Yes, AI can write it, but segmentation and permission matter more than the writing. Split past enquiries by enquiry date and outcome, send each group a short seasonal update that shows dates you really still have and one clear next step, and email only people who gave permission, with an unsubscribe link in every send.
The catch is that a past enquiry is not a subscriber. A couple who asked for a quote 14 months ago didn't necessarily agree to a newsletter, and many of them married elsewhere last summer. So the real work happens before any writing: sort the list by how old each enquiry is and what became of it, take out anyone who booked or opted out, and check what the rest actually agreed to.
AI earns its place at three points: turning your calendar export into a plain list of open dates by season, drafting each segment's email from a facts sheet, and giving each draft a second read against those facts. It doesn't know your diary, and it doesn't know who gave permission. You supply both: the diary as a fresh export on the morning of each send, and permission as a column in your records.
If your list came from a wedding fair, the three follow-up emails in the fair stand and follow-ups tutorial come first, and this newsletter picks up afterwards. Everything here assumes a spreadsheet or CRM export with an enquiry date, the wedding date they asked about, an outcome and a permission flag.
Sort 640 enquiries into groups you can actually write to
The venue in this example has 640 enquiries in its records over three years (illustrative figures). Export them with six columns: enquiry date, wedding date asked about, outcome, permission recorded (yes or no), unsubscribed (yes or no) and, if you note it, the reason they didn't book. Then count each enquiry once, at the first group below that fits it, so the groups add up to the export.
| Group, in counting order | Illustrative count | What happens to them |
|---|---|---|
| Unsubscribed, bounced, or asked not to be contacted | 47 | Suppression list: never emailed, in any tool |
| Booked a wedding with the venue | 118 | Left out: they get client emails, not this newsletter |
| No permission recorded | 160 | Held out until permission is confirmed or an adviser says otherwise |
| The wedding date they asked about has passed | 74 | Left out of wedding emails |
| Emailable | 241 | Split by enquiry date, below |
| Total | 640 |
Unsubscribes come first in the order on purpose. A client who once opted out of your marketing is still opted out, and counting them under "booked" would hide that. The headline is the gap between 640 and 241. A venue that mails all 640 sends roughly three in five of them (399 people) something they asked not to receive, never agreed to or can't use.
Now split the 241 by how recently they enquired. The right-hand column is an assumption about what each group needs, and you should test it against your own records.
| Segment | Enquired | Count | What they probably need |
|---|---|---|---|
| A | In the last 90 days | 52 | Still comparing venues: current availability and an easy viewing |
| B | 91 days to 12 months ago | 121 | Have narrowed down or drifted: a reason to look again |
| C | More than 12 months ago, wedding date still ahead | 68 | Long-lead or postponed plans: next year's calendar and what has changed |
Pick the cut-offs from your own bookings
Ninety days and twelve months are placeholders. Your own bookings tell you better. Export the weddings you won with two columns, enquiry date and booking date, and delete names and contact details first, because the assistant needs dates and nothing else. Claude and ChatGPT both read uploaded spreadsheets.
This CSV has one row per booked wedding: enquiry_date and booking_date.
Calculate the number of days between them for each row. Give me the median, and the number of days
within which 80% of bookings were made. Flag any row with a missing date or a booking date
before the enquiry date, and leave those rows out of the calculation. Show the working.
If it reports a median of 41 days and 80% booked within 96 days, that would support a 90-day "recent" window for segment A, and it would tell you that an enquiry older than about three months is probably no longer being actively decided. Treat those figures as an example of the shape of the answer, not as numbers to copy.
If you also record why couples didn't book, add a second cut. A couple who lost out because their date was taken is the best audience you have on the day a cancellation frees it. A couple who found you over budget is the audience for midweek dates and for a revised minimum spend, which the tutorial on peak, midweek and minimum-spend pricing explains. Skip this cut if your records only say "lost", because the assistant can't recover a reason nobody wrote down.
For help turning a raw export into labelled groups, segmenting an email list with AI walks through fixed labels and evidence checks. Record the outcome of every enquiry from now on, even if it is only "booked", "lost" or "no reply", so next year's segments are easier.
Settle permission and unsubscribe rules before the first send
This is a practical outline, not legal advice. The rules on marketing emails differ from place to place, and data-protection law also covers the names and addresses on the list, so treat the table below as the cautious default and let an adviser loosen it if your rules allow. Sort everyone on the list into one of three kinds of record.
| What the record shows | How I'd treat it |
|---|---|
| Ticked a box for updates, with the wording and date saved | Include them, and keep the record as evidence |
| Enquired, but no box and no mention of updates | Hold out of the newsletter until an adviser has looked at how you collected the details |
| Unsubscribed, bounced, or asked you to stop | Suppress permanently and carry that list across every tool you use |
The middle row is where venues get into difficulty, so here is why it is held out. Rules differ on what a bare enquiry allows. Some treat it as enough for follow-up emails about similar services, usually only if the couple was given a clear way to opt out when they enquired; others want a recorded opt-in before any marketing at all. A recorded opt-in is the stronger position, though even that can lapse once it no longer reflects what the person wants. At least one data-protection regulator has also fined businesses for emails that asked people whether they wanted to hear from them, treating the question itself as marketing, so a re-permission campaign to people who never opted in is not automatically safe. The older the enquiry, the weaker the case that the couple expects to hear from you. Ask a data-protection adviser how the rules that apply to you treat this group, and keep it out of the send until you have an answer.
For the people you can email, every message should carry the same few things: who you are, a postal address or other contact details, a line saying why they are receiving it, and an unsubscribe link that works in one click without a login. Most email tools add the link and the address automatically, but check them in the preview. Gmail's sender guidelines require one-click unsubscribe from senders of more than about 5,000 emails a day to Gmail accounts and tell senders to keep spam complaints under 0.10%. A venue won't send at that volume, but the same habits protect your reputation at any size. Gmail also clips messages over 102KB, which can push the footer out of sight, so keep the newsletter short and plain.
A filled-in footer
You are receiving this because you enquired about a wedding at [Venue name] and ticked the box for updates.
[Venue name], [postal address]. Reply to this email to reach [your name] directly.
Unsubscribe: [one-click link]
Going forward, add a permission line to the enquiry form itself, unticked by default: "Yes, send me occasional updates about open dates and seasonal events. You can unsubscribe at any time." When you move email tools, export your unsubscribes first and import them as suppressed before you import anyone else. Email tools typically ask you to confirm you have permission for every contact you upload, and an old list pushed into a new tool is the commonest way unsubscribed couples get emailed again. The consent and unsubscribe settings checklist and the footer check for AI-written marketing emails cover the settings in more depth.
Turn the diary into a seasonal availability table
Availability is what a past enquiry most wants to know and what an assistant is least able to tell you, so feed it as data and verify what comes back. Export your calendar with date, weekday and status, mark provisional holds as "held", and remove couples' names. A held date isn't free: it can return to the diary when the hold lapses, but you can't promise it.
Here is our diary as a CSV: date, weekday, status (free, held or booked).
Treat "held" as not free. For the next six months, give me a table by month showing how many
Fridays, Saturdays, Sundays and Monday-to-Thursday dates are free. Under the table, list the exact
dates you counted for each cell, so I can check them.
The last instruction is the safeguard. Without the dates listed, you have a number you can't check, and counting long runs of dates is something assistants can get wrong. Here is an illustrative result.
| Month | Fridays free | Saturdays free | Sundays free | Midweek free |
|---|---|---|---|---|
| May | 3 | 0 | 4 | 14 |
| June | 2 | 1 | 4 | 12 |
| July | 1 | 0 | 3 | 11 |
| August | 2 | 1 | 3 | 13 |
Spot-check one month by hand. If the assistant is out by even one date in that month, recount the lot, or run it again three months at a time. Then write the availability line yourself from the table, in plain words: "May has three Fridays and four Sundays free. Saturdays are gone in May and July, with one left in each of June and August." Re-export on the morning of the send, because an availability table is only true on the day it was made.
Two segment emails, drafted and then corrected
Keep a short facts sheet next to the table: real changes since couples enquired (new accommodation, a revised midweek rate, a new wet-weather plan), the price guide's year, and your tone rules. Paste two newsletters you've sent and liked, so the assistant has your voice to copy. Then ask for one segment at a time.
Write a newsletter email for segment A (enquired in the last 90 days, not yet booked), sent in late winter.
Under 170 words. One theme: summer availability. Use only the facts sheet and the availability table below.
One next step: a viewing, using [VIEWING LINK]. Do not mention the couple's own wedding date.
No exclamation marks, no "special day", no urgency that the table doesn't support.
Give me three subject lines, each of which states something the email proves.
This is the sort of draft that comes back for segment A. It is illustrative, with the slips a first draft often has.
Subject: Summer at [venue]: where the diary stands
Hi [first name],
Our calendar is filling fast for summer, but we've held back a few of the best dates just for you. June still has two Fridays and four Sundays free, and the barn looks gorgeous in the evening light.
Couples often tell us that seeing the space makes the decision easy. Viewings take 45 minutes: [VIEWING LINK]
Warm wishes,
[your name]
Three lines need to go. "Filling fast" is a claim the table doesn't support: it shows two free Fridays and one free Saturday in June, which is neither fast nor slow, so state the numbers and let them speak. "Held back a few of the best dates just for you" invents a favour you haven't offered anyone and would then have to honour. "Couples often tell us" is a claim about what couples say that nobody has recorded, so cut it. What survives is short and true: "June has two Fridays and four Sundays free, and one Saturday on [date]. Viewings take 45 minutes: [VIEWING LINK]."
Segment C needs a different job. These are couples with a long-lead date who haven't heard from you in a year, so the opening has to say what has changed since they enquired. A good version reads: "You asked about [season] a while ago. Since then we've added [real change], published the price guide for [year] and opened the calendar to [month]. If your plans have moved, reply with your new date and I'll tell you whether it's free." Everything in the brackets is yours to supply, and the invitation to reply matters more than any link, because a reply starts a conversation you can follow up personally.
Subject lines that say what the email contains
| Weaker | Better, if true | Why |
|---|---|---|
| Still thinking about your wedding? | Fridays in June: where the diary stands | Names the content; doesn't nag |
| Don't miss out! | Two Fridays left in June | The number must match the table |
| Venue newsletter: spring | What has changed at [venue] since you enquired | Gives segment C a reason to open |
A second read: audit the claims, then check the send yourself
A second pass is cheap. Paste the draft, the facts sheet and the availability table into a fresh chat and ask for an audit instead of a rewrite.
The audit prompt
Audit this email. List every factual claim in it (dates, numbers, features, promises, claims about other couples).
For each, say whether it appears in the facts sheet or the availability table, and quote the line.
Mark anything you cannot find as UNSUPPORTED. Do not rewrite the email.
On the segment A draft above, a good audit returns the June availability as supported and marks "filling fast", "held back a few of the best dates" and "couples often tell us" as unsupported, which is the result you want. It won't catch what neither document contains, so the human checks still apply.
- Counts: the send list is 241 in total (52, 121 and 68 by segment), matching the table, with the suppression list applied.
- Dates: every number in the subject and body matches the table exported that morning.
- Footer: venue name, address, reason line and a working unsubscribe link are visible in the preview.
- Merge fields: a test to yourself and to an address on a different provider shows names, not raw tags.
- Links: the viewing link opens the right page on a phone, and shows only real slots.
- Length: under about 200 words, no more than one image.
If you have never sent a list of this size, a free plan is enough to learn on. At the last check Mailchimp's free plan allowed 250 contacts and 500 sends a month without scheduling or automation flows, and MailerLite's allowed 250 subscribers and 2,500 emails a month, so import only the emailable 241 and not the whole export. Plans change, so confirm the limits for your list size on the tool's own pricing page.
Four sends in a year, each with one job
| Send | Main segments | Theme | Diary check before sending | One next step |
|---|---|---|---|---|
| Late winter | A and B | Summer availability | Free Fridays, Saturdays and Sundays for the next four months | Book a viewing |
| Late spring | B and C | Autumn and winter weddings at the venue | Free dates from September to February | Reply with your date |
| Late summer | A, B and C | Next year's peak dates | Saturdays still free for next year | Book a viewing |
| Early winter | B and C | Midweek and off-peak dates, with the new price guide | Midweek and Sunday counts, new guide year | Download the guide |
Each send has one theme, one availability line and one next step; anything beyond that belongs on your website. The table uses seasons rather than months because your calendar should decide the rest: find the months when enquiries arrive at your venue and send about two weeks before them. Real-wedding stories can earn a place in the late-spring send about autumn and winter weddings, but only with the couple's permission for the photographs, which the real-wedding and consent tutorial covers.
Reading results without trusting open rates
Some mail apps load images automatically, which makes open counts unreliable, so don't build decisions on them. Count the things that need a human reaction instead: replies, viewings booked, unsubscribes and spam complaints. Here is an illustrative first send to all 241.
| Segment | Sent | Replies | Viewings booked | Unsubscribes |
|---|---|---|---|---|
| A | 52 | 6 | 4 | 0 |
| B | 121 | 5 | 3 | 2 |
| C | 68 | 3 | 2 | 1 |
| Total | 241 | 14 | 9 | 3 |
Read it by segment. Segment A turned 6 replies into 4 viewings, which is what a warm group should do. Segment B produced the unsubscribes: two from 121 is under 2%, but it is the group to watch, and if it rises on the next send, send B fewer emails before you change the wording. If the three viewings from segment B end in one booking worth a contribution of $5,000, that one wedding has paid many times over for a year of Mailchimp's entry paid plan (from $13 a month for up to 500 contacts at the last check). The $5,000 is an assumption, so substitute your own contribution per wedding and your own booking rate, and compare the cost of those viewings with what you pay for them from other channels.
What to do with the replies that come back
"We booked elsewhere." Reply with one line of thanks and congratulations, mark the enquiry lost with that reason, and take them off future sends. Don't try to win them back: the reply has given you the answer, and the reason code improves next year's segments.
"Is [date] free?" Check the live diary yourself, then write back. If the date is free, offer a viewing slot. If it has gone, offer the nearest free dates on the same weekday and say what a Friday or a Sunday would change. The assistant can draft the reply if you paste in the message and your list of free dates, but the sentence that says a date is free should always be yours.
"Please take me off your list", typed as a reply. That is an unsubscribe even though they didn't click the link. Add them to the suppression list the same day, send at most one line confirming it, and don't ask them to reconsider.
"How did you get my email?" Answer politely and plainly: they enquired on [date] and, if it is true, ticked the box for updates. Then remove them. If someone asks what data you hold about them or asks you to delete it, treat that as a formal request, because in many places it is one with a deadline, and pass it to your data-protection adviser the same day.
Out-of-office replies and bounces. Ignore the first. A hard bounce means the address doesn't work, so suppress it. Your email tool usually does this itself, but check the report after each send, because a list full of dead addresses can damage how your later emails are treated.
Slips that show up after the first send
A booked couple receives "still looking?" because their status was never updated in the records. Refresh outcomes the day before each send and rebuild the emailable list from the records, instead of reusing last time's file. A couple whose wedding date has passed receives an invitation to book a date that no longer exists; the "date has passed" group is there to stop that.
The availability line goes stale. A table exported on Monday and used on Thursday can mention a Friday that was booked on Tuesday, so re-export on the day. Then there is the old enquiry that gets a line about "your date", taken from the original enquiry. Leave their date out of the newsletter: you haven't checked it for them, and you may be about to disappoint them.
The unsubscribe list gets left behind when you change email tools, and unsubscribed couples reappear in the next send. Treat the suppression list as a document you own, in a file you can find, not as a setting buried inside one tool. Finally, a draft goes out with the assistant's warmth where yours should be. "We'd love to be part of your special day" fits every venue, so replace it with a line about the actual barn, the actual Fridays and the actual price guide, which is what a couple still comparing four venues will remember.
Questions venues ask before the first send
How often should a venue email past enquiries?
Four times a year is a sensible start: enough to stay familiar, few enough that each send can carry real news about dates or changes at the venue. Watch the unsubscribe count after each one. If a segment's unsubscribes jump, send less to that group before you try anything cleverer. Add a send only when you have genuine news.
Is it worth emailing couples who enquired over a year ago?
Only if the wedding date they asked about is still ahead and they gave permission. A couple planning a long-lead date, or whose plans were postponed, may welcome next year's calendar. Anyone whose date has passed should come off the wedding list, and could be offered anniversary or family events only if they agreed to hear about those.
Should the newsletter quote prices?
Link to the current price guide rather than printing figures in every email, and make sure the guide is for the year the couple is marrying. Prices quoted in an old email can come back to haunt you. If you do mention a figure, take it from your published guide, never from the assistant's memory.
Can AI send the newsletter automatically each season?
It can draft, and your email tool can schedule, but I'd keep a person between the two. The risky lines are availability and prices, which the assistant can't see. A five-minute check against the live diary before each send is cheaper than apologising to a couple whose date was already taken.
Further reads
- How Wedding Venues Use AI for Viewings, Enquiries and Follow-Ups — What happens once a past enquiry replies and wants a viewing.
- Can AI Qualify Wedding Leads by Date, Budget and Guest Count? — Score enquiries by date, budget and guest count.
- Email Newsletter QA Checklist Before You Hit Send — A pre-send checklist for any newsletter.
- Hotel Email Marketing With AI: Past-Guest Segments and Subject Lines — A similar past-guest segmentation, written for hotels.
- Wedding Venue CRM: Which Features Matter and Where AI Helps — Which CRM statuses and features are worth paying for.
- AI for Event Venues: From Enquiry to Quote Without Missing Details — The enquiry-to-quote workflow for an event venue: 14 fields, one reply that asks for what is missing, a calendar check, a rate-card quote and a hand-off brief.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Mailchimp and MailerLite pricing pages (free-plan limits), Gmail sender guidelines, and published guidance and enforcement summaries on how enquiries and consent requests are treated under direct-marketing rules, as checked in October 2026.