By James Proctor, Co-Founder and Managing Director, The Inteq Group
You handle exception volume spikes without adding staff by designing the process so that the routine majority of any surge is absorbed by handling capacity that scales with volume, while human capacity is reserved for the judgment residual, the fraction of the surge that genuinely requires a person, which grows far more slowly than raw volume does. The approach works because of a fact about surges that almost no operation has ever measured: surges are overwhelmingly composed of ordinary cases arriving in extraordinary quantity. Solve for that composition and the historical weld between volume volatility and operating cost breaks.
What Is a Surge Actually Made Of?
Decompose any spike, a seasonal peak, a post-incident flood, a regulatory deadline pile-up, and the anatomy repeats. The bulk of the added volume is routine work: cases that match known patterns and clear policy, distinguished from a normal Tuesday only by their arrival rate. A thin layer is genuinely novel or ambiguous, and a thinner layer still is consequential enough to demand deliberate judgment. Traditional operations experience all three layers as one undifferentiated queue, because nothing in the design can tell them apart, so the organization staffs to the total. The design insight is almost embarrassingly simple: the layers are separable, case by case, by explicit criteria, and only the top layers ever needed people.
What Does the Design Require?
Three elements. Explicit criteria that separate routine from judgment at intake, per case, so the routine layer flows at machine speed regardless of arrival rate. Context assembly for everything that surfaces to a person, so the scarce human hours land on judging rather than gathering, precisely when those hours are scarcest. And composition monitoring, because a surge whose judgment residual starts growing is telling you something new is happening, and that signal, invisible in an undifferentiated queue, is exactly what management needs during a disruption. None of this eliminates people. It concentrates them, which is why service quality during spikes goes up under this design, not down.
“Volume is not the enemy. Undesigned volume is.”
What About the Ultimate Stress Test?
Tax and accounting firms live the purest version of this problem: the business model compresses a year of client work into a brutal quarter, and the industry's standing answer is seasonal hiring, overtime, and burnout, institutionalized to the point of professional folklore. I will state the uncomfortable reading plainly: seasonal hiring at that scale is an admission that the process design failed, converted into an annual ritual with its own budget line. Decompose the season and the anatomy holds. The overwhelming share of surge work, standard returns, document intake, routine notices, extension processing, is pattern work arriving at once, while the judgment residual, complex structures, unusual transactions, gray-area positions, is a modest fraction that the firm's experienced professionals should own completely. Firms redesigning intake, preparation, and review around that separation are discovering the season still peaks, but the peak is absorbed by design: smaller seasonal ratios, senior staff spending the season on judgment instead of triage, and the quiet retirement of the ritual in which the firm trains temporary staff each January to perform work the process itself should have absorbed.
The economics follow the anatomy. Capacity planning stops being an exercise in staffing for the worst month, with the carry cost of the peak baked into every other month or paid in service failures when the guess runs low. The fixed human establishment sizes to the judgment residual, which is small and remarkably stable even when raw volume swings violently, and volume itself becomes a throughput fact rather than a hiring trigger. That is the decoupling in practice, and it shows up on the income statement, not just the process map.
Learning to run surge-composition analysis and design the intake criteria that separate the layers is a practical skill built in our agentic AI training courses, and it transfers to every volume-volatile process an analyst will ever touch.
