More on Retail Rostering

For what rostering is concerned, no other sector has been carefully studied as retail.

In the retail we can easily with great care all data that make a manager happy. They can finally use the formulas they learnt at the college.

And they had better doing so.

In the retail sector markups are usually thin.

The cost of personnel is high.

If they don’t control the cost of personnel, the company will quickly incur into a loss.

Retail Rostering: how the weekly roster is created

In the retail sector, rostering is usually done by week.

Using budget data we can forecast the weekly earnings and the sum we can devote to personnel costs.

A division by the average weekly cost of a single shop assistant will yield the maximum number of shop assistants we can use in the week, say N.

We cannot use a number greater then N, or the shop won’t be profitable.

If N is low, this could mean offering a bad service to customers.

A workforce management application will alert the managers and they will decide if they want to suffer a loss.

Sometimes managers decide to lose money for a period, because they know they will recover the loss at a later time.

For what concerns us, we have a number N of employees we can use in the week.

Our skill as planners is to make the best use of the N employees, placing them in the days and times when we mostly need them.

Retail Rostering: spreading employees horizontally

Depending on work contract, each employee has 1 or 2 days off in a week.

First step will be to allocate days off for each employee.

We need to allocate days off so that:

  • we have a higher number of employees in days of greater customers inflow
  • in the long run all employees rest the same number of (possible) week days

Care must be taken so that we get full force on – for instance – weekends, when the number of visitors is the hightest. In that case no

Also we need to equally share the workload between employees so that all share the same number of (what are regarded) good and bad shifts.

Needless to say, a workforce management application will do that automatically. We input the average customer inflow by weekday. We input the type of contract for each employee. The application does the rest: it will optimize the assignment of days off and the number of available employees for each week day.

Retail Rostering: spreading employees vertically

We completed step 1 and for each day of the week we have the names of employees to place in the roster.

The next step is to have more employees in the parts of the day where the customers inflow is greater.

In the meantime we need to

  • saturate (but not exceed) the weekly work hours of each employee
  • use only predefined shifts
  • equally assign early hours and late hours to all employees

This is an optimization problem that John Nash will solve in a few minutes… but if we are not John Nash … we are out of luck.

As expected, a workforce management application will do that in a time so short (a few seconds) that John Nash would blush.

When shops move hundreds of assistants and a very detailled analysis is required (say time slots up to 15’ short) a yellow cloud of smoke could rise from our PC, meaning that the problem is too hard to solve.

Nowadays good applications use powerful solvers, so a problem must be really complex to be unfeasable.

Retail Rostering: tuning up the rostering

The final touches to an optimized automated rostering are done by hand.

The world still needs us. Computers are stealing our jobs no time soon. When you see the automated roster, it has things than can be improved: shorten the pause of John, shift the entrance time of Paul, etc.

A workforce management application will do whatever we don’t specifically forbid. By so doing, results can be a bit weird sometimes.

While we apply the final touches, the application will show us possible drifts from optimal results. Sometimes a more reasonable roster is much better than an absolutely optimal roster: employees satisfaction counts more than a minimal gain in optimization.

Contact us
Démonstration Gratuite
Call Me Now!
Me téléphoner! entrez votre numéro de téléphone mobile
Buy Tickets Now!
Tickets