Template · Contact centre operations
Call center capacity planning template
In short
A call center capacity planning template turns a volume forecast into the number of agents you need: forecast contacts and handle time, use Erlang C to find agents required per interval for your service level, add shrinkage, convert to FTE, then compare with current headcount and size the hiring classes needed to close the gap.
Contact Center Capacity Plan template
Contact Center Capacity Plan
Complete one plan per queue or skill group for each planning period (usually monthly, looking 3 to 6 months ahead). Use the busiest representative interval for the Erlang C check, and weekly totals for the FTE calculation. Record assumptions so the plan can be checked against actual results.
1. Plan details
- Queue / skill group / line of business
- Site(s)
- Planning period (month or weeks covered)
- Prepared by
- Date and version
- Hours of operation (days and times)
2. Demand forecast
- Forecast contacts per week
- Basis of forecast (history period, growth, seasonality, marketing or billing events)
- Peak interval: day and time (e.g. Monday 10:00 to 10:30)
- Forecast contacts in peak interval
- Interval length (minutes)
- 153060
- Average handle time, seconds (talk + hold + after-call work)
- Channel mix if blended (voice / chat / email %)
3. Service targets
- Service level target (e.g. 80% answered in 20 seconds)
- Average speed of answer target (seconds)
- Maximum planned occupancy (%)
- Abandon rate target (%)
4. Agents required in the peak interval (Erlang C)
- Traffic intensity in Erlangs = contacts x AHT (seconds) / interval length (seconds)
- Agents required on phones to meet service level (from Erlang C calculator)
- Resulting service level (%)
- Resulting average speed of answer (seconds)
- Resulting occupancy (%) = Erlangs / agents
5. Shrinkage
- Planned PTO and holidays (%)
- Unplanned absence and tardiness (%)
- Breaks (paid) (%)
- Training, coaching and meetings (%)
- System downtime and other (%)
- Total shrinkage (%)
- Agents to schedule in peak interval = agents required / (1 - shrinkage)
6. FTE requirement for the week
- Workload hours = weekly contacts x AHT (seconds) / 3,600
- Staffed hours needed = workload hours / planned occupancy
- Paid hours needed = staffed hours / (1 - shrinkage)
- Paid hours per FTE per week
- FTE required = paid hours needed / paid hours per FTE
- Schedule efficiency adjustment (%) and adjusted FTE required
7. Supply and gap
- Current production FTE (excluding agents in training and nesting)
- Agents in training / nesting and expected graduation dates
- Expected monthly attrition (% and agents)
- Planned transfers, promotions or leaves
- Projected production FTE at start of planning period
- FTE gap (required - projected); negative = surplus
8. Hiring and action plan
- Training + nesting length (weeks)
- Expected loss during training and nesting (%)
- Graduates needed = (FTE gap + attrition during lead time) / FTE per agent
- Class size to start = graduates needed / (1 - training and nesting loss)
- Offers to make (allow for no-shows and declines)
- Class start date(s) and required offer-acceptance date
- Other levers if hiring cannot close the gap (overtime, cross-skilling, outsourcing, self-service, schedule changes)
9. Assumptions, risks and approval
- Key assumptions and risks
- Last period: forecast vs actual volume, AHT and shrinkage
- Approved by (operations)
- Approved by (finance / HR)
- Date
How to use this template
- Forecast weekly contacts and the peak interval from at least 12 months of history if you have it, adjusted for growth, seasonality and known events like billing cycles or product launches.
- Measure average handle time as talk plus hold plus after-call work, from the same period and queue, not a company-wide average.
- Calculate traffic in Erlangs for the peak interval and use an Erlang C calculator to find the number of agents that meets your service level without pushing occupancy above your cap.
- Build shrinkage from its parts using your own last 3 to 6 months of data, then divide required agents by one minus shrinkage.
- Convert weekly workload into paid hours and FTE, and adjust for schedule efficiency, since shifts never match demand curves exactly.
- Compare with projected production headcount after attrition, then size hiring classes using your real training and nesting loss rates and lead times.
- Review the plan monthly against actual volume, handle time and shrinkage, and feed the error back into the next forecast.
How the calculation works
Capacity planning runs in four steps: demand, agents on phones, agents scheduled, and FTE.
- Demand. Workload is contacts multiplied by average handle time (AHT). 200 calls in a 30-minute interval at 300 seconds AHT is 60,000 seconds of work in 1,800 seconds, or 33.3 Erlangs. That means 33.3 agents would be busy all the time with zero waiting.
- Agents on phones. Calls do not arrive evenly, so you need more agents than Erlangs to answer within your target. Erlang C estimates the probability a call waits and the service level for a given number of agents.
- Shrinkage. Agents paid but not available to take calls: breaks, PTO, absence, training, coaching and meetings. Contact centers commonly plan for 30 to 35 percent total shrinkage, but use your own measured figure.
- FTE. Convert the weekly hours into full-time equivalents and compare with the people you have.
Worked example
A health insurer's member services queue forecasts 18,000 calls a week with a 300-second AHT, a peak of 200 calls in the Monday 10:00 to 10:30 interval, and a service level target of 80 percent in 20 seconds.
Peak interval (Erlang C). Traffic is 33.3 Erlangs. With 38 agents, service level is about 76 percent, short of target. With 39 agents it is about 83 percent, average speed of answer is around 13 seconds and occupancy is 85 percent. So 39 agents must be on the phones. With 30 percent shrinkage: 39 / 0.70 = 55.7, so schedule 56 agents for that interval.
Weekly FTE. Workload: 18,000 x 300 / 3,600 = 1,500 hours. At 85 percent planned occupancy, staffed hours are 1,500 / 0.85 = 1,765. With 30 percent shrinkage, paid hours are 1,765 / 0.70 = 2,521. At 40 paid hours per FTE that is 63 FTE. Schedules rarely fit the demand curve perfectly, so many planners add 5 to 10 percent for schedule efficiency, bringing it to about 66 to 69 FTE.
Gap and hiring class. The queue has 55 production FTE today. Attrition runs about 4 percent a month (around 2 agents), and training plus nesting takes 5 weeks. Required 66 minus 55 is a gap of 11, plus about 3 agents lost during the lead time: 14 graduates needed. With 20 percent loss through training and nesting, start a class of 14 / 0.8 = 17.5, so 18. If about 15 percent of accepted offers do not start, make 21 offers. That number goes straight into the hiring class plan.
Erlang C limitations
Erlang C is the standard starting point, and every workforce management tool and spreadsheet Erlang C calculator uses some version of it. Know where it is wrong:
- It assumes no caller abandons, so it overstates the agents needed when abandon rates are meaningful. Erlang A models abandonment and gives lower numbers.
- It assumes calls arrive randomly and agents are interchangeable. Multi-skill routing and callbacks break that.
- It does not handle chat concurrency or email backlogs. For email, plan on workload hours and response-time targets instead.
- Small teams suffer most: a 6-agent queue at 80/20 needs far more slack per Erlang than a 100-agent queue. Merging small queues often saves more than any hiring plan.
Run Erlang C per interval across the day to build the requirement curve, then use a shift schedule template or scheduling tool to fit shifts to that curve.
Connecting capacity planning to hiring
The capacity plan is the input for recruiting, and the lead time matters more than most finance teams expect. A hiring class that starts training on the 1st needs offers accepted two to three weeks before, which needs sourcing to start four to six weeks before that. If the plan shows a gap in May, recruiting starts in February.
Feed these numbers into a workforce hiring plan across all queues and sites, and track actual training loss, nesting loss and early attrition by class so the next plan uses real rates. For roles such as a contact center specialist, be honest about attrition: planning with 2 percent monthly when the floor runs at 5 percent produces a permanent gap.
Common mistakes
- Using daily averages instead of interval peaks, which leaves Monday mornings badly short.
- AHT that leaves out after-call work.
- One shrinkage number copied from an industry article rather than measured.
- Planning occupancy at 95 percent. It looks efficient and burns agents out; 80 to 88 percent is a common range for voice.
- Counting agents in training and nesting as full production FTE.
- Hiring to fill today's gap, forgetting the attrition that will happen while the class is in training.
- Never checking forecast accuracy, so the same error repeats every month.
Frequently asked questions
How do you calculate call center staffing?
Multiply forecast contacts by average handle time to get workload, use Erlang C to find how many agents meet your service level in each interval, divide by one minus shrinkage to get agents scheduled, then convert weekly paid hours into FTE and compare with current headcount.
Is there an Erlang C calculator in Excel?
Yes. Erlang C can be built in Excel with the formula for the probability of waiting, or with an add-in. Many teams use a simple sheet that takes calls, AHT, interval length and service target and returns agents required, service level, speed of answer and occupancy.
What is a good shrinkage percentage for a call center?
Many contact centers plan for about 30 to 35 percent total shrinkage, including breaks, PTO, absence, training and coaching. The right number is the one you measure in your own operation over the last several months.
What is the difference between a call center staffing calculator and a workforce management template?
A staffing calculator answers how many agents you need for a given interval. A workforce management or capacity planning template goes further: weekly FTE, shrinkage, current headcount, attrition and the hiring needed to close the gap over the coming months.
How far ahead should call center forecasting be done?
Most teams forecast at three horizons: long-term (6 to 18 months) for budgets and hiring, medium-term (1 to 3 months) for hiring classes and PTO, and short-term (weeks and days) for schedules and intraday changes.
What occupancy rate should a call center plan for?
For voice, 80 to 88 percent is a common planning range. Sustained occupancy above about 90 percent usually shows up later as higher absence and attrition.