CoolFace
Datasetpublic

gigxchange/uk-live-music-rates-2026-08

UK Live Music Rates - August 2026 (GigXchange Index) Monthly open dataset of what live music pays in the UK, by city, event type and band size. Every figure is artist take-home in GBP: agency-published rates are normalised by stripping a 20% commission before any percentile is computed. This is Issue 05 of the GigXchange Index, published 2026-09-03. The file is a snapshot of exactly what gigxchange.app published on 2026-09-03 - the same figures shown on the site and in the… See the full description on the dataset page: https://huggingface.co/datasets/gigxchange/uk-live-music-rates-2026-08.

sourceHugging Facecc-by-4.0updated 22d agoView on Hugging Face
0likes54downloads
Dataset Card

UK Live Music Rates - August 2026 (GigXchange Index)

Monthly open dataset of what live music pays in the UK, by city, event type and band size. Every figure is artist take-home in GBP: agency-published rates are normalised by stripping a 20% commission before any percentile is computed.

This is Issue 05 of the GigXchange Index, published 2026-09-03. The file is a snapshot of exactly what gigxchange.app published on 2026-09-03 - the same figures shown on the site and in the monthly report, with no export-side filtering. Each issue measures the market at a moment in time; it does not revise the month before it.

What UK gigs paid in August 2026, and why thirteen cities now sit GBP 22 apart. Fees by city, gig type and band size, from 4,431 observations.

Headline figures

  • —GBP 990 - Wedding band fees, UK national: A 3-4 piece wedding band typically takes GBP 990 in August 2026, GBP 52 below July. The middle half of bookings land between GBP 717 and GBP 1,367, with the premium tier at GBP 1,880.
  • —GBP 22 - The league that flattened: Thirteen cities now run from Edinburgh at GBP 1,067 to Manchester at GBP 1,045 - a spread of GBP 22 where July showed GBP 523. Nothing moved in the market; the model changed underneath it.
  • —407 - 407 cells, one grid for every city: Published rate cells rise to 407 from July's 280, because every city now answers the same 23 questions instead of only the ones its own data could reach. That is a coverage decision, not a surge in evidence.
  • —GBP 350 - Pub & bar gigs hold the floor: A 3-4 piece pub gig pays GBP 350, most between GBP 250 and GBP 600 - GBP 25 under July. Solo pub work sits at GBP 250, a duo at GBP 375 and a five-plus lineup at GBP 525.
  • —GBP 866 - Function work below the wedding line: Private parties pay GBP 866 for a 3-4 piece and corporate GBP 812, putting them GBP 124 and GBP 178 behind weddings. The two function markets have stopped reading as interchangeable at this lineup size.
  • —GBP 1,489 - What band size is worth: The wedding curve runs GBP 330 solo, GBP 526 duo, GBP 990 for a 3-4 piece and GBP 1,489 at five-plus. Per player a quartet and a six-piece both work out near GBP 248. Compare rates on artist profiles.

Read the full issue: https://gigxchange.app/rates/reports/august-2026 (PDF: https://gigxchange.app/rates/reports/gigxchange-index-2026-08.pdf).

Files

FileRowsWhat it is
gx-index-august-2026.csv407Published rate cells as at 2026-09-03: 299 city, 86 region, 22 UK baseline
sources.csv9Every source class, its approved-observation count and its quota cap
DATASHEET.md-Datasheet (Gebru et al. framework): motivation, composition, collection, uses, maintenance

Measured or derived - read is_imputed

203 cells are measured, 204 are derived. A city cell is measured only when it holds at least 3 observations collected in that city (local_sample_size >= 3); below that it still publishes, labelled is_imputed = true, imputation_method = blanket_pool, meaning the figure comes from the location-less national layer applied to that city. Region and UK rows are measured by definition. Filter `is_imputed = false` for measured-only analysis - that flag is the recipe; do not reconstruct it from the split counts.

Columns

FieldTypeDescription
scopestringAggregation level: uk (national baseline), region, or city
city_slugstringCity identifier, e.g. london (empty for non-city rows)
regionstringUK region (empty for city and national rows)
use_casestringEvent type: wedding, corporate, privateparty, pubsand_bars, club, festival, theatre
band_size_bucketstringBand size: solo, duo, trioquartet, fiveplus
p25numberLower-quartile fee, GBP, artist take-home
p50numberMedian fee, GBP, artist take-home
p75numberUpper-quartile fee, GBP, artist take-home
p90numberPremium-tier fee, GBP, artist take-home
sample_sizeintegerDistinct observations behind the cell - evidence held, not gigs that happened
local_sample_sizeintegerObservations collected in this city (city rows only)
blanket_sample_sizeintegerLocation-less observations applied to every city, e.g. national rate cards and union minimums (city rows only)
is_imputedbooleantrue = derived (fewer than 3 local observations); false = measured
imputation_methodstringHow a derived cell was filled (blanket_pool); empty for measured cells
last_refreshedstringDate these figures were published (ISO 8601): 2026-09-03

sources.csv

FieldTypeDescription
sourcestringSource class identifier
display_namestringHuman-readable source name
observationsintegerApproved observations from this source at publication (4,431 in total)
max_sharenumberMaximum share of any published cell this class may own; the caps sum to 1.00
statusstringlive, or no rows this issue
rationalestringWhy this source counts this much

Data collection

Observations are ingested through one gated path from nine documented source classes: confirmed platform bookings, anonymous public submissions, verified rate imports, artist profile asking rates, Musicians' Union and Equity published minimums, agency rate cards, venue gig budgets, web extraction and a reserved ticket-derived class. Each is capped at a maximum share of any published cell, so no source gains influence through volume alone: agency rate cards are about 79% of the corpus and may decide at most 10% of any figure. Every observation passes plausibility bounds, source whitelisting, payment-type restriction and duplicate detection; rejections are logged with a reason. Percentiles are a weighted cumulative crossing, not row replication. Method in full: https://gigxchange.app/rates/methodology

Bias and honest limits

  • —These are advertised, asked or budgeted rates - not a record of money that changed hands. About 79% of the corpus is agency rate cards, so the sample skews toward the function market.
  • —Not a census. Most UK gig fees are agreed privately and surface nowhere.
  • —Not a recommended price, not a regulated benchmark, not financial advice.
  • —Many cells rest on few observations. Every row carries its sample_size - weigh it.
  • —A derived cell is an estimate of what a city would pay given how it prices elsewhere. It is never counted as evidence and never used to compute another figure.

Maintenance

Published monthly by GigXchange as a new, standalone dataset per issue (a new repo here, a new dataset on Kaggle, a new record on Zenodo). Earlier issues are never edited; corrections to a published issue are documented in the next one. Contact: https://gigxchange.app/contact

Publisher and source

Citation

Zenodo DOI (record of citation): 10.5281/zenodo.22389294

bibtex
@misc{gigxchange_uk_rates_2026_08,
  author    = {Zahid, Naumaan and {GigXchange}},
  title     = {UK Live Music Rates --- August 2026 (GigXchange Index)},
  year      = {2026},
  month     = {8},
  publisher = {GigXchange},
  url       = {https://huggingface.co/datasets/gigxchange/uk-live-music-rates-2026-08},
  note      = {CC BY 4.0. Zenodo DOI 10.5281/zenodo.22389294.}
}

Also available on Kaggle.

Licence

Licensed CC BY 4.0 - free to use, including commercially, with attribution:

GigXchange Index, UK Live Music Booking Rates 2026 (August). https://gigxchange.app/rates