Articles, perspectives, and analysis on capacity planning, resource management, workforce transitions, and the operational questions that matter.
As businesses accelerate AI adoption, the operational questions are changing. The question is no longer whether AI will replace headcount — it is how to model the transition accurately before committing to it. Token consumption, agent throughput, and handling time equivalence require the same analytical rigour as any workforce change.
A capacity plan that has never been tested against historical performance is a hypothesis, not an operational tool. Backtesting is the step that separates a working model from a well-formatted spreadsheet — and the step most planning functions skip.
Finance teams can attach costs to resource numbers. What they cannot do without the right inputs is build a credible operational business case. The planning function is the source of those inputs — and the quality of the plan determines the quality of every financial decision that follows.
When a planning function loses its key resource, the instinct is to find cover. The more important question is whether the model, the controls, and the governance framework can survive the transition — and for how long. Cover fills a desk. Continuity protects the cycle.
Resource decisions, SLA commitments, and headcount plans all flow from the forecast. If the forecast is systematically biased — high or low — every decision downstream is built on the same error. Forecasting accuracy is not a reporting metric. It is an operational control.
Most capacity models start with assumptions about how long things take. The ones that hold up under pressure start with measurement. True handling time — including wrap, variance by complexity, and channel — is the foundation on which every reliable resource number is built.
Tell us about your operation and your planning requirements. We will listen, ask the right questions, and recommend the most appropriate structure.