Capacity Planning

What capacity planning covers

Short-term workforce management asks who sits where next Tuesday at 10:30. Capacity planning looks 3 to 18 months ahead and asks bigger questions: how many agents will we need in each month, how many must we hire and when, how many seats and licenses do we need, and what will it cost.

It typically produces a weekly or monthly plan by line of business with required FTE, available FTE, the gap, the hiring classes to close it, and the overtime or outsourcing needed when hiring cannot close it in time. Finance uses it for the budget. Recruiting uses it for requisitions. Facilities uses it for seat planning.

Inputs to a staffing model

A credible model needs eight inputs, each reviewed against actuals every month:

• Volume forecast by month, week and channel, from call volume forecasting.
• Average handle time by contact type, including expected changes from new products, policy changes or new hire mix.
• Service level target and the resulting occupancy, usually from Erlang C or the WFM tool.
• Shrinkage, both planned and unplanned, by month, since vacation and sickness are seasonal.
• Attrition for tenured staff and separately for agents in their first 90 days.
• Training length and nesting curve: weeks in classroom, weeks in nesting, and the productivity ramp after that.
• Training attrition, the share of a class that does not make it to the floor.
• Recruiting lead time: how many weeks from requisition open to start date for a full class.

The last three are where most plans go wrong, and they are where recruiting and L&D can give planners the best data.

Turning the plan into hiring classes

Work backward from the date you need productive agents. If training is 4 weeks, nesting 3 weeks and the ramp to full AHT another 4 weeks, a class needs to start about 11 weeks before the demand arrives. Add recruiting lead time, often 3 to 5 weeks for a class of 20 to 30 entry-level agents, and the requisition must open 14 to 16 weeks before the peak.

That is why capacity planning connects directly to headcount planning and to how you run each hiring wave. Most centers hire in fixed class sizes set by trainer capacity, say 15 to 25 per class, and stagger classes so the training team is not swamped. When the gap is larger than training can absorb, the plan has to show the shortfall honestly and cover it with overtime, an outsourcing partner or a lower service level for a period.

Seat planning

Seat planning checks whether the building, or the work-at-home equipment pool, can hold the plan. Required seats depend on the peak number of agents logged in at the same time, not on total headcount. Shift patterns, part-time staff and work-at-home agents all bring the ratio down. A center running two overlapping shifts might need 0.6 to 0.7 seats per agent; a single-shift operation needs close to 1.0 plus room for training classes.

Do not forget training rooms. A 25-person class occupies 25 seats for the full training period, plus nesting pods on the floor.

Common capacity planning mistakes

• Using last year's attrition. If wages in your labor market moved, attrition moved too.
• Treating new hires as fully productive. An agent in month two might handle 70% of a tenured agent's volume. Model the curve.
• Ignoring training attrition. Losing 25% of a class before the floor is normal in some operations. A plan that assumes zero will always be short.
• One scenario. Build a base, a high and a low case. Leadership decisions are easier when they can see the cost of being wrong in each direction.
• Planning in isolation from recruiting. If recruiting cannot deliver 40 starts in three weeks, the plan should not assume it.

Best practice for planners and recruiters

Run a monthly capacity review with operations, WFM, finance, recruiting and training in the same meeting. Review forecast accuracy, the attrition and training assumptions, the hiring classes already in flight, and the next 13 weeks of start dates.

Keep a living hiring calendar that recruiting can see: class start dates, sizes, language or skill requirements and shift mix. Recruiters then build pipelines ahead of demand instead of reacting to urgent requisitions. In our experience, the centers with the steadiest service levels are not the ones with the cleverest forecast models. They are the ones where the capacity plan and the recruiting calendar are the same document.