Total compensation is base salary plus variable pay, equity, and benefits. Salary benchmarks usually cover base only, so evaluate the other parts separately.
Standard deviation measures how spread out salaries are around the median. A small spread means market consensus; a large one means wide variation and risk.
Salary compression is when new hires earn nearly as much as tenured staff, usually because market rates rose faster than internal raises. It drives attrition.
Salary benchmarking compares your pay against market data for equivalent roles, matched by role, seniority, tech stack, and geography to stay meaningful.
Annual reviews are too slow for LatAm tech. Check bands against market data quarterly to catch drift early, even if formal adjustments still happen yearly.
Treat retention as a range design problem, not "pay more". Set a retention floor and an acquisition ceiling, and track the drift between your pay and the market.
A passive candidate is a professional who is employed and not job hunting. Reaching them means an offer that beats their current pay, not just matches the market.
On-site and hybrid roles use country benchmarks; fully remote uses a regional USD benchmark. Layer modality, industry, size, and English on top of geography.
The gender pay gap is the pay difference between men and women in equivalent roles. In tech it often starts at the offer stage; market-based bands reduce it.
Requiring English narrows the pool and puts you in competition with US and EU remote employers, pushing the range up. Make English optional to widen reach.
Flexible schedule and remote options are now baseline expectations, not differentiators. Equity and education budgets can still attract by seniority and role.
Larger companies usually pay higher base salaries; startups often close the gap with equity, scope, and flexibility. The size effect grows with seniority.
Benchmark LatAm tech salaries with data segmented by role, stack, and seniority, drawn from real hiring behavior rather than annual self-reported surveys.