how they are developed.  Some organizations use reasonable expectations (REs) or rule of thumb standards because they do not have the systems or resources to create more specific or detailed labor standards.  This is not a bad approach to get started.  Our Deli clerk would much rather get time for the customers, pounds and packages she/he must produce rather than just an amount of time per dollar sold.  But, as you have an opportunity to do so, it can help to have more detailed standards that take into account the type of equipment you have, the number of steps that need to be traveled, the packaging requirements, and the other specifics of what the work requires at this particular Deli to calculate the right amount of time fairly.  Generally, standards of a more detailed nature are referred to as Engineered Labor Standards.  An engineered labor standard defines how much time it takes to perform a specific method for a specific volume unit.  Note again that just as we learned that just as all approaches to forecasting are not equal, neither are all approaches to building engineered labor standards!

Applying forecasted volume to time standards

When you have a forecast volume and a time standard for that volume you can multiply the two together to get what is known as raw work content time.  This may also be referred to as engineered time or raw demand.  The time is considered “raw” or “engineered” because while it tells you how much work is required, it is only a portion of the resource time you would need to schedule to have workers perform this work.

Transforming raw demand into scheduling demand

Several things are missing from raw demand to get to scheduling demand.  Allowances for breaks or other personal fatigue and delay are not in the raw demand calculation.  Also, let’s say the raw demand required 0.67 people to do the work.  Can you schedule 67/100ths of a person?  In most systems, scheduling is done on the quarter-hour, so you can round down your raw demand to .5 hours for this work, or you round up to 0.75.  But what if we must always have one person scheduled for this work?  Or what if we must have one but no more than 2?  These are minimum and maximum constraints.  All are examples of what we collectively refer to as staffing parameters.

Getting an understanding of your raw work requirements is critical, but additional processes occur to transform that raw demand into scheduling demand. This process of transforming raw demand to scheduling demand is usually referred to as staffing or demand planning or a part of developing a workplan.

We will cover all these staffing terms in greater detail as we discuss how automated systems can handle staffing to best meet your needs and optimize your scheduling.  And, just like we learned for forecasting and for engineered standards – not all systems do it the same way, or nearly as well.

With these basic terms in mind you will be ready for our next edition of Scheduling Insights.

In his series, Scheduling Insights, Dan Bursik provides insights and strategies around effective retail labor scheduling, addressing a diverse array of challenges and topics. To read the previous edition, click here. To search for all editions of Scheduling Insights, click here.

How difficult can it be to plan to preposition the resources in advance of when you need them? Well, as many labor managers and schedule writers know from experience that depends.

The Scheduling Insights series of blogs will explore issues relating to effective scheduling and allow you to reflect on some of the approaches and strategies learned over years of experience that can help you write better schedules, especially with the help of automated scheduling software. I will start with basic concepts around different scheduling needs, suggest some of the pitfalls to avoid with systems, and help you get to know some of the features and functions you want to have working for you in a top-notch scheduling solution.

Let’s start by thinking about what scheduling is all about: prepositioning your resources that meet the needs of your business, company or activity in advance. It is likely that most people think about scheduling people, and that’s often what those doing scheduling are asked to do. But not always. There are many different types of scheduling. Sometimes you are scheduling people, but sometimes the task of scheduling resources is equally important. Have you ever been scheduled to attend a meeting, and no one scheduled a projector – or a room to conduct the meeting in? How well would it work to schedule a train crew without scheduling power (the locomotive) to move their load? Or scheduling a hospital crew without representing all the technical skills and specializations needed to cover the work?

Scheduling is often perceived to be the process of deciding who gets what shift, or whether Tom or Ken or Aimee or Marty or Jane or Betsy get that prized morning shift. The task of scheduling is more about understanding the job to be done, the resources required, and having a process to develop an effective plan and to match your resources to that plan. While you never want to forget that real people have needs that must be considered in the schedule writing process (like when they are available to work); it is important for good schedule writers not to focus on those concerns, but to instead focus on what needs to be done when to create the best plan to meet the needs of your customers, your business process or other key considerations. A good schedule will take those personal considerations into account as appropriate to your business rules or desires, but that comes after you create your workplan.

I always say, you cannot create a good schedule from a bad forecast, and that axiom bears repeating here. A good workplan begins with an accurate forecast of the work you are intending to service. Understanding the work content of that workload is how you translate that volume into scheduling requirements. Those requirements may be for equipment, facilities, and the skills and qualifications of the people you need to perform the activities defined by that work and by your business. That translation of forecast volume to workload to workplan all occurs before what people think of as “scheduling” but it lies at the heart of what makes for good scheduling or poor scheduling.

Never forget that good scheduling requires a good forecast. That forecast is key to defining what work your schedule needs to cover. The process of translating that workload into a workplan – up to and including your specific scheduling requirements is different depending on the type of work you are doing and the data available for forecasting.

That’s the subject of our next edition of Scheduling Insights.