Erlang still dominates contact center staffing. Abandonment and retrials are where many models break down.
- 1 minute ago
- 5 min read
By Mark Alpern, COO, Cinareo.
This article reflects patterns we see repeatedly when reviewing real-world staffing models across contact center environments. In many of these environments, the largest source of staffing error is not forecast accuracy, but how abandonment and retries are treated, or ignored, in the underlying model.
Despite decades of technological change in contact center operations, Erlang-based staffing models remain the de facto standard for determining frontline capacity requirements. This persistence is not accidental, nor is it simply historical inertia. Erlang continues to endure because it uniquely balances operational rigour, financial discipline, and decision defensibility.
At a tactical level, Erlang provides a mathematically sound way to translate forecasted demand and average handle time into staffing requirements. At a strategic level, it enables executives to govern one of the largest controllable cost categories in the enterprise - its human resources - with transparency, consistency, and confidence.
For modern contact center leaders, Erlang is not just a queueing formula. It is a control mechanism for operational risk and financial performance.
The Core Problem Erlang Solves
Every contact center, regardless of industry or maturity, must answer the same fundamental question: Given expected demand and service commitments, how much capacity is required, and what are the consequences of being wrong?
Erlang models address this problem by linking Arrival rates (volume), Work effort (average handle time) and Performance objectives (service level, speed of answer)
to required staffing levels. This linkage is critical. Without it, staffing decisions are driven by intuition, historical averages, or budget constraints, all of which expose the organization to service failures, cost overruns, or both.
Why Erlang Endures
There are three primary reasons Erlang continues to dominate contact center staffing.
Proven Reliability and Transparency
Erlang has been stress-tested for over a century across telecommunications and contact centers. Its assumptions, limitations, and behaviors are well understood and widely documented.
From an executive perspective, this matters because Erlang-based staffing plans are explainable to non-technical stakeholders, auditable by finance and governance teams, and defensible in board, regulatory, and budget discussions. When staffing plans are challenged, Erlang provides a clear line of sight from assumptions to outcomes.
Speed and Scalability
Erlang calculations are computationally lightweight, allowing organizations to run thousands of interval-level calculations quickly. This enables intraday and short-term staffing adjustments, medium-term and long-range capacity planning and budgeting, and rapid scenario modeling without excessive overhead. Speed matters not just operationally, but strategically. It allows leaders to evaluate alternatives quickly and respond to uncertainty without reverting to guesswork.
Industry Standardization
Most workforce management platforms, spreadsheets, benchmarks, and historical comparisons are built around Erlang. This creates a shared language across Workforce planners, Operations leaders, Technology vendors, Finance and executive stakeholders. Standardization reduces friction in decision-making and ensures staffing discussions are grounded in a common analytical framework.
Understanding Erlang C, A, and X
While Erlang is often discussed as a single model, its variants serve different operational realities. Choosing the right one is a tactical decision, but the implications are strategic.
Erlang C: The Industry Workhorse
Erlang C is the most widely used model in contact centers. It assumes that customers will wait indefinitely (no abandonment), first-come, first-served queues, and has a fixed number of agents. In environments where abandonment is material, treating abandoned contacts as lost demand creates a systematic distortion in staffing and cost assumptions.
Because it ignores abandonment, Erlang C tends to overestimate staffing requirements in environments where customers abandon while waiting. Despite this limitation, it remains popular because it is: Conservative by design, easy to explain and defend, closely aligned with service-level-driven cultures
From a financial perspective, Erlang C often functions as a risk-averse baseline, prioritizing service protection over cost efficiency.
Erlang A: Introducing Abandonment
Erlang A extends Erlang C by modeling customer patience and abandonment behavior. It is better suited for high-volume environments, digital channels with short tolerance thresholds, and operations where abandonment materially affects outcomes. Erlang A typically produces lower staffing requirements than Erlang C.
The trade-off is increased complexity, as it relies on accurate estimates of customer patience or tolerance to abandon, a capability many organizations struggle to maintain consistently. When customer patience assumptions are unstable or poorly estimated, Erlang A can underestimate the number of agents required to achieve a service standard, particularly in volatile or highly variable environments.
Erlang X: Accounting for Abandonment and Retrials
Erlang X further extends the model by accounting for customers who abandon and retry later. This is particularly relevant in industries such as government services, healthcare access lines, utilities and public services. In these environments, abandonment does not represent lost demand, but deferred demand that returns and distorts future workload. Really, any organization that has a good understanding of patience and redials can benefit from Erlang X.
In these environments, Erlang X is not an enhancement to Erlang C or Erlang A, but a correction for demand that does not disappear.
While Erlang X is conceptually the most accurate, its adoption has been limited due to difficulty in modeling retrial behaviour and by limited support in commercial WFM platforms. When applied correctly, Erlang X provides the clearest view of true demand and staffing requirements in high-retry environments.
The Strategic Importance of Erlang for Executives
Financial Governance and Labor Cost Control
Labor represents 60–80% of total contact center operating costs. Erlang provides a direct, traceable connection between demand assumptions and labor investment.
For finance and executive leaders, this enables validation of staffing budgets against demand-driven logic, quantification of the cost of improving or degrading service levels, and clear articulation of the financial risks of under- or over-staffing. Without an Erlang-based foundation, staffing decisions lack financial rigor and undermine cost governance.
Decision Defensibility Under Uncertainty
Forecasts are never perfect. Executives are not deciding whether plans are “accurate,” but whether they are reasonable, explainable, and risk-aware.
Erlang provides a defensible baseline when combined with what-if scenario evaluation, a consistent framework for comparing strategic options, and protection against reactive, anecdote-driven decision-making. In executive and board discussions, Erlang shifts the conversation from opinion to evidence.
Foundation for Strategic Scenario Modeling
Erlang is not the ceiling but a starting point. Modern capacity planning platforms, such as Cinareo, use Erlang as the analytical backbone and layer on multi-skill optimization, channel-mix and concurrency modeling, multi-geo, site, and BPO allocation decisions, and cost, CX, and EX trade-off analysis. This transforms Erlang from a staffing calculator into a strategic decision-support engine.
Final thought
Erlang remains dominant not because it is perfect, but because it is governable, transparent, and defensible. For operations leaders, it ensures staffing plans are operationally viable. For finance leaders, it ensures labor investments are economically justified. For executives and boards, it provides confidence that capacity decisions are disciplined, auditable, and aligned with enterprise objectives.
Pressure-test your assumptions
If abandonment and retrials are material in your environment, the next step is to see how they change your capacity requirements. We built an Erlang X calculator that includes Patience and Retry factors to help teams model deferred demand and understand how retries affect staffing, service levels, and cost.
Have a complex environment or want to sanity-check your assumptions? Talk to our team.
