Research
Evaluating Turnover Claims in Philippines Outsourcing Proposals
A buyer's guide to definitions, denominators, cohort data, and reporting rules behind a vendor's turnover number.
On this page
Key stats
The U.S. Bureau of Labor Statistics Job Openings and Labor Turnover Survey separates quits, layoffs and discharges, and other separations, which shows why a single turnover figure can hide very different causes.
Methodology
This note reviews the definitions used in the Bureau of Labor Statistics Job Openings and Labor Turnover Survey, labor-force material from the Philippine Statistics Authority, and International Labour Organization guidance on labour statistics. It asks how a buyer can test a turnover claim in an outsourcing proposal without pretending that one rate describes the whole relationship. The method is a due-diligence aid for a specific account, not a ranking of providers. It does not assert that low turnover is always better or that public data can predict a named team's result.
Evidence summary
Ask the vendor to define what counts as a separation before comparing any number. A figure that includes probationary endings, fixed-term completions, and involuntary exits describes a different phenomenon from one that counts only voluntary resignations. The definition should match the decision you are making, whether that is pricing risk, planning backfill, or judging management quality. A proposal that will not state the definition is offering a mood, not evidence.
Check whether the rate is monthly, quarterly, or annualized. A monthly rate multiplied by twelve is not the same as an annual rate computed from a full year of headcount, because seasonality and hiring ramps distort the result. Ask for the numerator, the denominator, and the period. Then recalculate the figure yourself before accepting the comparison. Write the period and the formula beside the figure.
Separate voluntary from involuntary turnover. A vendor that removes poor performers quickly may show a high involuntary rate and a healthy voluntary rate, while one that avoids difficult conversations may show the opposite. Neither number alone proves good management. Ask which category the vendor assigns to a failed probation and record the answer. Record the reason code with the count.
Use a cohort view to see whether new hires stay. A team can look stable overall while replacing most of its newest members every few months. Compare the share of hires still present after three, six, and twelve months. A high early loss points to recruiting, onboarding, or scheduling rather than to the general labor market. Ask for the cohort table in the same workbook as the headcount.
Check the denominator. Turnover can be calculated against beginning headcount, average headcount, or the number of filled positions. Each choice produces a different rate, and a vendor may choose the one that looks best. Ask for the formula and apply it consistently across the proposals you compare. A rate without its denominator cannot be audited. Record which denominator the vendor used.
Compare like with like before drawing a conclusion. A small team with ten people will swing more from one departure than a site with hundreds, so a single month is not a trend. Role mix, shift pattern, tenure, and location all affect the result. If the proposals describe different populations, compare the definitions and the cohort data instead of the headline rate.
Verify the claim with payroll or HR records rather than a slide. Ask for aggregate counts by month, with the separation reason and the last working day. Redact names and personal data, but keep the counts consistent with the headcount report. A vendor that cannot produce a reconciliation between the two should not be scored on the number. Ask for the account-level figure in writing.
Distinguish site-level from company-level turnover. A provider may report a strong company average while one account or one shift carries most of the loss. Ask for the figures for the team that would serve you, not only the enterprise summary. If the provider cannot isolate the account, ask how it will report after the contract starts. Note the date and the person who reconciled the counts.
Write the reporting requirement into the contract. Specify the definition, the frequency, the formula, and the person who signs the report. Include a short note on the main reasons and the actions taken. Review the first two reports together so the format is settled early. Name the report owner and the escalation path for a late submission. Name the owner of the monthly report.
Interpret the number with the operating context. A site opening a new account may hire aggressively and lose more people during the ramp, while a mature team may be stable but expensive to change. Ask what happened during the period and what the vendor did about it. Turnover is a signal to investigate, not a verdict on its own. Ask what changed in the period before accepting the explanation.
Put the turnover discussion on the same calendar as the service review so the number is read with the operating context. A monthly review can compare the rate with hiring, scheduling, and quality data instead of treating it as a standalone score. Record the questions the number raised and the answers the provider gave. Over time, that record shows whether the reporting is improving or simply repeating.
Ask for the definition
Request the written definition of a separation, including probation, fixed-term completion, and involuntary exit. Check whether rehires and internal transfers are counted. A vendor that cannot produce a definition cannot produce a comparable rate. Put the definition in the vendor file beside the number. Keep the signed definition with the proposal. Ask the provider to confirm which separations are included.
Recalculate the rate
Ask for the numerator, denominator, and period, then compute the rate again. A figure that changes under recalculation needs an explanation. Note the formula and the source of the headcount. A simple check catches many presentation differences before they become a dispute. Show the recalculation in the review note. Attach the source table to the note.
Read the cohort
Request retention by hire cohort at three, six, and twelve months. The cohort shows whether the team retains new people or only long-tenured staff. Compare the cohort with the overall rate and with the account's hiring plan. A stable average can hide a failing onboarding pipeline. Ask for the same cohorts from every vendor you compare. Use a consistent cohort definition.
Check the scope
Ask whether the number covers the proposed account, the site, the country, or the whole company. A company average may not describe the team you will manage. Request the narrowest scope available and a plan for account-level reporting. Write the reporting scope into the contract. Record the scope in the contract schedule. Ask how the account will be isolated.
Verify with records
Compare the reported separations with aggregate payroll or HR counts. Look for a reconciliation between the monthly headcount and the separation list. Keep personal data out of the review and use counts and reasons. A number that cannot be reconciled should be treated as unverified. Note the reconciliation date and the reviewer. Keep personal data out of the working file.
Set the reporting rule
Define the frequency, owner, and format for ongoing turnover reporting. Include the reason mix and a short action note. Review the first reports together and correct the format early. A metric with an owner is more useful than an annual assurance. Review the first two reports before the next renewal. Agree the format before the first report is due.
FAQs
Is a lower turnover rate always better?
No. Very low turnover can reflect a stable team or a lack of performance management, and very high turnover can reflect rapid growth. Read the voluntary and involuntary mix, the cohort data, and the operating context before judging the number.
Should we compare a vendor's rate with a national average?
Only with caution. National data covers a different population and may use a different definition. Use it as context, not as a benchmark for a specific account, and state the difference in scope.
What is the most useful single check?
Ask for the definition and the denominator, then recalculate the rate. If those three elements are consistent, the conversation can move to causes and actions. If they are not, the headline number is not yet usable.
How should a new account's turnover be reported?
Report it separately from the provider's company total and show the cohort. A new account is expected to have a different pattern during the ramp. Separating the account prevents a strong company average from hiding a local problem.