Ready to get started?
Join artists and venues on the UK’s peer-to-peer live music marketplace.
Most UK pub and grassroots venue bookers never formalise their programming decisions. They book on instinct, scattered across 3-4 hours a month of WhatsApp threads, late-night email scrolls and half-remembered till totals. AI doesn’t replace that instinct — it gives you a structured 60-minute monthly session where you pull last month’s data, feed it to an LLM, get a draft calendar and rebook recommendations, apply your local knowledge, and book. 4 jobs AI handles well: themed-night ideation, lineup curation, residency ladders, and return-slot decisions. Cost: £0-20/month (free ChatGPT or Claude Pro at ~£18/month). Safety rail: AI doesn’t know your room, your regulars, your sound limits, or local politics — it reads your data and drafts options; you choose.
Read 12 ChatGPT prompts for venue bookers for the inbox layer, vetting acts for filtering pitches, and venue compliance for the paperwork layer. This post is the next level: the programming decisions themselves.
This is the sixth post in our AI for UK music venues series. The first five covered 12 inbox-reply prompts, how to use AI to fill your venue, the venue tools stack, vetting acts, and compliance paperwork. This one tackles the strategic layer most venues never formalise: the programming decisions that determine what your room sounds like week after week.
We’ve seen the same pattern across UK grassroots venues we work with on GigXchange: a one-person booking team programmes on gut feel, scattered across 15-20 conversations a month, with no structured review and no data trail. The good bookers have strong instincts. The great bookers run those instincts through a structured process — and that’s where AI fits in.
Before any of the 4 jobs below, you need data. The Codex concern we hear most from pub managers is: “I don’t have clean exports.” You don’t need a database. You need an 8-column spreadsheet — or even a notebook — with these fields:
| Date | Night | Act | Fee (£) | Door count | Est. bar uplift (£) | Notes | Rebook? |
|---|---|---|---|---|---|---|---|
| 4 Apr | Fri acoustic | Jake Morris | £200 | 55 | £680 | Full room by 9pm | Yes |
| 11 Apr | Fri acoustic | Sarah Kent Duo | £300 | 32 | £410 | Quiet night, rain | Maybe |
| 18 Apr | Sat covers | The Voltage | £500 | 78 | £1,120 | Great energy | Yes |
| 25 Apr | Sat covers | Blue Note 5 | £600 | 25 | £320 | Claimed 80 draw | No |
If you have a spreadsheet, great. If all you have is memory and till totals, AI can still structure the review — but confidence is lower and you should treat its suggestions as conversation starters, not decisions. The minimum useful dataset is 8-12 bookings with at least fee + door count filled in. That’s roughly 2-3 months of weekend-only programming at a typical pub running 4-6 live nights per month.
Strong venue operators respect this. If your Thursday acoustic night draws 40-60 people every week, they buy 3-4 drinks each, the act is reliable and charges £200, the sound engineer likes working with them, and the till shows £600-£900 bar uplift consistently — don’t over-programme it. AI will suggest variations because that’s what generative models do. The right answer is often: keep what works.
Only use AI for the nights that aren’t working or the slots you haven’t filled yet. A 120-cap pub running 4 live nights a month might have 2 that are solid and 2 that underperform. Focus the 60-minute review on the 2 weak nights. The strong nights get a 30-second “still working, keep going” and nothing else.
The most expensive mistake in programming isn’t booking the wrong act — it’s changing a night that was already working because a shiny AI suggestion made you second-guess a 12-month track record. On a 2.5% profit margin, the cost of one destabilised good night is £300-£800 in lost bar revenue over 4-6 weeks while you rebuild the audience.
This is the highest-leverage AI use for programming. Most pub bookers have run the same 3 formats for 2-3 years (acoustic Friday, covers Saturday, open mic Sunday). AI doesn’t replace those — it surfaces 6 concepts you haven’t considered, drawn from what works at similar venues across the UK.
I run a [CAPACITY]-cap [pub/bar/venue] in [CITY] that does live music every [DAYS]. Current audience skews [AGE RANGE], mainly [AUDIENCE TYPE e.g. locals, students, after-work crowd]. We currently do [LIST CURRENT FORMATS e.g. acoustic Friday, covers Saturday, open mic Sunday]. Our PA handles [DESCRIBE SETUP e.g. vocals + acoustic guitar comfortably, full band at a push]. Typical act budget is £[FEE RANGE] per night. Suggest 6 themed night concepts I haven't tried. For each give me: 1. Name (something a punter would recognise on a poster). 2. Genre / format. 3. Target audience segment and why they'd come. 4. Why it works in a venue this size. 5. One act type to anchor it (e.g. "jazz trio", "singer-songwriter with loop pedal"). 6. Estimated fee range for the anchor act (check GigXchange Rate Index at gigxchange.app/rates for UK benchmarks). Be specific to the UK scene. No generic suggestions. Default to ideas that work in a room of [CAPACITY] people with a [BUDGET] budget.
A good suggestion names a real format (“Songwriter Showcase — 3 original acts, 25 minutes each, £75 per act”), targets a specific gap in your week, and gives you a realistic fee. A bad suggestion says “jazz night” with no format, no fee, and no audience rationale. If AI gives you the second, ask it to be more specific or discard it.
Every themed-night concept passes through 3 filters before it touches your calendar:
Once you’ve chosen a format, you need a bill — and for multi-act nights, the running order matters more than most bookers think. The tempo-arc principle is simple: open mellow, build energy, close big. A 3-act bill where all 3 play at the same intensity loses 30-40% of the audience by the second act because there’s no shape to the evening.
Use the GigXchange Profiles compare tool to shortlist up to 4 acts side by side. Compare genre, fee, location, and verified gig history in 60 seconds. Then feed the shortlist to AI for running-order advice.
I'm building a 3-act bill for a [FORMAT] night at a [CAPACITY]-cap [pub/venue] in [CITY]. Doors at [TIME], first act at [TIME], last act finishes by [TIME]. Here are my shortlisted acts: [PASTE ACT NAMES + GENRES + TYPICAL SET LENGTH + FEE] Build me a running order with: 1. Open / support / headline sequence. 2. Genre and tempo arc (mellow → build → peak). 3. Set lengths for each (total must fit the window). 4. Fee split suggestion based on draw and slot. 5. One backup act suggestion if my headline cancels. The venue is [DESCRIBE SOUND SETUP]. Budget ceiling is £[TOTAL].
A common starting model is 50/30/20 (headliner/support/opener). On a £600 total budget, that’s £300 headline, £180 support, £120 opener. Then adjust for draw, travel distance, backline provision, and who is actually selling the room. Door splits, guarantees, and ticketed vs free-entry nights change the arithmetic — there is no universal rule. For a ticketed 120-cap room at £8 door, gross door is £960; a 70/30 venue/artist split puts £672 in the venue’s pocket and £288 to the act — and that’s before bar revenue.
Cross-reference fees against the GX Rate Calculator and the GX Index to check you’re in the right range. The Musicians’ Union minimum is £167.16 per musician for up to 3 hours — paying below that is your call, but good acts know the number and the ones worth rebooking expect at least that.
A residency ladder is the single best programming tool most UK pubs never use. The concept: an act starts with a monthly slot. If audience grows over 3 months, they move to fortnightly. If it sustains for another 3 months, they move to weekly. If it dips at any rung, they stay where they are or step back down.
A reliable residency act at a 100-cap pub playing 4 times a month at £200 per booking costs £800/month and generates £2,400-£3,600 in bar revenue (assuming 45-60 people per night at £13-£15 average spend per head). That’s a 3:1 to 4.5:1 return on the act fee — and the consistency builds a regular audience that comes for the night, not just the act. Over 12 months, a strong weekly residency generates £28,800-£43,200 in bar revenue from £9,600 in act fees.
Here are my last 6 months of bookings at [VENUE NAME, CITY, CAPACITY]: [PASTE YOUR BOOKING DATA — date, act, fee, door count, bar uplift estimate, notes] Identify acts that show residency potential. For each candidate: 1. Average door count across bookings (consistent or growing?). 2. Cancellation rate (how many no-shows or late cancellations?). 3. Bar-revenue correlation (do their audiences buy drinks?). 4. Current fee vs recommended residency fee (10-15% discount is standard — the act saves on marketing and gets guaranteed income). 5. Recommended ladder step: monthly / fortnightly / weekly. 6. Risk flag if any (one big night then nothing, audience that doesn't spend, acts that bring their own crowd once but never again). Cross-reference fees with GigXchange Rate Index at gigxchange.app/rates for fair residency pricing in [CITY].
A residency discount of 10-15% on the standard fee is the UK norm. On a £250 standard booking fee, that’s £212-£225 per residency slot — the act earns less per gig but gains guaranteed income and zero marketing overhead. Both sides benefit. Check the GX Index to confirm your base rate is fair before applying the discount.
The rebook decision is the one most bookers get wrong because they rely on a single metric: door count. But 80 people with low spend vs 45 people with strong spend and low staff friction — door count alone is misleading.
For each act, score 3 dimensions on a 1-5 scale:
| Dimension | What it measures | Weight |
|---|---|---|
| Revenue per booking | Door + bar uplift minus act fee. A £500 act that generates £1,200 in door + bar scores higher than a £200 act that generates £400. | 50% |
| Reliability | Shows up, plays the agreed set, professional with staff and sound engineer, no last-minute changes. 5 = never caused a problem across 6+ bookings. | 30% |
| Audience fit | Do their punters match your regulars? Do they come back when the act isn’t playing? A 5 means the act builds your audience; a 1 means they bring a one-time crowd that never returns. | 20% |
A weighted score of 4.0+ is a rebook. 3.0-3.9 is a maybe — check if the weak dimension is fixable (e.g. reliability issues might be a one-off). Below 3.0 is a polite no. Feed your booking data to AI and ask it to calculate the score for each act — it will do the maths in 15 seconds across 20 acts that would take you 45 minutes by hand.
The act that brings 100 people once but 20 the second time. The act that draws a big crowd but the crowd doesn’t buy drinks. The act whose audience gets into fights or causes noise complaints that put your 11pm licence condition at risk. The act that cancels 2 out of 5 bookings. None of these show up in a headline door count — they show up in the notes column of your tracking spreadsheet, and that’s exactly what the rebook-score prompt catches when you feed it your full data.
Here is the step-by-step session. Run it on the first Monday of every month. Block 60 minutes in your diary. Do not multitask.
| Step | Time | What you do |
|---|---|---|
| 1. Pull data | 5 minutes | Export or copy last month’s booking spreadsheet (8 columns above). If you don’t have one, start one now — even 4-6 rows from memory is better than nothing. |
| 2. Feed to AI | 5 minutes | Paste the data into ChatGPT or Claude with the combined prompt below. Let it process. |
| 3. Review AI output | 20 minutes | Read the draft calendar, rebook recommendations, residency candidates, and themed-night suggestions. Mark each as accept / reject / discuss. |
| 4. Apply local knowledge | 15 minutes | Filter through: acts you know personally, relationships, sound limits, local politics, upcoming events in town that change the audience mix, bank holidays, seasonal patterns. |
| 5. Book | 15 minutes | Book confirmed slots via GigXchange or direct contact. Send polite holds to maybe-list acts. File the review output for next month’s comparison. |
You are a UK grassroots venue programmer. The venue is [NAME, CITY, CAPACITY]. Live music runs [DAYS PER WEEK]. Current regular nights: [LIST]. Act budget: £[RANGE] per night. Here is last month's booking data: [PASTE YOUR 8-COLUMN SPREADSHEET DATA] Give me: 1. REBOOK LIST — score each act on revenue-per-booking (50%), reliability (30%), audience-fit (20%). Recommend: rebook / maybe / pass. For rebook acts, suggest fee and frequency. 2. RESIDENCY CANDIDATES — any act with 3+ bookings showing consistent or growing draw. Suggest ladder step (monthly / fortnightly / weekly) and a fair residency fee (10-15% below standard, cross-referenced with UK rates at gigxchange.app/rates). 3. WEAK NIGHTS — identify nights where door count or bar uplift is trending down. For each, suggest 2 format changes to test. 4. DRAFT CALENDAR — build next month's programming using rebook acts + 1-2 new suggestions for weak nights. Include fees and format for each slot. 5. SAFETY FLAGS — anything I should watch (acts with declining draw, over-reliance on one act, nights with no backup plan). Be specific. Use the data. Don't suggest things the data doesn't support.
The output from this prompt is a 2-3 page document you can review in 20 minutes. It replaces the 3-4 hours of scattered instinct-based decisions most bookers spread across the month. Anthropic’s productivity research finds Claude speeds up individual knowledge-work tasks by ~80% on average — this prompt is a textbook example of that kind of structured task.
Five rules to keep the AI-assisted review honest:
The 60-minute review gives you a plan. These tools help you execute it:
Running a monthly programming review you’d improve, or a themed-night concept that worked in a room under 150 capacity? We refresh this post once a year and rely on UK venue feedback to keep it current. Last refreshed at the date stamped above.
Join artists and venues on the UK’s peer-to-peer live music marketplace.