We took rows out. The market moved anyway.

July was the month we audited our own machinery — and published what we found. Two hundred and twenty-five observations that never deserved a vote — an over-eager club model, and behind it our own test accounts — are out of the pool, two sources emptied with them, and the published cell count now reads 280 against June's 310. Then, with the noise gone, the city league moved for the first time.

• 4,290Observations in the pool
• +254Net growth since the June issue
• 280Published cells, down from 310
• 670Published rates re-checked this week

Source: GigXchange Index ingest telemetry, June-issue close to 31 July 2026.

Where the market settled in July

Every figure below is performer pay, in pounds — ticket money, door splits, gear hire and travel are refused at the database layer. A city or region cell needed three trusted sightings to publish; under this issue's method one observation seen repeatedly could meet that, and the UK baseline had no floor (see the Method note). Reading note: n counts weighted samples, not unique gigs — an observation is counted at every geographic level it informs and once per re-sighting, which is why n runs far above the 4,290-observation pool.

By gig type

Weddings set the ceiling at £1,042, up £37 on the June issue, and private parties sit £42 behind at £1,000 — functionally the same market at this lineup size. Corporate holds at £839 for a third issue; fixed budgets do not negotiate. At the floor, pubs read £375 against the £425 we printed in June: the bulk of that fall is the July quarantine taking fabricated rows out of the cell (Section Three walks through it), not landlords paying less.

By band size

Solo £328, duo £613, £1,042 for a 3–4 piece, £1,433 at five-plus. The big cliff is still duo-to-band — a £429 jump, the price of turning background music into an event. Per chair, though, growth gets cheaper: a quartet works out near £261 a musician, while the five-plus median spread across six players comes to about £239 — because the van, the PA and the admin cost the same whether four people share them or six.

Across the whole grid

Weddings lead the 3–4 piece column, with private parties close enough behind that a 3–4 piece client is effectively choosing on service, not price. Other lineups have other leaders: private parties at solo (£496), corporate at duo (£765), and club (£2,291) and festival (£1,515) at five-plus. Corporate compresses into a narrow band; pubs and clubs anchor the floor. Two figures earn a caution flag rather than a quote: club five-plus at £2,291 stands on six samples, and the theatre 3–4 piece on three — read both as bearings, not benchmarks.

City league — first blood

Thirteen UK cities publish a wedding cell this issue, ranked by the typical 3–4 piece fee. After three issues of a frozen table, July finally produced movement: London's median climbed from £903 to £917 and took fourth place from Leeds. One swap is not a trend — but it is the first time fresh evidence has re-ordered the league rather than confirmed it.

Coverage outlook

We are keeping the same honest score as last issue: the corporate cells for Nottingham and Sheffield and the festival cells for Brighton and Edinburgh, first promised in May, are still short of the three-observation bar. Scraping cannot manufacture city-tagged event fees and the artist panel reports asking rates, not gigs. Those four cells get built when working musicians report real fees at gigxchange.app/rates — that route, or none.

Weddings

Weddings stay the deepest cell in the Index and the dearest 3–4 piece booking. The July curve runs £328 solo to £1,433 at five-plus, and the flagship 3–4 piece median lands at £1,042 — £37 up on June, the largest move this segment has shown, though still modest against a pool this deep.

Pubs and bars

Pubs hold the floor at £375 for a 3–4 piece — £50 under June's printed figure, mostly because the July quarantine took rows out of the cell rather than because rooms pay less. What the floor lacks in headline it returns in rhythm: repeat slots, a built-in crowd, minimal overheads. A month of steady residency nights can still beat a single wedding fee.

Built with Method v1 - method history and corrections policy.

From an observed fee to a percentile you can quote in a negotiation — the pipeline in plain English.

What we cover

UK live music only. A published cell is the meeting point of three axes: where (a canonical city, one of twelve UK regions, or the national baseline), what kind of gig (seven types, weddings to theatre/pit), and how many musicians (solo, duo, 3–4 piece, 5+). Each cell carries four percentiles — p25/p50/p75/p90 — never an average.

Where the data comes from

The July pool draws on six sources, two fewer than June: weekly refreshes of published UK market rates, Musicians' Union and Equity published minimums, a monthly panel of artist asking rates, fees reported by gig-workers, and human-reviewed extractions from public pages. Venue budgets and the ticket-derived club model both left the pool in the July audit — Section Three explains each exit. Every observation arrives tagged with source, gig type, band size, venue type and location.

Cities with thin data

Publication happens at three levels: city (major UK metros, satellite towns folded into their hub — Hove counts as Brighton, Oldbury as Birmingham), region (the twelve official UK regions) and the UK baseline. Three trusted sightings earned a city or region cell its place in print — under this issue's method, possibly one observation seen three times (see the Method note). Ask the live calculator about somewhere thinner and it walks the chain — city, then region, then the national figure with a regional adjustment — and tells you which level answered you.

Regional adjustments

The adjustment layer blends ONS regional household income and family-spending data with public industry price benchmarks, tuned per gig type: corporate fees track regional wealth closely, wedding spend barely does, and pub fees hardly flex at all because venue economics set the ceiling everywhere. Adjustments only ever fill gaps — wherever real observations exist, they win outright.

Keeping junk out

Five automated gates separate an incoming observation from a published cell: required fields, a vetted-source whitelist, an artist-fee-only rule enforced in the database itself, per-musician-per-hour plausibility bounds, and a duplicate fingerprint. One of the five spent the spring switched off without anyone noticing — Section Four tells that story straight, including what the gate rejected the day it came back.

Refresh cadence

Cells rebuild nightly at 05:00 UTC; published market rates refresh on Mondays; the artist panel snapshots on the 28th of each month; approved submissions flow in daily. Each rebuild is archived, never overwritten — that archive is what makes the issue-on-issue comparisons in Section Three possible. A cadence note for the series: the report now compiles overnight on the 28th of each month. This first issue under the new schedule was compiled on 31 July, so its window runs to the end of the month; from Issue 05 the window closes on the 28th.

Honest scope

A record of what the UK market paid or asked across the observation window — nothing more. Not a price recommendation, not a regulated benchmark, not financial advice. A legitimate gig can sit outside every range we print. Free to reuse under Creative Commons Attribution 4.0.

Why some sources count more than others

A fee that provably changed hands outranks a listings-page price, so every observation carries an effective trust weight (source weight × confidence). One change of basis to declare this issue: the table below prints the average weight actually applied to this month's rows, not the headline dial setting. Where the two differ — verified reports most visibly — it is because older rows keep the weight they entered with; the time series never retro-edits itself.

What changed in the July dataset

The pool stands at 4,290 approved observations, up 254 on the June issue. The number that matters more is smaller: 225 — the rows we removed from our own index, in public.

Deduplication and re-sighting

A listing that is still there next Monday tells you its price is holding — but it must never vote twice. When the weekly refresh recognises a listing it already holds, it bumps a sighting counter instead of inserting a twin: 670 re-sights this week, none of them widening the sample. Since July, a re-confirmed listing also ages from its latest sighting, not its first — a price the market keeps confirming no longer fades as stale.

How we keep junk out

Every observation faces the same gates whatever its source, and every rejection is logged with its reason — the audit trail is part of the product. This issue owes you a confession: the range check spent roughly ten weeks not running, and nobody noticed until the July audit.

1. Has the basicsNo amount, no source, no gig type — no entry. Incomplete rows are refused at the door, not patched later.
2. From a known sourceOnly the vetted source list may write to the index, through a single ingest funnel. A new feed earns its place via a weighting review first.
3. Is an artist feeThe database layer itself refuses ticket money, gear hire, licensing and door splits. Only pay to performers counts.
4. Passes the range checkPer-musician-per-hour bounds, set per gig type. This is the gate that silently fell out of the pipeline in a spring rewrite — for about ten weeks nothing was range-checking incoming rows. It is back, unknown gig types now bounce outright, and on its first day back it rejected a £25,000 "wedding band" and a £100 one from the same feed.
5. Not a duplicateEvery observation is fingerprinted. A re-checked listing bumps a sighting counter; it never becomes a second row.
6. Sealed against our own test dataNew since the quarantine: every adapter checks whether a profile belongs to a seed or test account before ingesting, and fails closed when it cannot tell. The 225 rows we removed have no way back in.

Anonymous submissions

Anyone who has played or booked a UK gig can report the fee at gigxchange.app/rates — no account needed. Submissions are rate-limited, sanity-checked and human-reviewed before they touch a cell. Three approved reports unlock every cell and trend line for 30 days.

External references used for regional adjustments

• ONS regional household income and family-spending data (latest releases).
• Music Venue Trust grassroots venue occupancy and capacity data (2025 survey).
• Public 2026 UK wedding and function-band price benchmarks.
• Help Musicians / Musicians' Union — Musicians' Census (~6,000 respondents).

A transparent record of fees actually paid or asked on UK gigs, with every cell showing its sample size, sources and refresh date — and a standing public record of the corrections we make to our own data, including the ones that shrink our own headline numbers.

1. Not a recommended price.We publish where the market sits, not where you should sit in it. Undercut or exceed any figure here with a clear conscience.
2. Not an official benchmark.No industry body endorses or regulates these figures. Independence is the point.
3. Not financial advice.Evidence in, decision yours.
4. Not the whole market.Most UK gig fees are agreed privately and surface nowhere. An index can only weigh what it can see.

What's still missing

• Confirmed platform bookings — the 1.00-weight reference signal — still contribute zero rows. The bookings we hold are dominated by our own test accounts, and counting those would repeat the exact mistake this issue spent a section undoing. The source switches on when real transactions arrive in volume, not before.
• Venue budgets restart from nothing: the first genuine venue budget the index ever ingests will be the first row in that source.
• The club picture is thinner without the ticket-derived model — the five-plus club cell rests on six observations and is directional at best.
• Festival and theatre stay the sparsest gig types; the 3–4 piece theatre cell stands on three observations.
• No genre-level cells yet — every observation carries the tags, and publication is waiting on volume, not engineering.

Licence

The report, the live figures at gigxchange.app/rates and the raw dataset are all published under Creative Commons Attribution 4.0 — use them, republish them, build on them commercially, credit "GigXchange Index, gigxchange.app/rates".

Help shape the next issue

One real fee from one real gig outweighs a hundred scraped listings — July proved it outweighs a hundred of our own estimates too. Report yours anonymously at gigxchange.app/rates — sixty seconds, no signup — and three reports unlock full granular access for 30 days.