A wedding planner can use AI at each hand-off in the process: summarising the enquiry and consultation call, drafting the proposal, turning questionnaires into a design brief, comparing supplier quotes, chasing RSVPs and confirmations, then building the final timeline from confirmed supplier times. In the illustrative case below, admin per wedding falls from 54 hours to about 34.
The planner, her tools and every figure here are illustrative, built from typical tasks rather than a real client. What carries over to your business is the pattern: AI drafts and organises, the planner decides, and nothing involving allergies, money or times reaches a couple or a supplier without a person reading it.
The planner, before any AI
She works alone with a part-time assistant two days a week and takes 16 full-planning weddings a year, most 12 to 18 months from enquiry to wedding day. Design meetings and the days themselves weren't the problem. The problem was admin: writing, chasing, reformatting and re-checking. She timed one wedding's admin, stage by stage, before changing anything:
| Stage | Admin hours before |
|---|---|
| Enquiry to consultation brief | 3 |
| Proposal and contract | 4 |
| Design brief and supplier shortlist | 12 |
| Budget, guests and dietary list | 15 |
| Supplier confirmations | 8 |
| Final timeline and supplier pack | 10 |
| After the wedding | 2 |
| Total | 54 |
What she set up first, and what it cost
She spent two evenings on set-up before the first couple went through the new process:
- ChatGPT Plus at $20 a month (Claude Pro at the same list price would do the same job), with one project per couple so each wedding's details stay together.
- Google Workspace Business Standard, about $14 a month on an annual plan, which she already had, for Gemini in Gmail, Docs and Sheets.
- Otter.ai Pro at $8.33 a month billed annually, for transcribing consultation and design calls when the couple agrees to recording.
- Her existing planning CRM for contracts, invoices and the client portal. No change there.
- A master instructions file in each project: her tone notes, her standard timeline format, her supplier roster, and a short list of things AI must never do (confirm a supplier, state a price, merge dietary needs).
New spend: under $30 a month. The bigger cost was the two evenings, and the first three weddings, which ran slower while she built the prompts.
Month 0: from enquiry to a consultation brief
An enquiry arrives: a couple, a date next September, a venue already booked, 110 guests, a note that they're "not very organised". She checks her diary first (a job AI never does for her), then asks the assistant to summarise the enquiry and draft a reply offering three consultation slots. Before the call, she pastes the enquiry and what she knows of the venue into a prompt that produces a one-page prep sheet: likely questions, gaps in what the couple said, and anything about the venue she should confirm rather than assume. The model had suggested the venue "offers in-house catering". It didn't; she checked, and that became her first question on the call.
The consultation is recorded with the couple's agreement. She's careful here: whether AI note-takers are safe for client calls depends on where the recording goes and who can see it, and couples talk about family money on these calls. The transcript goes into this prompt:
From this consultation transcript, fill in:
COUPLE | DATE | VENUE | GUESTS | BUDGET (stated, and how firm)
PRIORITIES (their words) | NON-NEGOTIABLES | WORRIES
FAMILY DYNAMICS to be aware of (facts only, no speculation)
DECISIONS ALREADY MADE | DECISIONS STILL OPEN
QUESTIONS I PROMISED TO ANSWER
Quote the couple directly where you can. Leave a field blank
rather than guess.An illustrative slice of what came back for the "not very organised" couple:
PRIORITIES: "Good food, and people dancing, not standing around." NON-NEGOTIABLES: ceremony outside if dry; grandmother (wheelchair user) seated near the front. WORRIES: "We'll forget something important." FAMILY DYNAMICS: the bride's mother may be difficult about the menu. DECISIONS STILL OPEN: band or DJ; whether children are invited. QUESTIONS I PROMISED TO ANSWER: does the venue allow a late bar; what a wet-weather ceremony looks like.
One field breaks the prompt's own rule. In the transcript, the bride said "Mum has strong views on the menu". "May be difficult" is the model's reading of it, and it's the sort of line that could end up forwarded to a caterer. She changed it to the quote itself and added "never characterise family members; quote them or leave it blank" to her master instructions. Everything else went into the couple's project as the first page of the file.
Time: 3 hours before, 1 hour 45 minutes after. The saving is mostly in not rewriting her notes.
Month 1: proposal and contract
The brief feeds the proposal. She runs the four-prompt sequence from writing event proposals with AI: a gap check, two concept directions, a draft with her standard sections copied word for word, and a pass that lists every claim to check. Pricing comes from her own sheet. The contract and invoice go through her CRM as before.
One thing she learned: the concept prompt kept offering "rustic elegance" for any venue with exposed beams. She added a line to her master instructions listing five overused phrases, and the drafts got noticeably less generic. Before the rule, a concept opened: "A celebration of rustic elegance, where timeless charm meets modern romance beneath the barn's exposed beams." After it, for the same couple: "Long tables down the barn, food served to share so people talk, and the band starting straight after the speeches, because you said you want people dancing, not standing around." The second came from the brief's own words, which is what the banned list pushed the model back towards. Time: 4 hours before, 2 hours 15 minutes after.
Months 2 to 4: design brief and supplier shortlist
This is where the biggest block of hours used to go. The couple's questionnaire runs to 40 questions, plus a mood board and voice notes sent at midnight. She pastes the lot into the couple's project and asks for a two-page design brief in her usual headings: colour, flowers, atmosphere, must-have moments, things they dislike. She rewrites about a third of it, which is still faster than starting from nothing.
For suppliers, the assistant drafts enquiry emails to each shortlisted photographer, florist, band and caterer, each one carrying the couple's specifics so suppliers can quote properly first time. When quotes come back, it lines them up in a comparison table. The method in comparing supplier quotes side by side with AI is what she follows.
The enquiry drafts needed one standing edit. An illustrative first draft to a florist began: "I'm planning a September wedding for 110 guests at a barn venue, with a total budget of $38,000..." A florist told the whole-wedding budget will quote towards the top of what that figure allows. She now gives each supplier only their own line: "Flowers budget around $3,500, covering the bridal party, ceremony and 12 long-table arrangements", plus the colour notes from the design brief and the venue's rules on candles and fixing to beams. Suppliers quote to the brief they're given, so the brief should be theirs alone.
What went wrong: the comparison listed one photographer's quote as including a second shooter. It was an optional add-on on page two of the PDF. She now asks for a column called "Included or add-on?" with the page number for every line, and checks any cell that says "included". Time: 12 hours before, 8 after.
Months 5 to 9: budget, guests and the dietary list
The budget lives in a Google Sheet built on the layout in building event budgets and run sheets with AI, with Gemini in Sheets writing the formulas and a status column that shows which lines are quotes and which are guesses. The assistant drafts the monthly budget update for the couple from the sheet's figures, and she checks every number in the email against the sheet.
That check earns its keep. An illustrative monthly update drafted by the assistant said the couple were "$1,200 under budget overall". The sheet agreed only if you counted the stationery line at its $600 estimate and the late-bar extension as not needed; with the stationer's actual quote of $1,050 and the $600 bar extension the couple had asked about, the margin was $150. She rewrote the line as "about $150 under budget, with stationery now quoted and the late bar still to decide", which is less cheerful and far more useful to a couple deciding whether to add a photo booth.
Guest logistics use the same project: RSVP reminder messages in three tones (gentle, second nudge, final date), a summary of who hasn't replied, and a consolidated dietary table for the caterer from the RSVP form.
The dietary table produced the season's most serious near-miss. Asked to "tidy" the list, the assistant grouped "vegetarian" and "pescatarian" together and shortened "severe nut allergy, carries an adrenaline pen" to "nut-free". She rewrote the prompt so it may only group identical entries, must copy anything containing "allergy", "severe", "pen" or "coeliac" word for word, and must list the guest's name against every allergy line. She also checks the final dietary table line by line against the RSVP export herself, every time. Time: 15 hours before, 10 after.
Months 10 and 11: supplier confirmations
Eight weeks out, every supplier gets a confirmation email: date, venue address, arrival time, contact on the day, what they've been booked for, and the questions she needs answered (set-up time, power, parking, meals). The assistant drafts one per supplier from the master sheet. When replies come back, she pastes them in and asks for the arrival, set-up and finish times as a table, which goes straight into the timeline sheet. Any supplier whose reply doesn't answer every question gets flagged for a phone call.
The table lost something the first time. The band's reply said: "We'll arrive around 5, need 90 minutes to set up, and we can't set up during the speeches because of the noise." The extracted row read "Band: arrive 17:00, set-up 90 min", which was accurate and incomplete, since speeches were scheduled for 17:45. She added a column called "Conditions" to the extraction prompt, with the instruction to copy any "can't", "must", "only" or "before" sentence word for word, and moved the band's arrival to 16:00. Time: 8 hours before, 5 after.
The last four weeks: building the final timeline
The wedding-day timeline is built backwards from the anchors: ceremony at 14:00, the venue's access time, the caterer's service windows, the band's contracted finish, the curfew. She gives the assistant only confirmed times and her standard columns (time, what happens, where, who's responsible, cue, contact) and asks it to fill the gaps with at least ten minutes of buffer before each anchor.
Hair and make-up is the classic example of where AI helps with layout but not facts. The make-up artist's own schedule said six faces at 45 minutes each, bride last, plus a 30-minute buffer: a 07:45 start to finish at 12:45. The assistant laid that out correctly because she gave it the artist's numbers; asked to estimate on its own earlier in the season, it had allowed 30 minutes a face.
Two more catches. The first draft put couple portraits at "golden hour, about 20:30"; the actual sunset on the wedding date was more than an hour earlier, which she confirmed with a sunset calculator for the venue. Models don't reliably know sunset times for a given date and place, so she now enters sunset herself as an anchor. And the stress-test prompt ("if the ceremony starts 20 minutes late, what breaks?") showed that the drinks reception would overrun the photographer's hours.
From the master, the assistant produces a one-page version for each supplier and a friendlier couple's version without supplier phone numbers and set-up notes. Time: 10 hours before, 6 after.
After the wedding
A thank-you email to the couple drafted from her own notes on the day, a review request two weeks later, and short thank-you notes to each supplier with one specific compliment, which suppliers remember when they're asked to recommend a planner. Time: 2 hours before, 1 after.
The season's numbers
| Stage | Before | After |
|---|---|---|
| Enquiry to consultation brief | 3h | 1h 45m |
| Proposal and contract | 4h | 2h 15m |
| Design brief and supplier shortlist | 12h | 8h |
| Budget, guests and dietary list | 15h | 10h |
| Supplier confirmations | 8h | 5h |
| Final timeline and supplier pack | 10h | 6h |
| After the wedding | 2h | 1h |
| Total per wedding | 54h | 34h |
Twenty hours per wedding across 16 weddings is about 320 hours a season, against well under $400 a year in new subscriptions. She chose to spend some of it taking two more weddings and some of it not working Sunday evenings. The checking time didn't fall, and she didn't want it to: every stage above still ends with her reading the output.
What she'd tell another planner
- Structure beats clever prompts. The same fields every time (brief, quotes, confirmations) did more than any prompt trick.
- Keep a wrong-once list. Every error in this story became a rule in her master instructions: include-or-add-on columns, allergy lines copied word for word, sunset entered by hand. The list is now 14 lines long and the errors haven't recurred. The first few lines read: "1. Quote comparisons: include-or-add-on column with page number. 2. Dietary lists: group identical entries only; copy allergy lines word for word with the guest's name. 3. Sunset is entered by me, never estimated. 4. Never characterise family members. 5. Supplier replies: copy every condition sentence into Conditions."
- AI is never the last reader of anything with allergies, money or times in it. Setting up human review without slowing down explains how to make that fast rather than a bottleneck.
- Start where the hours are. For her it was suppliers and guests, not proposals. Time one wedding before you choose.
If you're not sure where your own hours go, start smaller than this planner did: which event planning tasks to hand to AI first will help you choose a starting point, so you can pick one stage, run it for three weddings and only then add the next.
Further reads
- How Wedding Venues Use AI for Viewings, Enquiries and Follow-Ups — The venue side of the same couple's journey.
- How Wedding Suppliers Can Reply to Enquiries in Minutes With AI — A faster first reply, the stage before this workflow starts.
- Automate Client Onboarding With AI: Forms, Contracts, and Emails — Automate the contract, questionnaire and welcome pack.
- AI Meeting Note-Takers Compared for Small Teams — Pick a note-taker for consultation and design calls.
- How to Turn Meeting Notes Into Tasks Automatically With AI — Turn every call into a task list automatically.
- AI Error Log: Track Mistakes and Stop Them Happening Again — Keep the wrong-once list this planner relies on.
- How to Organise Shared Files So AI Tools Can Use Them — Three afternoons to get a shared drive ready for Copilot, Gemini or ChatGPT, with a wedding planner's folder tree and a ten-question test.
- Best AI Tools for Small Event Planning Businesses — Eight AI tools for small event planning businesses, ranked by the hours they give back, with prices, first uses and the two to handle with care.
- Can AI Qualify Wedding Leads by Date, Budget and Guest Count? — How AI reads messy wedding enquiries, the five ways dates and budgets get misread, three set-ups by effort, and a back-test on last season's leads.
- AI Mistakes That Put Couples Off Your Wedding Business — Eight ways AI makes a wedding supplier look careless or fake to couples, each with an example from real-looking enquiries and the fix.
- ClickUp vs Asana vs Monday: Which AI Features Save Time? — All three now charge for AI differently. A timed photography-studio example shows which features save hours and what each costs a five-person team.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: OpenAI and Anthropic plan pages, Google Workspace pricing page, Otter.ai pricing page (figures checked September 2026).