ChatGPT prompts UK venue bookers actually use12 prompts for triaging band pitches, drafting offers, writing gig listings and post-gig debriefs — copy, paste, edit.

UK venue bookers spend more time triaging unsolicited band pitches than they do programming the calendar. A grassroots venue with a weekly live-music night typically sees 30-80 pitch emails a month, of which 80% are wrong-fit and need polite rejections, 15% are maybe-worth-a-look, and 5% are real candidates. The rejections eat the time. The 5% need real attention.

Below: 12 prompts that compress that triage from hours to minutes, plus the editing rules that keep AI-drafted replies from sounding like a bot. All field-tested on real UK venue inboxes before publishing.

What ChatGPT actually does well for venue bookers

ChatGPT helps UK venue bookers by drafting tier-appropriate replies to artist pitches, summarising EPKs and social presence in 30 seconds, writing gig listings across platforms, generating offer emails with consistent terms, and capturing post-gig debrief notes — provided every prompt has a role, constraints, your variables, an output shape, and a list of phrases to avoid.

It’s strongest at repeat-format writing: the 50 polite rejections, the 12 monthly gig listings, the contract-style offer emails. These are the tasks that swallow a venue booker’s mornings. ChatGPT compresses them — provided you bring the inputs (your room’s capacity, your typical fee bands, your booking-tier rules) and you edit before sending.

It’s also useful as a second pair of eyes on EPKs: paste an artist’s pitch + their socials in, ask for the credibility signal in 100 words. Half the value is the summary, half is what gets flagged as suspicious (claimed draw vs social engagement, recent venues vs current activity).

What it gets wrong (and how we know)

The recurring failure modes when UK venue bookers use AI without guard rails:

• Bot-tone rejections. Default ChatGPT rejection emails read as “Hi, thank you for your submission. While we appreciate your interest, we are unable to accommodate at this time...” That sentence has flagged a venue as bot-replying within the first sentence. UK artists talk; this gets noticed.
• Hallucinated artist details. Asked to summarise an EPK, AI will sometimes invent press hits, festival slots, or streaming counts that aren’t in the source. Always work from text the artist actually sent.
• Generic gig listings. Default listings sound interchangeable across venues. Without a forbidden-phrases list and your venue’s actual character, every listing reads the same.
• Over-eager offers. Without explicit fee floors, AI drafts offers that under-shoot the market. Anchor every offer with a realistic UK fee floor (our UK rate calculator gives you a defendable range by city + gig type + band size).
• American venue assumptions. Default ChatGPT thinks “venue” means a 500-cap club with a green room. UK pub rooms with 80-cap and a corner stage need different rider language and different expectations.

Every prompt below has those failure modes designed out. Now the prompts. Each assumes you’ll fill in the {curly braces} with your venue’s real details before hitting send.

Prompt 1 — Reply: yes, follow up

Use when: an unsolicited pitch is genuinely promising and you want more info before deciding.

You are the booker at {venue_name}, a {capacity}-cap {pub | bar |
listening room | theatre} in {city}. Use only supplied facts; mark any
missing venue facts or date windows as [NEEDS FACT]. An artist has pitched. Write a 100-word warm-but-non-committal reply asking for:
(a) link to a recent live recording, (b) two recent UK venues they've
played, (c) realistic local draw in {city}, (d) settlement expectations
(door split, guarantee, flat fee). Output: subject line first, then body.
End with availability windows for {date_range_we_have_open}. UK formatting
(£, DD Month YYYY, 24h times). British English. No "team {venue_name}"
sign-off, no "rocking the stage", no exclamation marks.

Why this works: the three info requests are the shortest path to vetting fit before you commit calendar time. The “non-committal” instruction stops AI promising the slot before you’ve seen the recording.

Prompt 2 — Reply: graceful no

Use when: the pitch is wrong-fit (genre mismatch, no UK gigs, unrealistic ask) and you want to decline without burning a future relationship.

You are the booker at {venue_name}, declining a pitch from a {genre}
artist whose work doesn't fit our programming, currently focused on
{our_actual_focus}. Use only supplied facts. Do not invent contacts,
venues or opportunities. Write a 60-word polite no with subject line.
No "we'll keep you on file" unless we genuinely will. Suggested
alternative, if genuinely known: {specific_alternative_or_context}. If
this field is blank, give a general direction only (genre-appropriate
venue type), not a named venue. British English. UK formatting. No
"unfortunately", "at this time", "thank you for thinking of us".

Why this works: the suggested-alternative line is the trick. Artists remember venues that helped them find a better fit, even when declining. The forbidden-phrases line kills the corporate-bot tone that signals AI rejection within two sentences.

Prompt 3 — Gig listing for socials

Use when: a confirmed gig needs an announcement post across IG, FB, your website, and 2-3 listings sites.

You are the booker at {venue_name}. Use only supplied facts; do not
invent support acts, timings or access details. Write a gig listing for
{venue_name} hosting {artist_name} ({genre}) on {date} at {time}.
Tickets: £{advance} advance / £{door} door. Ticket link: {ticket_url}.
Age / accessibility notes: {age_limit_access_info}. Support:
{support_act_or_none}. Doors: {door_time}. Output 4 versions:
(a) Instagram caption 80-100 words; (b) Facebook event description
60-80 words; (c) website listing 40-60 words; (d) one-line for Skiddle /
Songkick / Bandsintown (140 chars max). Lean into what makes {venue_name}
distinctive. Venue distinctive fact to use:
{sound_system | room_style | location | programming_note}. If blank,
omit venue-distinctive claims. UK formatting (£, DD Month YYYY, 24h
times). British English. No exclamation marks, no "don't miss out", no
"the place to be".

Why this works: generating all 4 platform variants in one pass keeps voice consistent. The “no don’t miss out” constraint kills the cliché-template tone that local audiences ignore.

Prompt 4 — Flyer caption for a regular night

Use when: you run a weekly or monthly residency / open mic / themed night and need fresh copy each week.

You are the booker at {venue_name}. Use only supplied facts. Write 4
weekly flyer captions for {regular_night_name} at {venue_name}
({open_mic | residency | themed_night | quiz_with_music}). Each caption
60-80 words. Lead with this week's twist: {week_1: theme_or_act},
{week_2: theme_or_act}, {week_3: theme_or_act}, {week_4: theme_or_act}.
Date + start time per week: {week_1_date_time}, {week_2_date_time},
{week_3_date_time}, {week_4_date_time}. If any weekly twist or
date/time is blank, write [NEEDS FACT] for that caption rather than
inventing one. Voice: {venue_brand_voice: cosy/sharp/sardonic/warm}. Each caption ends
with start time + entry: {free | £{fee}} for that specific week. If
the week is free, use "free entry"; if £fee, use "£{fee} on the door"
or similar. UK formatting (£, DD Month YYYY, 24h times). British
English. No exclamation marks, no "the place to be".

Why this works: regular-night captions burn out fast because the core info is identical week-to-week. Forcing AI to lead with the twist (this week’s act, theme, or hook) gives each caption something new.

Prompt 5 — Vet an artist’s EPK in 30 seconds

Use when: you’ve had a pitch and need to fast-vet whether it’s worth a real read.

You are an EPK triage assistant for the booker at {venue_name}. Use only
the pasted evidence below; if no evidence is provided, say "not enough
evidence" rather than estimating. Do not invent press hits, festival
slots or numbers. {paste_pitch_and_epk}. Output a 140-word bulleted
checklist covering: (a) genre + 2-3 closest reference acts inferred from
material; (b) confidence-rated draw notes based ONLY on pasted evidence:
strong / weak / missing; (c) claimed recent UK venues from the paste,
each marked [CLAIMED - VERIFY]; (d) red flags (stale socials,
inconsistent claims, no recent gigs). Mark all inferences as [INFERRED].

Why this works: the [INFERRED] tag stops AI from confidently making up facts to round out the summary. The “social engagement vs claimed numbers” line is the highest-signal vetting move — an artist claiming 10k mailing-list with 200 IG followers is suspicious; AI flags it.

Where it still fails: AI cannot phone the venues the artist claims to have played. That’s still the highest-signal vetting move and you do it for the 5% that pass this triage.

Prompt 6 — Draft an offer email

Use when: you’re ready to confirm and need to send the offer with fee, terms, asks.

You are the booker at {venue_name}. Use only supplied facts. Write an
offer email to {artist_name} for a gig on {date}. Fee: £{amount}
({flat | door_split | guarantee_plus_door}). Door split percentage (if
applicable): {split_pct}. Settlement detail (gross/net, ticketing fees,
PRS handling): {settlement_detail}. Slot:
{support | headline | residency_week_n}. Set length: {minutes}. Get-in:
{time}. Soundcheck: {time}. We provide:
{pa | backline | drum_kit | none}. We don't provide:
{green_room | meals | accommodation}. Settlement timing:
{after_gig | within_7_days | invoice_30}. Cancellation clause reference:
{clause_or_url}. Confirmation deadline: {date}. Output: subject line
first, then body. UK formatting (£, DD Month YYYY, 24h times). End with
a line confirming the offer is subject to contract and provisional hold
until written acceptance, with the formal contract following on
acceptance. Do not negotiate the fee in this draft.

Why this works: spelling out everything you provide AND don’t provide upfront kills the next-day “but I assumed you’d feed the band” email. The “do not negotiate” line keeps AI from softening your fee floor.

Prompt 7 — Reply to a no-show or late cancellation

Use when: an act has cancelled within 7 days, or no-showed, and you need to reply firmly without burning the bridge if it’s genuinely unavoidable.

You are the booker at {venue_name}. Use only supplied facts. Do not
make legal threats, admit liability, or invent consequences beyond the
supplied contract clause. Write a 90-word reply to {artist_name} who has
{cancelled within 7 days | no-showed} the gig on {date}. Their reason:
{their_reason_or_silence}. Our policy: {short_notice_clause_from_contract}.
Tone: firm, professional, no emotional language. Venue decision:
{close_door | leave_room_for_future_booking}. End according to the
Venue decision: (a) close the door, or (b) leave room for a future
booking. Output: subject line, body, then an internal note: "Before
sending, verify: contract clause, payment position, replacement costs."
UK formatting, British English.

Why this works: distinguishing “close the door” from “leave room for the future” up front prevents AI from defaulting to soft-no-conflict middle ground that signals you’ll book them again. State the policy, draw the line.

Prompt 8 — Draft a TEN application or licensing letter

Use when: you need a Temporary Event Notice, a noise-mitigation letter to a council officer, or a PRS / PPL paperwork response.

You are a UK venue-licensing assistant. Start your output with
"Document type: X" and "Confidence: high/medium/low" using these criteria:
high = official wording is pasted by user and applies cleanly;
medium = type is clear but some required wording missing;
low = type ambiguous or evidence absent. If confidence is low, ask for
the missing document context before drafting. First identify
which document type this is: TEN cover letter / noise mitigation response /
PRS reply / PPL reply. Use only supplied facts and any official wording the user
has pasted. Do not give legal advice. Do not cite regulation numbers,
fees or deadlines unless explicitly provided. Draft for {venue_name}, a
{capacity}-cap venue at {address}. Event/query: {event_or_query}.
Date(s): {dates}. Hours: {start_time} to {end_time}. Mitigation in place:
{sound_limiter | door_close_time | residential_distance | etc}. Tone:
factual, non-defensive. End with my contact: {name, email, phone}. Mark
required attachments as [ATTACH] and any official application, deadline
or statutory wording as [CHECK OFFICIAL FORM]. UK formatting, British
English ("licence" not "license").

Why this works: licensing officers respond to factual + non-defensive language. The “do not give legal advice” constraint is non-negotiable here — AI gets compliance details wrong often enough that you must run any output past a real licensing officer or solicitor before filing.

Prompt 9 — Residency proposal

Use when: you’ve worked with an act and want to offer a residency (weekly or monthly) rather than a one-off return.

You are the booker at {venue_name}. Use only supplied facts. Write a
200-word residency proposal to {artist_name}, who has played
{n_previous_gigs} gigs at our venue. Evidence of audience response:
{door_counts | bar_sales | sellouts | staff_notes | audience_feedback}.
Do not call response "strong" unless supported by this evidence.
Format: {weekly | monthly}, {n_dates} dates, fee £{amount} per slot
({flat | door_split}). If door split: percentage {split_pct}, settlement
timing and basis (gross/net, ticketing fees) {settlement_detail}. Our commitments:
{promo_commitment, mailing_list_features, dressing_room_perks}. Required
from artist: {minimum_set_length, original_material_requirement}.
Note any exclusivity terms (e.g. no competing local residency) at the
end as a contract-review flag rather than as a hard email demand.
Output: subject line, body, proposed first slot date. UK formatting (£,
DD Month YYYY). British English.

Why this works: residency offers fail when one side’s expectations weren’t spelled out. Listing both sides’ commitments forces clarity. The “no competing local residency” clause is the single most-overlooked variable that kills residencies later.

Prompt 10 — 4-week social plan around a confirmed gig

Use when: a gig is confirmed 4-6 weeks out and you need a structured social-promotion plan rather than ad-hoc posts.

You are the booker at {venue_name}. Use only supplied facts; do not
suggest platforms we do not use: {platforms_used}. Build a 4-week
social-promo plan for {artist_name} playing {venue_name} on {date}.
Ticket link: {ticket_url}. Key selling point for this gig:
{specific_reason_to_care}. If either is blank, mark [NEEDS FACT].
Output as a week-by-week table with: week, post type (announcement /
artist-feature / behind-scenes / week-of / day-of), platform, copy
length, draft copy angle (one line), CTA, asset needed (artist photo /
video clip / venue photo), owner, deadline. Lead with build-up, peak at
week-of, finish with day-of story-format. After the table, list
2 artist-asset requests I should send the act now. UK formatting (£, DD
Month YYYY). British English. No "drop everything", "the place to be",
exclamation marks.

Why this works: the asset-needed column is the trick — AI tells you upfront which photo or clip you’ll need to produce, so you can request it from the artist on confirmation rather than chasing the day before.

Prompt 11 — Mailing list email about upcoming nights

Use when: your venue has a regulars mailing list and you want a monthly or fortnightly email that builds attendance without spamming.

You are the booker at {venue_name}. Use only supplied facts. Write a
320-word regulars mailing-list email. Mark missing facts as [NEEDS
FACT]. If the lead-gig field {this_month_lead_gig} is blank, omit
section 2 entirely rather than inventing a pick. If {booker_note} is
blank, write [NEEDS FACT: booker note] in section 4. Output: subject
line, preview
text (50-90 chars), then body. Body structure: (1) opening referencing
this month / week / season, 1-2 sentences, no "hope this finds you
well"; (2) lead pick of the month: {this_month_lead_gig} with one-line
why; (3) supporting gigs from this list:
{date, act, ticket_link, one_line_why} x 3-4. If fewer are supplied,
include only supplied gigs and mark the gap [MISSING]; (4) personal-
from-the-booker line (30-50 words, fill in: {booker_note}); (5) sign-off
referencing regulars-only perk if any:
{early_door | priority_seating | members_drink | none}. If "none", use a
plain sign-off without inventing a perk. UK formatting (£, DD
Month YYYY). Tone: like a venue manager, not a marketing department.
British English. No "Team {venue_name}" sign-off.

Why this works: the “personal from the booker” line is the highest-engagement part of any venue email and the part AI can’t write for you. Give it the line in 1-2 sentences; let AI tighten it.

Prompt 12 — Post-gig debrief notes

Use when: you want to capture what worked / didn’t after every gig so future programming gets better, without spending 20 minutes writing notes.

You are an analytics assistant for the booker at {venue_name}. Turn the
filled-in notes below into a post-gig debrief — do not generate a blank
template. If a field is blank, write [MISSING] and do not invent it.
Separate factual notes from recommendation. Notes: artist={artist_name},
date={date}, door_count={door_count}, wet_sales=£{wet_sales},
timing_issues={timing_issues}, audience_reaction={audience_reaction},
sound_issues={sound_issues}, staff_comments={staff_comments}. Output
a 1-page markdown report: factual log, up to 3 evidenced wins, up to
2 evidenced issues (write [MISSING] where notes do not support them),
1 thing to do differently next time, yes/no/maybe rebook recommendation,
plus 3 booking signals to reference next time a similar act applies. UK
formatting (£, DD Month YYYY). British English. No corporate language.

Why this works: 90% of venues never write debrief notes because the format-from-scratch friction is too high. Having a fixed template that takes 60 seconds to fill in turns gig-night data into a programming feedback loop. Run this on every gig for a year and your booking judgement gets sharper.

How to make any prompt better — the 5-part formula

Every prompt above shares the same five parts. Skip any one and the output reverts to generic AI-listicle slop.

1. Role. “You are the booker at {venue_name}, a {capacity}-cap {type} in {city}.” Anchors voice and grounds the model in venue reality.
2. Constraints. Word count, tone, British English, language register. Without limits, AI writes long and generic.
3. Variables. The {curly_braces} your real details slot into. The more specific the inputs, the less generic the output.
4. Output shape. “4 versions: IG / FB / website / one-line”, “ordered table”, “1-page markdown report.” Defines the artefact you actually want.
5. Forbidden phrases. The single biggest quality lever. List the cliches you never want to see — “rocking the stage”, “the place to be”, “don’t miss out”, “at this time”.

If a prompt isn’t working, the missing piece is almost always part 5.

How we know these prompts work — the 5 tests every prompt passes

Every prompt above passes the same 5 quality tests before it ships. They’re worth knowing because once you’ve used the prompts here, you can apply the same tests to any prompt you write yourself.

1. It can’t make stuff up. Every prompt forces ChatGPT to use only the details you give it — leave a field blank and it writes [NEEDS FACT] rather than inventing artist credits, fee figures, claimed draw numbers or local venues that don’t exist.
2. It can’t damage your reputation. No legal threats in the no-show prompt, no admissions of liability, no fake urgency in gig listings, no inventing “strong audience response” without evidence. The compliance prompt explicitly refuses to give legal advice or cite regulations that haven’t been provided.
3. It produces a finished thing, not a fragment. Every email prompt outputs a subject line plus body. Every listing prompt produces ready-to-paste copy for each platform. The post-gig debrief fills in the template — it never gives you a blank form.
4. It sounds British, not American. No “rocking the stage”, no “the place to be”. £ and DD Month YYYY. 24-hour times. “Licence” not “license”. The forbidden-phrase list is task-specific so the output reads like a working UK venue manager wrote it.
5. It handles blanks gracefully. If you don’t fill in a field, the prompt tells ChatGPT exactly what to do — omit the line, mark it, or ask for it — rather than letting it improvise.

The test in plain English: would your venue be safe sending this without rewriting? If the answer was no, the prompt got reworked. The same 5 tests run again at every annual refresh.

Where ChatGPT stops and GigXchange begins

ChatGPT writes the email. It doesn’t find the right local artists, doesn’t know what they actually charge, doesn’t verify their recent gigs, and doesn’t see what other venues like yours are programming this month.

The natural complement: use ChatGPT for the writing, then use a real UK live-music marketplace for everything else.

• Find acts to book: our UK artist directory shows verified acts by city, genre and band size — the fastest way to spot artists who genuinely gig in your area.
• Anchor your offers: the GX rate calculator gives you a defendable fee range based on real UK rates by city, gig type and band size. Use the output as the {fee} variable in Prompts 6 and 9.
• See what others are programming: the live UK gig directory across 40+ cities shows what venues like yours have on this month.
• Compliance reality check: our plain-English guide to live music licensing covers PRS, PPL, the Live Music Act 2012 and TENs — useful before running Prompt 8.

Treat ChatGPT as a fast first-drafter for the volume tasks. Use the real-world tools for everything else.

If you find a prompt above that works particularly well for your room — or one you’d add — we’d genuinely like to know. UK venue bookers sharing what works is how everyone’s programme gets stronger. The prompts here will be updated whenever a better version emerges. Last refreshed at the date stamped above.