Both established corporations and high growth startups need to optimize customer support to reach their customer service goals at reasonable costs. 

The calculation of the required number of agents is not as simple as taking the number of calls and multiplying with the average handle time and then dividing it by the unit of time. First and foremost, it requires demand planning and typically the historical data is the best source of information for key statistics needed. Of course, the past can not predict the future, but it is a good starting point to kick off discussions with various stakeholders about budgeting and planning as well as to experiment with different scenarios. In this planning phase the following key information should be considered:

• Expected number of incoming interactions (expected sales is a good proxy for this)
• Planned marketing campaigns
• Seasonality
• Unexpected complaints due to new service/product launch
• Need for scalable workforce
• Expansion strategy to new markets
• Experience of the agent team

The goal of this planning phase is to have a good understanding of the number of calls, language skills and service level required. Service level is defined as the percentage of calls answered within a predetermined number of seconds. The standard service level is the 80/20, meaning that 80% of the calls are answered in 20 seconds. Business needs, competition, and customer expectations may also suggest other service levels, such as 80/30 or 70/40. Service level expectations change by type of channel too. For example, the industry standard for email support is to answer 100% of emails in 24 hours and for live chat support to reply 80% of chats in 20 seconds.

All of these complexities can be managed by using a rather scientific approach, the call center staffing calculation, which uses the Erlang C formula. This is an elegant mathematical depiction that requires inputs, such as the number of calls, average handling time, timeframe and calculates the number of agents needed, expected service level, and average speed of answer. There are a couple of underlying assumptions, such as that the formula assumes that the call arrivals can be modeled by a Poisson process and the agent skill set is homogeneous, however, it is still the best tool in the hand of customer support managers to plan ahead with the required resources.

As SLA is directly related to budgeting, thus the next step you may consider is to run the numbers and evaluate various staffing scenarios. Below, you can find the Agents Republic calculator free of charge to get useful information that can be easily saved for future reference or email to your colleagues. If you are planning to outsource partly or fully customer support you can get a quick cost estimation with a click of a button.

Fill in the blanks to calculate the number of agents needed and get a cost estimate.

Use Erlang calculator to calculate the number of agents and get cost estimate

SLA is an easy to understand and measure metric, but it does not tell the whole story. To get a complete view of your customer support efforts you may also consider tracking customer satisfaction, net promoter score (NPS), first contact resolution (FCR), abandon rates, quality scores, etc. The best set of metrics is defined by industry best practices, company business goals and most importantly company culture.

Get started with outsourcing customer support today! Ask for a quote!

Share