How to measure internal mobility without counting transfers as exits
Keep genuine organisational exits separate from internal transfers, and build mobility measures that distinguish participation, vacancies and career progression.
Team WINC Wire
6 min read
When a restaurant supervisor moves to a sister hotel, the original operation has a vacancy. The group may have retained a capable employee. A useful people report records both outcomes without describing the move as someone leaving the business.
It becomes difficult to measure internal mobility when payroll records, vacancies and individual careers are treated as interchangeable. Establish a shared boundary and a small set of distinct measures before setting targets.
Start with the organisation boundary
Define what “internal” means for the report: one hotel, an employing entity or a specified group of entities. List the included operations and keep the definition visible.
A shared brand is not enough. Independently owned franchise properties may not belong to the same workforce reporting boundary. An employee moving between them should not automatically be marked as internally retained. Equally, a payroll migration inside the agreed group should not create a false external departure.
Statistical sources use different boundaries. The US Bureau of Labor Statistics’ JOLTS definitions count transfers to another location of the same firm within “other separations” because the survey measures establishments. Transfers within the same location are excluded. That convention is valid for its purpose, but it is not a ready-made definition of group-level employee loss.
Document the difference when using an external benchmark. Your internal analytics boundary is also not a legal determination about continuity of employment; complex employer changes need their own review.
Create an event dictionary
Ask HR, payroll and operations to agree a small dictionary, supported by examples from their systems:
- External hire: a person joins from outside the reporting boundary.
- External exit: a person’s employment within that boundary ends.
- Permanent internal move: an existing employee changes their substantive role or location inside it.
- Promotion: a move meeting a documented increase in role level or accountability, not merely a changed title.
- Temporary assignment: a time-limited placement with a recorded expected end or return.
- Administrative change: a payroll, cost-centre or reporting correction without a substantive career move.
Allow one event to have several descriptive flags, such as promotion and property transfer, without counting it twice in the overall move total. Keep temporary assignments separate from permanent mobility, and exclude administrative changes from both.
Use an effective date, not the date someone typed the change into the system. Preserve a person identifier across properties and a separate employment-spell identifier for genuine exits and rehires.
Measure people, events and vacancies separately
1. Employee mobility participation
For a simple quarterly cohort measure, define eligibility at the start of the quarter. Divide the number of those employees making at least one qualifying move during the quarter by the eligible opening cohort.
Hypothetical example: 50 of 500 eligible opening employees make a qualifying move. Participation is 50 ÷ 500 × 100 = 10%. If those 50 people make 60 moves, participation remains 10%; the additional moves belong in the event count.
Employees in the opening cohort who subsequently leave stay in its denominator. New starters are outside both numerator and denominator for this measure; show their moves separately rather than dropping them into the numerator. If policy restricts eligibility by tenure, disclose the excluded population and review whether the rule itself limits opportunity.
This is an opening-cohort measure, not a universal mobility standard. A separate whole-workforce rate using employee time at risk can address unequal exposure, but must state its units and method. Do not switch definitions halfway through a trend.
2. Internal fill rate
Divide qualifying vacancies filled by existing employees by all comparable vacancies filled during the period. Count completed appointments, not applications, offers or unfilled requisitions.
Independent hypothetical example: 24 of 40 filled vacancies go to internal candidates. Internal fill rate is 60%. That does not mean 60% of employees moved. It describes the source of appointments, not the share of the workforce experiencing mobility.
Fix the vacancy scope. Permanent positions, temporary cover and graduate rotations should not drift in and out of the denominator according to which result looks stronger. Track the vacancy left behind by an internal appointment so that hiring chains remain visible.
3. Progression and development
Split permanent moves into promotion, lateral movement and other agreed categories. A high move count may reflect valuable cross-training, repeated restructuring or people escaping an unsuitable role. Add the destination, employee’s stated objective and a later check on the placement.
Assess opportunity as well as completions: who could see vacancies, who applied, who was shortlisted and who accepted. Use relevant eligible denominators for each stage, and protect small groups from identification. A participation gap is a question for investigation, not a diagnosis by itself.
Reconcile site movement with group headcount
Use a movement bridge to explain changes instead of relying on payroll “leaver” codes.
Hypothetical property example: a hotel opens the month with 120 employees. It adds 12 external recruits and four internal arrivals, while eight people leave the group and six transfer to other group properties. Closing headcount is 120 + 12 + 4 − 8 − 6 = 122.
That property had 14 departures from its team, but only eight were external exits. At consolidated group level, the six transfers out should pair with six internal arrivals elsewhere. Internal movements cancel in the group headcount bridge; they still matter in operational staffing plans.
For a separate monthly external attrition rate, divide external exits by average employee headcount in the same boundary and period. Using the opening-and-closing average as an approximation here, eight exits ÷ ((120 + 122) ÷ 2) × 100 = 6.6%. Daily average headcount is preferable when staffing changes sharply. Exclude internal transfers from the exit numerator, and never mix employee counts with full-time-equivalent denominators.
Where records do not pair, investigate timing, duplicate identifiers, boundary errors or missing destination data. Keep an unresolved category until evidence settles the case. Do not force unmatched movements into external exits simply to make a spreadsheet balance.
Build a mobility review that managers cannot game
Review the measures together. Rewarding only site retention can discourage managers from releasing good people. Rewarding only internal fill can encourage nominal moves or appointments without readiness. Neither target should hide poor matching or block employees’ choices.
Use a compact quarterly review:
- Confirm the boundary, period and event definitions are unchanged.
- Reconcile external and internal flows with opening and closing headcount.
- Show unique movers, movement events and vacancy fill measures separately.
- Check access across properties, shifts, job families and relevant employee groups.
- Review placement quality after an agreed period, including whether employees are still in the destination role.
- Assign an owner to barriers such as delayed release dates or invisible vacancies.
A later retention comparison between movers and non-movers is descriptive unless the analysis addresses selection. Movers may already differ in performance, tenure, ambition or access to supportive managers.
For the organisational case behind the measures, see WINC Wire’s internal mobility strategy and its discussion of hospitality’s internal hiring channels. The practical priority is straightforward: make career movement visible without confusing a hotel’s staffing loss with the organisation losing an employee.
Sources
AI-generated editorial illustration.
