Before you dive into building salary bands, you need some relevant and reliable data and there are plenty of sources to choose from. Market data can give you helpful insight into industry standards, but it should guide your decisions, not dictate them. It’s always an estimate, not an absolute truth, and works best as a reference point rather than a rigid rule.
"Use survey data as a tool, not a rule."
That said, data still plays a vital role. One of its most important functions is to build credibility. When your compensation decisions are backed by solid benchmarks, it gives stakeholders more confidence, helps managers have clearer conversations about pay, and shows employees that your approach is grounded in evidence, not guesswork.
"The strength of your compensation strategy is only as solid as the data behind it."
I break data sources into two buckets + AI generated:
| Category | Description | Key Characteristics | Cost & Reliability | Why it matters |
|---|---|---|---|---|
| 1️⃣ Primary Data Sources | Benchmarking providers where you exchange your data for market insights (e.g., Ravio, Figures, Pave, Mercer, AON Radford). | - Robust, validated data |
Reliability: High | This is the most robust standard. It reflects actual payroll records, ensuring your pay decisions are fair, sustainable, and legally defensible. | | 2️⃣ Secondary Data Sources | Includes job boards, scraping platforms, local surveys, and informal data channels. | - Useful for filling gaps
Reliability: Medium/Low | These can provide quick context but lack the validation of payroll data. They are often skewed by "Selection Bias" (high/low outliers). | | 🤖 AI-Generated (Modelled) | Probabilistic "guesses" based on scraping the web, job adverts, and public forums. | - Fast and conversational
Reliability: Unverified | AI doesn't "know" the data; it predicts patterns. Without access to private payroll databases, it cannot provide the accuracy needed for benchmarking. |
When picking your data sources, think about what works best for your needs. You want a balance of scope, reliability, and cost, but you also need to ensure the data supports your compensation strategy rather than simply replicating industry norms. And remember, always be transparent with your team about the sources you’re using.