Answer ten questions about your own organisation. Every answer is scored against real Philippine payroll data, and you get a profile showing exactly where your compensation structure is carrying risk you may not be able to see yet.
Anyone can get an average salary for a role in ten seconds. Nobody can tell you whether your structure is exposed, or what to fix first.
It tells you what other people pay. It does not tell you whether your own band has quietly stretched to a six-fold spread, or whether your promotion differential still justifies the accountability that comes with it.
In 2026, several industries looked flat at the midpoint while their bottom decile fell and their top decile rose. If you only watch the middle, the movement that predicts attrition is the movement you cannot see.
Every result here ends in specific next steps: something to verify, produce, monitor, or compare. No generic advice, and nothing that requires a sales call to be useful.
Whether your increases are keeping pace with what actually happened to prices.
Whether the gaps between your levels still justify the step up.
Whether you can see movement inside a band before it becomes attrition.
Whether your pay structure reflects how AI is already changing what roles are worth.
Each answer is compared against Sprout BenchMark payroll data as you go. Nothing is gated, and your result appears at the end.
Inflation fell to a nine-year low of 1.7% in 2025, then reversed hard. The January to May 2026 average landed at 4.5%.
Measured against that benchmark, every job level in the Sprout BenchMark payroll base lost purchasing power. Rank-and-file employees came closest to holding ground, and even they fell short by roughly eight tenths of a point. Managers and officers landed near minus two and a half.
The average manager increase of 2.15% conceals the real movement. The bottom decile of managers took a nominal pay cut, while the upper quartile advanced 3.09%. The band is stretching, not rising.
In the Sprout BenchMark records, entry-level pay rose fastest at 3.68% while manager pay barely moved at the median. The consequence is that the multiple between each level narrowed, whether or not that was anyone's intention.
Median pay ratio, compressing
Median pay ratio, compressing
Median pay ratio, compressing
Averages in the most knowledge-intensive industries look unremarkable. Split the Sprout BenchMark distribution and the picture changes completely.
In Professional, Scientific and Technical Services the top decile gained 6.16% while the bottom fell 4.72%. Inside software development, legal, and product management, the tenth percentile fell while the ninetieth rose. The tasks entry-level professionals were hired to perform are the tasks generative tools now handle, and the market has stopped paying a premium for that rung.
Media and Communications shrank 13.8% while its top decile gained 15.2%. Fewer people, paid more, concentrated at the senior end. Software development stayed flat in a sample that grew 6.2%, losing workforce share, and received the smallest increase of any major function at +0.64% while still paying 54% more than customer support.
A Sprout study across more than 1,000 Philippine companies tracked attrition, tenure, absenteeism, and unproductive hours by salary percentile. The pattern is consistent enough to plan against.
Attrition falls almost twofold as you move up the salary range. Employees below the 25th percentile leave at 14.23%. Above the 75th, that drops to 7.26%. The middle two brackets sit close together at 11.43% and 11.08%, which means most of the retention benefit is concentrated at the two ends.
Read that alongside the 2026 Sprout BenchMark finding and the implication is uncomfortable. The cohort whose real wages fell furthest is the same cohort that already leaves at nearly double the rate of the top quartile.
Attrition declines at every step up the salary range, and the drop between the third quartile and the top quartile is the steepest single move: from 11.08% to 7.26%.
The lowest-paid quartile has both the highest attrition and the longest average tenure at 4.33 years. Those two facts coexist because the bracket contains two very different populations: people who leave quickly, and people who stay for years because the alternatives are limited.
The middle brackets are the most mobile, at 3.33 and 3.23 years. That is the group most likely to move for a modest raise, and the group whose pay moved least in 2026.
Absenteeism does not decline with pay. It is lowest in the second quartile at 10.03% and highest above the 75th percentile at 18.44%. The likeliest explanation is that these are two different behaviours wearing the same label: financial strain at the bottom, and greater flexibility to take leave at the top.
Unproductive hours sit between 35.6 and 36.2 per month in three of the four brackets. The 25th to 50th percentile is the exception at 20.88 hours, roughly forty percent lower than every other group.
Whatever is happening in that bracket is worth understanding rather than assuming. It is the one place in this data where a pay position appears to coincide with materially different behaviour.
Retention, tenure, absenteeism, and unproductive-hours figures are drawn from Sprout's salary benchmarking study covering more than 1,000 Philippine companies over the January to December 2023 period. Salary movement figures are from the 2025–2026 payroll base described in the methodology below.
The diagnostic scores four dimensions and names the weakest one. That weakest dimension is the profile, because it is the constraint that makes everything else harder to fix.
Your biggest exposure is not what you pay. It is that you cannot see what is happening inside your own bands. Everything else stays a hypothesis until percentile movement is visible.
Your structure is sound. Your absolute numbers are losing a race against prices. The most fixable of the five, and the most time-sensitive.
You pay competitively at every level. The problem is what sits between them. Shows up first as declined promotions, later as resignations.
Your pay structure still reflects what roles were worth before AI entered the workflow. The gap compounds fastest in the functions where adoption is deepest.
You read your own structure accurately and price against a market most organisations cannot see. The risk here is complacency rather than exposure.
Ten questions. Three minutes. No email needed to see your result.
Two datasets sit behind this diagnostic. Here is what each one covers, and where each one stops.
Figures are weighted by headcount. Segments below the minimum size are excluded rather than reported with wide error. Outsourcing and staffing employers form a substantial share of the industry sample, so that view leans toward the sector.
Attrition, tenure, absenteeism, and unproductive hours come from a separate, earlier study. It is used here to show how pay position relates to behaviour, not to describe 2026 conditions.
Ten questions, four dimensions, and a benchmark built from real payroll records. Finish it and take the one-page summary with you.