Methodology & Data Sources

Data last updated July 2026.

Every figure on GamedevSalary.com is resolved from three data layers, in strict priority order. Numbers are always labeled with the layer (and source) they came from — we would rather show you an honest gap than an invented number.

The three data layers

  1. Live submissions — anonymous salary reports submitted directly on this site. Submissions are reviewed manually before they are published in aggregate; spam and implausible values are removed.
  2. Community data — public community salary datasets, individual reports shared with us, and anonymous submissions on this site (2018–2026). 1,958 individual salary reports in total. Job titles are mapped to our standardized role list via deterministic rules; salaries outside $10k–$500k are excluded. A role/region/seniority bucket needs at least 3 reports (2 for rare specialist roles) before we publish its median. Low/high ranges derive from the bucket's spread, or ±20–25% of the median when the bucket is small.
  3. Public industry reports — aggregate figures cited from published salary surveys (see the source list below). We cite a small number of aggregate statistics per source with attribution; we do not republish any report's underlying dataset.

Where none of the three layers covers a combination, the site shows a clearly marked model estimate: an editorial interpolation from category baselines and multipliers. Model estimates should never be cited in negotiations.

Experience bands & interpolation

Data is anchored at five experience bands (0–2, 3–5, 6–10, 11–15, 16+ years) with band midpoints at 1, 4, 8, 13 and 18 years. Values between anchors are linearly interpolated. Two sanity passes run over resolved anchors: cliffs and inversions between adjacent bands are smoothed, and career progression is kept weakly monotonic (a band never drops below ~96% of the previous one). This favors a realistic career curve over raw bucket noise when buckets are thin.

Currency & regions

All comparisons are in USD. Community reports in other currencies are converted at fixed reference rates (GBP 0.79, EUR 0.92, JPY 150, KRW 1,330, AUD 1.52 per USD). Regions are coarse on purpose — buckets need enough reports to be meaningful, and country-level splits would fall below our minimum sample sizes for most roles.

Trust model

Chart bars encode confidence visually: color identifies the data layer, opacity scales with sample size and source quality, and interpolated or extrapolated years are dimmed relative to direct anchors. Tables on the salary guide pages show the same provenance as badges with sample sizes.

On the discipline benchmark table, Lead+ rows are floored at the senior IC median: lead buckets are small and title-mapped, and several sit below senior IC pay for reasons that are mapping noise, not market reality.

Community data sources

The community layer combines these public datasets with individual reports shared with us and anonymous submissions on this site. Public datasets are used as aggregate medians per role group, with attribution.

Industry sources

We gratefully credit the following published surveys and platforms. Figures cited from these sources are small numbers of aggregate statistics (medians per role group), used under citation with attribution — each is labeled inline wherever it appears. If you are a publisher and have questions or corrections, please get in touch.

Government reference data (all industries)

Role pages additionally show official government wage statistics for the closest matching occupation. These cover all industries, not just games, and are therefore displayed as a separate, clearly-labeled reference layer — they are never merged into the games-specific benchmarks or the chart.

Attribution: BLS data is in the US public domain. Canadian data is used under the Open Government Licence – Canada. UK data contains public sector information licensed under the Open Government Licence v3.0. Stack Overflow survey aggregates contain information from the Stack Overflow Developer Survey, made available under the Open Database License (ODbL).

Purchasing-power adjustment

The "US buying power" column on role pages divides the nominal USD median by a regional price-level index (US = 1.0), computed from World Bank 2023 data (PPP conversion factor ÷ official exchange rate, indicators PA.NUS.PPP and PA.NUS.FCRF, used under CC BY 4.0). Regions are approximated by representative countries — UK: GBR; Western Europe: DEU/FRA/NLD/ESP; Nordics: SWE/FIN/DNK/NOR; Eastern Europe: POL/CZE/ROU; Asia Pacific: JPN/KOR/AUS/SGP. Worldwide and Rest of World are shown unadjusted. Treat these as directional, not precise.

Job listings

Role pages link to matching openings on external games job boards. Listing snippets are fetched from GameJobs.co's public feed during the build and carry their fetch date; if the feed is unreachable the last committed snapshot is used. Some links may become partner links that help fund the site — they never influence the salary data.

Limitations

  • All figures are base salary only — no bonus, profit share, or equity.
  • Community data skews toward respondents of the communities the survey reached; it is not a random sample.
  • Some sub-roles inherit discipline-level figures where reports only publish broader categories.
  • Regional buckets hide large in-region differences (e.g. California vs. Midwest US).

Corrections

Spotted something wrong — a figure, a mapping, an attribution? Please reach out via the about page. Corrections are welcome.