Breakdown of the problem into smaller units is a great way to start. Segregation into locations (region/country), business units (education, healthcare, entertainment etc.), segments (new business, existing business), contract types (cost plus, P&L) is one way of narrowing the problem. For e.g.: higher costs in P&L accounts are a bigger worry as the risk associated with any cost increase or labor issue lies with the contractor. Similarly, new businesses carry a higher cost ratio initially, as the revenue stream is not yet stabilized.
Though labor cost is a number in itself and performance is a qualitative measure, various matrices can help understand the extent of the success or failure of the cost order.
Several KPIs can be used to assess this:
- Labor as a % of revenue
- Revenue per labor hour and labor cost per hour
- Productivity variance
- Premium labor as a % of total labor cost
- Voluntary attrition rate
- Non-productive hours as a % of total hours
The variances here, when benchmarked against internal/external indicators can indicate:
- Fundamental issues with labor forecasting
- Compliance issues where actual utilization exceed planned labor hours
- Inability to adjust labor hours in conjunction with fluctuating demand
- Inappropriate labor mix of salaried and contract workers
- High usage of premium labor3
Combination of technology and human intervention
Since wages account for nearly 70% of the total labor cost, the first step towards reducing labor cost is to get the forecasting right. Demand forecast in a catering business needs to show the shape of the day and the anticipated demand. Demand is forecast typically in blocks of hours (9am-12pm, 12pm-3pm, 3pm-6pm etc). That way, managers can see exactly how many employees they need in each area to meet that demand. It also takes into account exactly how much time employees need to deliver each activity. This even includes non-revenue-generating (but necessary) activities, like preparation and clean up.
This type of forecasting is a typical use case for supervised machine learning regression, which can make accurate predictions utilizing many factors like seasonality, changing trends, local events, holiday periods, time of the day etc. It can create schedules that cater to the unique demand of each location, helping to avoid over or under-staffing across the organization, both of which have significant cost implications.
Right employee mix
Finding the appropriate mix of skill sets (Manager, chef, cashier etc) and employee type (full time/part time) can be a challenge. The consistency provided by the permanent staff heavily outweighs the expense due to additional benefits like health care, pension etc. Artificial Intelligence (AI) solutions that work toward optimization3 can be explored. Closed loop intelligence optimization4 is an iterative method where results are compared and fed back to the system to find the optimum mix. The right employee mix is the optimum combination of workers at the lowest possible cost with maximum productive efficiency.
The attrition rate in the contract catering industry is high since they hire the highest percentage of students to keep costs low. This employee base is transient and easily switch to other opportunities for a marginal increment in wages. The key to retention here is flexibility and training opportunities to upskill.
Using AI solution built on preference learning can recommend potential shifts to employees based on their previous scheduling preferences. This self-scheduling can provide fair, equitable, balanced schedule that meets their requirements as well as the company's business requirements.
While AI/ML can definitely contribute towards employee retention, some additional humane measures can go a long way. These include:
- Ensuring employee wages are comparable to market wages
- Providing training and opportunities to enable employees to upskill themselves
- Modifying shifts to accommodate employee needs and constraints
- Conducting activities to improve employee morale
Triggers can be set up based on continuous learning that can push out notifications where:
- Compliance issues are noted for a location on a regular basis
- Falling demand is not matched by reduced labor hours
- Breach of KPIs can be pre-empted
Timely information as well as using AI/ML learning can pave the way for intervention to remediate the issues that can have significant bearings on the cost element.
Effective use of AI in key areas can be a significant differentiating factor in the catering industry. Its use can range from forecasting demand, planning promotional activities, reducing employee attrition to getting the optimum labor mix. These in turn can reduce wastages in both labor and food, thereby increasing productivity and margins. However, AI can only be successful when used in conjunction with humane initiatives to improve employee morale and satisfaction. This winning combination can be the definite success mantra for businesses.