Define the retention problem before choosing the solution

A resignation, retirement, internal promotion and completed temporary contract all change the workforce. They do not represent the same management problem. Combining them into one number can make a useful development opportunity look like a failure or conceal repeated preventable exits from an important role.

Arcus recommends separating the measurement question from the intervention question. First establish what happened using consistent definitions. Then identify the likely mechanisms, assess service consequences and select a response that leadership can implement and sustain.

The aim is not to keep every employee indefinitely. It is to reduce avoidable loss of capability, support a healthy employment relationship and maintain continuity when people do leave. Appropriate retirement, progression and managed transitions remain part of a functioning organization.

For a small administration, the departure of one experienced employee may matter more than a large percentage change suggests. Leadership needs both the overall pattern and the operational story behind critical vacancies.

Build a baseline that can be interpreted

Define the period, workforce included, departure categories and denominator before calculating a rate. A simple annual separation rate can be calculated as departures during the year divided by average headcount during the same year, multiplied by 100. Monthly average headcount is preferable when staffing changes materially during the year; using only the start and end counts may conceal fluctuations.

Keep voluntary resignations, retirements, involuntary separations, internal moves and expected contract completions distinguishable in the underlying record. Do not silently change categories between reporting periods. Record service area and tenure where that information can be used appropriately and confidentially.

For illustration, six separations in an organization averaging 40 employees produce a 15% annual separation rate. If three were expected contract completions and three were voluntary resignations, the voluntary-resignation rate is 7.5%. Neither figure establishes whether management could have prevented the resignations.

Track retention of a defined starting cohort separately. If 36 of 40 employees present at the beginning remain at year-end, cohort retention is 90%, even if new hires increased the overall workforce. This answers a different question from turnover. Present counts alongside percentages, especially in small teams.

Use a longer view where individual events dominate annual figures. A rolling period can help identify patterns, but it should not delay action on an immediate workplace concern.

Test competing explanations with more than exit interviews

Exit interviews are useful but incomplete. Some former employees decline to participate, give several reasons or avoid criticism. Combine their feedback with confidential conversations with current staff, recruitment outcomes, workload records, vacancy duration and evidence about progression opportunities.

Where early-tenure exits cluster, examine the accuracy of recruitment promises, onboarding, supervision and readiness of equipment or systems. Where experienced staff leave, consider role growth, recognition, career pathways, workload and the competitiveness of the employment offer. These are hypotheses to investigate, not conclusions that follow automatically from tenure.

If departures cluster in one team, explore both local management practices and the conditions facing that team. A manager may need development, but the team may also have unusual service demand, unstable funding or hard-to-fill specialist roles. An organization-wide pay increase would be a costly response to a problem confined to one operating process.

Review whether fixed-term employment is being used for work that continues year after year. Greater employment certainty may be relevant, but it requires a credible funding and operating decision. Do not promise permanent positions before the Nation can sustain them.

Protect the people providing feedback. Small numbers can make supposedly anonymous comments identifiable. Explain who will see information, how themes will be reported and what confidentiality limits apply. Respond to specific serious concerns through the appropriate process rather than waiting for a statistical pattern.

Choose an intervention that addresses the mechanism

When the evidence points to workload, redesign priorities, coverage and handoffs. Adding a wellness activity while leaving an impossible service burden unchanged is unlikely to resolve the operating constraint. Determine which work can stop, be simplified or receive additional support.

When progression is the issue, create credible development routes: supervised assignments, training linked to real responsibilities, mentoring and transparent access to internal opportunities. In a small organization, development may involve broader capability or project leadership rather than a promotion into a vacancy that does not exist.

When compensation is the issue, use relevant job comparisons and assess the wider employment package. A targeted adjustment may be justified, but consider internal relationships and recurring affordability. A temporary bonus may delay an exit without changing the underlying employment proposition.

When management relationships are the concern, clarify expectations, provide appropriate coaching and address specific conduct or performance issues fairly. General training alone may be insufficient where accountabilities are unclear or agreed actions are not followed through.

Choose a small number of changes with named owners and observable results. An administration that launches many disconnected retention initiatives can consume the same management capacity it is trying to protect.

Estimate costs without manufacturing a return on investment

Separate recruitment expenses, temporary coverage, onboarding and staff time. Use actual invoices and recorded hours where possible. Include the service consequences of losing knowledge or relationships, but do not convert them into precise monetary savings without a defensible method.

Consider a hypothetical role with $7,000 in recruitment costs, $9,000 in incremental temporary coverage and $4,000 in external onboarding or training costs per replacement. That produces $20,000 in identified cash expenditure. An additional 120 hours of existing staff time should be shown separately unless it creates extra paid hours or requires replacement capacity.

Suppose a proposed development-and-supervision intervention costs $18,000 annually. Avoiding one replacement could offset that amount against the simplified $20,000 cash estimate. This is a scenario, not evidence that the intervention will prevent an exit. If no departure is prevented, the organization still incurs the intervention cost and may receive other benefits that need separate assessment.

Test the affordability of continuing the change. A pilot funded this year may create an expectation of ongoing development support. Include implementation time, backfill and management attention in the decision, and identify what will be reduced or postponed to make room.

Compare retention investment with continuity investment. Some departures will remain unavoidable; cross-training and documented handovers can reduce disruption even when they do not lower turnover.

Implement, review and communicate what changed

Begin with a short baseline review and a small set of confidential listening activities. Share the main themes with employees in a form that protects individuals. State which actions leadership will take, who owns them and when staff can expect an update.

Review implementation before claiming results. If supervisors were expected to hold development discussions, verify that the discussions occurred and produced useful commitments. If workload was to be reduced, check actual assignments and coverage. An announced initiative is not an implemented change.

Use early indicators such as clearer priorities, completed onboarding and improved access to support alongside later indicators such as voluntary departures, tenure and vacancy disruption. In a small workforce, a decline from three departures to two is encouraging but too little evidence by itself to establish causation.

Record other changes that may explain the result, including hiring conditions, contract cycles or altered service demand. Keep employee experience and service continuity in view: a lower turnover rate accompanied by unresolved workload or suppressed concerns is not a convincing success.

Retention becomes a management discipline when feedback produces visible action and leaders learn which changes matter. The result should be a more sustainable employment relationship and an administration better prepared for the transitions that remain.

Evidence and source notes

1. First Nations Health Authority, Evaluation 2019/20 to 2023/24, final report, June 2025, printed page 76. Exit-interview findings concern a sample of 79 former FNHA staff; they are not national workforce estimates.
2. First Nations Financial Management Board, Human Resources guidance. Used for management foundations in its stated FAL context, with applicability determined by the Nation’s circumstances.

Research reviewed October 18, 2026. Workforce counts and financial examples are hypothetical. The retention learning cycle is original Arcus analysis; no causal effectiveness claim is made for the proposed interventions.