Changing to new Shift Patterns
Staff Scheduling & Staff Rostering Software: Visual Rota
Forecasting Labour Requirements
All organisations need a manpower plan and the essential prerequisite for any manpower planning is a statement of demand for the products of the organisation. Whether it is a Nursing Home providing care, a hospital providing surgery or a factory producing cars, they all need to know the demand for their service or product. This can be a known demand based on the orders received from customers or an estimated, forecast or expected demand. Sometimes, it is the ability to provide a service which creates the demand, or the demand for a service which creates the service. It is often the establishment of a delivery date and schedule that turns the demand into orders.
The more accurate we can get in the forecast for demand, the more accurate the delivery date and schedule. There are many methods of forecasting the demand for a product and this is an area for mathematicians, statisticians, econometricians, marketing and operational research personnel. The more effort that can be spent on the subject of accurate forecasting, the more chance the organisation has of surviving and prospering. There are basically two types of product for most organisations, the first being a standard product produced in large quantities to satisfy a large reasonably stable demand, the 'bread and butter' side of the business. The other type of product is produced in fewer quantities to satisfy specific customers. These could be few in number but make a significant contribution to the total profitability. They could be a new market for the organisation. Forecasting demand for these is difficult. Forecasting for the first type, the standard product, is not easy, but various means have been devised, and the most obvious methods involve the assumption of persistence, that what is happening will continue to happen. Most mathematical methods involve looking at past data and making decisions based on the data and a measure of personal opinion to forecast the demand in the future.
The future can be any valid period of time, a day, a week, a month, a quarter or a year, or more. Having decided on setting the future demand for, say, a year, it makes sense to determine staffing levels for the same period. Hence, annual hours, which is a very efficient method of using manpower in production.
Annualised hours, what are they?. We usually hire staff to work so many hours per week because
If we look at a few examples, then we can assess the situation.
Example 1. Staff contract hours = 37 hours/week and you require the staff to work 8 hour shifts. This equates to having 37 shifts every 8 weeks. There is no exact equivalent for annual hours, the closest is 241 shifts of 8 hours.
Example 2. Staff contract hours = 37 hours/week and you require the staff to work 7.5 hour shifts. This equates to 72 shifts every 15 weeks. There is no exact equivalent for annual hours, the closest is 257 shifts of 7.5 hours.
In the above examples we have the staff hours expressed in hours/week and we require the staff to work shift lengths which are not exactly divisible into the contract hours. The reason we need to have this inequality is for efficiency and to maximise profit. If we have a look at the relative efficiencies, then we can assess the situation in terms of cash. If we assume in example one that the staff work 4x8hrs + 1x5hrs (which = 37hrs) each week, we lose 3 hrs productive labour per person per week, which is around 7%. In addition there can be the losses from idle production equipment during holiday periods. We also have the problem of continuity of staffing if there is a 3 hr gap every 5 shifts., We could have this covered by having other staff cover this period with another 5 hour shift and have a 2 hour overlap. This can minimise the production loss by only losing 2 hours for 2 staff each week, which is about 3% lost production. However, we have a problem devising a staffing schedule that will do this. The best schedule would be to have 10 staff operating 9 machines, but here we are allowing the staffing to dictate our production. From a pragmatic viewpoint, we would choose to close production early on one day. However, we could not do that if the production process had to be continuous, such as a chemical plant which could not be closed down, in which case, there would probably be an overtime payment of 3 hours per week to all the staff.
The problem with weekly hours is the inconvenience of having nothing but prime numbers to deal with when you are trying to deal with shifts. That is: 1-the number of shifts a person works per day, 2-is the number of daytime shifts, 3-is the number of shifts per 24 hours, 5-is the number of shifts worked per week, 7- is the number of days in a week, 9-is the number of days off whenever someone books a weeks holiday, 11-is the number of night shift hours, 13-is the number of day shift hours, 17-is the number of days off whenever someone books 2 weeks holiday, 19 & 23 minimum & maximum days worked per month, 29-is the awkward leap year problem, 31-is the number of days in most months. The problem of prime numbers is that nothing will divide into them. Solutions to staffing on a week by week basis are far from ideal. One answer is to organise staff over a larger length of time, hence staff are required to work an annual number of hours, but spread unevenly over smaller periods of time. But, you do need a computer to keep track of staff over longer periods.
Annualised hours would be the contract hours per week summed over the year. i.e.. 40 hours per week equals 2080 hours per year, ignoring leap years and that there are 52 weeks and one day in a real year. Take off the annual holidays and any state (bank) holidays, i.e. 5 weeks (25 days, or 200 hours). The annual hours to be worked will be 1880 (2080-200) hours. Now, you have to change the hours into shifts, i.e. 1880= 235 8hour shifts, or 188 10hour shifts, etc. A part-timer working 30 hours per week would have figures proportional to those above, but with a significant difference. This is the time allowance for holidays. These are individual to each organisation.
Let's go back to the full time employee, often they would like to negotiate a reduction in the working week, say to 37.5 hours. The problem of trying to incorporate a reduction in the hours of a shift from 8 hours to 7.5 hours is huge, but reducing the annual number of shifts by the equivalent is simple, we arrive at 220 8hour shifts, a reduce of 15 shifts over the year.
How to use annualised hours. Set up Visual Rota for the year (12 files covering a month each), assume the same staff stay for the year. Ask the staff for the next years holiday requests, or use the previous years holidays as a template. Allocate shifts to the staff ignoring the holidays. Use an average for each month of 21 or 22 shifts for a person working 40 hours per week in 8 hour shifts. Check the staffing levels for cover, and if certain times are overstaffed, concentrate these in the peak holiday times. Insert the holiday requests by overwriting the shifts. Check staffing levels and then start to move staff shifts to give the desired cover. In periods of serious staff shortages, either refuse to allow the staff to go on holiday or use agency staff or recruit new staff on special contracts.
How To Start
First you need to forecast manning requirements. Use your demand forecast to determine how many staff you need at all times during the day, at what grade and location. Next. determine shift start and finish times and staffing per shift. Next, do the shifts overlap? This calculation will tell you how many man shifts you need to cover for any period, a day or up to a year ahead.
Next we need to find how many staff are required to fill those shifts. Initially, we ignore holidays, absenteeism, training days, etc. It would help if you converted the man-shifts total to hours. Lets say this came to 100,000 hours. Next, calculate the hours worked per year per staff member. 40 hours/week = 2080 hours.
At this point you will need a rule about the additional day each year and two in leap years. It is for each organisation to do this, but one method would be to increase the yearly hours by 8 whenever December 31st is a week day. to 2088 hours.
Next we divide the man-hours total(100,000) by hours worked per man year(2080), and this comes to 47.9, lets call it 50 staff.
Next we calculate the extra staff we need to cover holidays, absenteeism, etc. These can be separate calculations. Holidays, if each staff member is allowed 25 days of 8 hours annual leave, including bank(state) holidays, we have a total of 50 staff with 25x8 hours holidays, which equals 10000 hours, or 5 extra staff members just to cover holidays. Of course, these new 5 staff also require holidays, so we could go to 6 new staff and have about 1000 hours remainder to cover training, etc., about 2 days/staff.
We now have our shifts and we have the staff, so the next process is to cover the shifts. This process only takes a few hours using Visual Rota, but it would take quite a few days to do by hand. Using the computer to do all the counting, reduces the time for this task to a very small fraction of the time it takes manually. Allocating holidays, training days and staff requests is extremely easy. Absenteeism is a problem for all organisations and cannot be overlooked or ignored. All organisations have the option to be short staffed whenever someone is off sick, or to bring someone in to cover the shift. Using annual hours gives you the means to be staffed correctly and have additional staff to cover absenteeism, at the same time. Just as you can forecast demand by looking at past data, you can forecast absenteeism by looking at previous years. If you have 50 staff and they had an average of 5 days off absent, this equates to an extra member of staff for a the year.
The reason for annual hours is mainly to maximise some company element such as profit, or to equalise service, irrespective of holidays, etc. hence you will need to form a policy on holidays because it affects the next step. You can allocate holidays equally throughout the year and if there are no requests in the quiet periods, you will need to allocate the holidays irrespective of the staff wishes. The other method is to forecast the main demand periods for holidays and plan accordingly. For the former method, you can have a 'first come, first serviced' policy and once the peak holiday periods are fully booked, the rest have to be booked 'out of season'. This gives an even spread, and in this example, you will have 5 staff on holiday every day. This has a lot of advantages, but it can lead to bad relations between the staff. The latter method would initially require the peak holiday periods to be over staffed, by an amount equal to the holiday forecast. Then as requests are made, the staffing levels on the shifts falls. There is a maximum number of holiday requests possible, probably about 30% of the staff could be on holiday at any one time without affecting staff levels. However, the down side is that the staff at work have to work more shifts per week as a consequence during peak holiday periods. In our example, the 100,000 hours, based on 8 hour shifts, means that 35 staff out of our total of 56 staff have to come to work each day. Hence, if 20 staff book a weeks holiday, the rest of the staff have to work 7 days/week. The two methods can also be combined, such that some maximum number of staff is allowed on holiday, say 10, but if not enough staff want a holiday at that time, the holidays are allocated as well.
The next step is to start allocating shifts using Visual Rota, and this is covered elsewhere, and is as quick as you can type. We have done the maths, we know how many staff we want on each shift, and we know that WE CAN'T GO WRONG. I have emphasised that point because, if you stick to the plan, the numbers fall into place like magic. At the end of doing it, all the staff have the right number of shifts and all the shifts have the right number of staff.
We can offer advice about the following important items.
Staff leaving and joining and rejoining, transfers between depts.
Hols pro rata.
Mixing shifts of different length, 6 & 8 hrs
Rules about working over & under annual hours
Pay. hourly or not
2018 Copyright. All rights reserved.
please email us
Tel: (+44) 01636 816466