Our methodology

Built on real market data, not guesswork

TeamCalc benchmarks synthesise institutional statistics, market intelligence, and live platform signals into a single defensible figure for every role. Every number carries a transparent quality rating so you always know how much weight to give it.

Where our data comes from

We synthesise data from four independent source categories, each contributing a different lens on the UK tech compensation landscape. No single source tells the full story — combining them produces benchmarks that are more robust than any one in isolation.

1

Government and institutional data

Our foundation layer. We anchor to official UK labour market statistics published by the Office for National Statistics (ONS), covering hundreds of thousands of employee records. This gives us statistically robust baselines broken down by occupation and region — employer-reported data that eliminates the self-reporting bias inherent in online platforms.

2

Market intelligence platforms

Real-time market signal from the platforms where tech professionals research and report compensation. We aggregate across multiple major salary platforms — including leading job boards, compensation databases, and professional networks. No single platform gives the full picture, which is why we cross-reference across several.

3

Industry research

Annual salary surveys and reports from specialist technology recruitment firms and industry bodies, based on actual placements and market analysis. Combined with annual developer surveys, these are particularly valuable for trend detection and seniority calibration — they capture shifts before they appear in the institutional data.

4

Specialist community data

UK-specific tech compensation communities and crowdsourced data. We use this as a validation layer — sense-checking the institutional and platform data against what practitioners report in specialised forums. Community data surfaces outliers and emerging trends, but carries less weight in the final blend than institutional sources.

Institutional data

Government statistics

Market platforms

Compensation databases

Industry research

Recruitment surveys

Community data

Validation layer

TeamCalc Engine

Validate · Weight · Blend

Four independent source categories flow into a single blending engine. Institutional data forms the foundation; market platforms provide real-time signal.

How we turn raw data into benchmarks

Raw data from four source categories passes through a structured pipeline that validates, weights, and blends it into a single defensible figure for each role, region, and seniority combination.

Multi-source triangulation

Every benchmark is derived from multiple independent data points. Institutional data carries more weight than crowdsourced data — as you'd expect from employer-reported statistics covering hundreds of thousands of records. We account for known biases in each source type: some platforms over-represent senior roles at top-tier companies, while others skew toward earlier-career professionals.

Recency weighting

The tech salary market moves fast. A three-year-old data point for an ML engineer may bear little resemblance to today's market. More recent data carries proportionally more weight in the blend, and data beyond a defined age threshold is excluded entirely.

Regional calibration

We model salary variation across 13 UK regions, anchored to official regional pay data. Some regions have deeper data coverage than others — London and the South East naturally have more data points than Northern Ireland or Wales. We're transparent about this: every benchmark carries a quality score that reflects the depth of its regional data.

Seniority modelling

We model the full career arc from junior to executive, with distinct benchmarks at each level. Leadership roles — Engineering Manager, VP Engineering, CTO — additionally account for company stage. A CTO at a seed startup operates in a fundamentally different compensation market than a CTO at a public company, and our benchmarks reflect that.

Continuous refinement

Benchmarks are refreshed as new data becomes available — major updates when institutional data publishes, and incremental updates as new market intelligence surfaces. Aggregate, anonymised usage patterns help us identify where our benchmarks may need adjustment, creating a feedback loop that improves accuracy over time.

Sources

4 tiers

Validation

Outlier removal

Weighting

Recency + bias

Blending

Multi-source

Quality scoring

4 dimensions

Your benchmark

p25 / median / p75

Our methodology is designed to surface the most defensible figure, not the most optimistic one.

How we measure data quality

Most salary tools give you a number and leave you to guess how reliable it is. We think you deserve to know. Every TeamCalc benchmark is assigned a quality score based on four dimensions.

Source depth

How many independent sources contributed to this specific benchmark. More sources mean more confidence.

Source agreement

How closely those sources agree with each other. Tight clustering suggests the market rate is well-established.

Freshness

How recent the underlying data is. Fresher data receives a higher quality rating.

Directness

Whether the benchmark comes from direct data for this exact role/region/seniority, or is estimated from related data.

Quality labels

Strong — Multiple recent sources in close agreement. High confidence in accuracy. Suitable for board-level planning.
Good — Reasonable coverage with some estimation. A solid benchmark for most planning purposes.
Indicative — Limited direct sources or significant estimation involved. Use as a starting point and validate with your own market knowledge.
Directional — Largely derived from related data. Treat as a rough guide — we recommend supplementing with your own research.

Live example from the product

Senior Software Engineer · London

£85,000£72,000 – £98,000

p25 – median – p75

Data quality

Strong

Hover the quality badge to see the full breakdown. Every benchmark in TeamCalc includes this rating, so you always know the confidence level behind the numbers you're using.

What we don't do

Transparency means being honest about limitations, not just capabilities.

We don't fabricate data

If we don't have strong data for a role/region/seniority combination, we say so. We'd rather show an “Indicative” quality rating than present a guess as fact.

We don't include equity or variable compensation

Stock options, RSUs, and bonuses vary too widely by company to benchmark reliably. Our figures represent base salary. For early-stage leadership roles where equity is the dominant compensation component, keep this in mind.

We don't claim perfect accuracy

Salary data is inherently imprecise — no methodology eliminates uncertainty entirely. Our goal is to give you the most defensible starting point available, with the transparency to adjust based on your own market knowledge.

See how it works

Model your first engineering team for free. See real salary benchmarks, full cost breakdowns, and quality scores — no credit card required.

Get started free