Methodology & privacy
Useful salary data needs rules people can trust.
PayYako is designed to create directional market evidence while limiting the risk that any contributor can be identified.
Collection
Contributors submit job, experience, location, organisation type, monthly gross pay, currency, and benefits. We do not request names, email addresses, phone numbers, or employer names.
Normalization & moderation
Job titles are deterministically matched to a controlled role taxonomy. Currency is converted using the configured reference rate. Confidence rules flag unmatched titles, implausible values, and salary–experience mismatches for human review.
Benchmark calculation
Only approved records enter calculations. We report count, minimum, maximum, average, median, 25th percentile, and 75th percentile on monthly gross USD-equivalent pay.
Publication threshold
A benchmark is public only after it contains at least five approved records. Individual submissions, notes, proof links, network hashes, and exact record-level details are never exposed publicly.
Privacy & abuse controls
Raw network addresses are not stored. A one-way keyed hash may be retained for rate limiting. Public writes are validated server-side, and the database does not grant anonymous access to salary submissions.
Important limitations
The dataset is voluntary and may not represent every sector, employer, city, or informal work arrangement in Zimbabwe. Self-reported values can contain errors despite moderation.
Exchange rates, benefits, bonuses, commissions, and role scope can change the meaning of base pay. Benchmarks are evidence for a conversation, not a guaranteed offer or compensation recommendation.