Campground Marketing With AI: Rebook Last Year's Guests by Email

Email only guests who have agreed to hear from you, split them by site type and holiday weekend, and send a six-email series from October to April: thanks, news, an early-booking window for returning guests, a same-weekend reminder, a last call and a gaps email. Give AI a fact sheet and merge fields, never the guest list.

Rebooking works because these guests already chose you. The email's job is to make the next decision easy and give them a reason to make it before the public can. Agency lists stop at "send an email". The parts that decide the result are consent, a true deadline, and an offer that does not give margin away to guests who were coming back anyway.

Follow me on Instagram@sagnikteaches

The examples follow a family campground with 60 sites (30 RV, 20 tent, 10 cabins) open May to September. All figures are illustrative, so swap in your own. Nothing here needs a developer: a reservation-system export, an email tool and an AI assistant are enough. If you are still choosing the reservation system, the comparison of RV campground software covers what each one exports.

Connect on LinkedInSagnik Bhattacharya

Start with the list: who you may email and who you may not

Last year's booking records are not automatically last year's mailing list. A guest who gave an email address to receive a booking confirmation agreed to the confirmation, which is a different thing from agreeing to marketing. If you market to guests in the EU, data-protection law such as the GDPR can apply even when your campground is based elsewhere, and the safest standard is a recorded opt-in. Elsewhere the rules differ, but most share a core: say who you are, include a working unsubscribe and honour it quickly. Emailing only people who asked to hear from you is the safe habit everywhere, even where the law allows more. Ask an adviser what applies to your guests. The table below is a practical sorting order, not legal advice.

Subscribe on YouTube@sagnikteaches

Run your export through that table. In the illustrative campground, last season's 1,420 booking records shrank like this:

• 1,420 booking records from last season.
• Minus 260 repeat bookings by the same guest = 1,160 unique guests.
• Minus 310 booked through marketplaces without permission to email = 850.
• Minus 190 with no marketing permission recorded = 660.
• Minus 24 bounced or invalid addresses = 636 guests to email.

That is 45% of the records. A smaller list of guests who agreed is worth more than a larger one that draws spam complaints. It also tells you what to fix: add a marketing tick box, unticked by default, to the booking form and the paper registration card, so next year's list starts bigger.

Clean and segment the list without pasting guests into a chat

Export only the fields the emails will use: first name, email, last stay month, site type, number of seasons stayed, booking source, and marketing permission with its date. Leave out any notes field. Notes tend to hold pet names, complaints and personal details that no email should quote. Clean the file in a spreadsheet or in the email tool itself: remove duplicates, drop the permission column's blanks, delete bounced addresses. If you want help, ask the AI to write the spreadsheet formulas for you, as the tutorial on segmenting an email list with AI explains, and keep the guest rows in your own systems.

The AI never needs to see a guest. It writes templates with merge fields, and the email tool fills in each guest's details when it sends. Three fields are enough for useful personalisation: [first name], [last stay month] and [site type]. Anything else an email says about a guest should come from those three, and nothing from free text.

The first three segments add up to the full 636 (280 + 190 + 166). The last two cut across them. Segments are for choosing a line or two of copy, not for building six different emails; with a small list, three variants is the practical limit.

The rebooking calendar: six emails from October to April

A season that runs May to September gives this shape. Move the months if yours runs differently. Only emails three to five ask for a booking; the rest keep the relationship warm so that the asking does not arrive cold.

The window is the point. "Returning guests choose from Thursday 14 January to Wednesday 20 January; public booking opens on Thursday 21 January" is a deadline that is real, so nothing in the series needs invented urgency. The October email matters more than it looks: guests still remember the stay, and the replies tell you what to fix before you ask for money. Email six pairs naturally with a shoulder-season campaign, which the caravan park shoulder-season tutorial covers in detail.

Turn the October replies into the December news

Email 1 invites replies, and the replies are free research. Copy them into a document, strip out names, email addresses and anything that identifies a guest, then ask the AI to group them. Do this only with the replies, never with the guest list. On a consumer plan, switch off the model-training setting first.

Below are guest replies to a thank-you email from a campground. Group them into themes with a count for each. Do not quote any reply and do not guess at anything a reply does not say. Then list the three themes that appear most often.

Say 52 replies come back and the AI returns: more shade (14), later checkout (11), quieter evenings (9), better Wi-Fi (8) and other (10). The counts add up to 52, which you can verify in ten seconds. Now choose one cheap improvement you will really make, and one request you will explain rather than grant, and put both in email 2. "You asked for more shade, so we have planted along the tent loop" is the sort of line that makes a returning guest feel heard. Put an improvement on the fact sheet only when you have committed to it, because email 2 will repeat it as a promise.

Choose the early-booking offer by what it costs you

A percentage discount looks generous and costs more than it seems, because everyone who was coming back anyway claims it too. Take the 636 guests. Suppose 15% rebooked in an earlier season with no campaign at all (to find your own figure, count how many of the previous season's guests booked again this season). That is about 95 guests who will use any early offer you make, whether it moves them or not. The campaign's aim is to add more, say 5 points, or about 32 extra rebookings. Say a typical stay is three nights at $48, so $144, and the campaign costs $199.50 (an email plan at $26.50 a month for three months, plus six hours of your time at $20).

The break-even column is the cost of the offer on the 95, plus the $199.50, divided by what each extra booking brings after the offer. For the discount that is ($1,368 + $199.50) / $129.60, which rounds up to 13. For the firewood perk it is ($475 + $199.50) / $139, about 4.85, so 5. For priority access alone it is $199.50 / $144, so 2. These are revenue figures, before your running costs, so use your own margin per night. The pattern holds whatever your numbers: a perk that costs you little and feels valuable needs far fewer extra bookings to pay off than a percentage discount. The 15% baseline is an assumption for the illustration; count your own before you plan around it.

Whatever you offer must be real and checkable. AI will cheerfully invent a "free night", a "welcome pack" or a price rise. The next section puts the offer in the one place the AI is allowed to read.

One fact sheet and one prompt for the whole series

Write the facts once, in a block the AI can read and may not go beyond. This is the single most useful thing you can do to keep the drafts honest.

CAMPGROUND FACT SHEET (the only facts you may use)
Name: [campground name]. 60 sites: 30 RV, 20 tent, 10 cabins.
Season: 1 May to 30 September.
Returning-guest window: Thursday 14 January to Wednesday 20 January.
Public booking opens: Thursday 21 January.
Offer in the window: a free firewood bundle on arrival (cost to us: $5).
Improvements this year: [list only real ones].
Policies: [deposit and cancellation rules in one line each].
Do not mention: prices, how many sites are left, amenities not listed here, guests' pets or any personal detail.

Write email 3 of a six-email rebooking series for past campground guests. Audience: guests who agreed to marketing emails. Use only the fact sheet above and these merge fields: [first name], [last stay month], [site type]. Voice: friendly, plain, a little outdoorsy, never pushy. No invented urgency. Under 150 words. Give a subject line and two alternatives, plus preview text under 60 characters. Add one different line for each site type (RV, tent, cabin) that uses only the fact sheet. End with the booking link placeholder [link], the footer placeholders [postal address] and [unsubscribe link], and nothing else.

Run the same pair for emails 1, 2, 4, 5 and 6, changing only the job and the date logic. Email 4 uses [last stay month] and [site type]; email 5 states the real end date; email 6 draws on a short list of open dates you type in by hand.

A sample early-booking email, and the edits it needs

Here is a typical first draft of email 3 (illustrative):

Subject: Your site for next summer: first pick opens Thursday
Preview text: Returning guests choose before the public does.
Hi [first name],
Thank you for staying with us last [last stay month]. We are opening next season to returning guests first. From Thursday 14 January to Wednesday 20 January you can book any [site type] before it goes on general sale, and every booking made that week includes a free firewood bundle on arrival.
If you have a particular weekend in mind, choosing early is the way to be sure of it: [link].
Questions? Just reply to this email.
[your name], [campground name], [postal address]
Unsubscribe: [unsubscribe link]

That draft stays inside the fact sheet, which is why it is short. Earlier tries without the sheet looked different, and these are the edits to watch for:

1. "Only 12 sites left!" appeared because the AI wanted urgency. Nobody gave it a count, so delete it. The window already supplies a true deadline.
2. "Come and enjoy our new splash pad" described something that does not exist. Anything not on the fact sheet goes.
3. "Book before prices rise in February" was invented. Delete it unless it is true, and if it is, put it on the sheet first.
4. It wrote 240 words. Halve it; a returning guest wants the dates, the perk and the link.
5. It added "we hope your dog enjoyed the trail". The AI had no pet data, so it guessed from the tone. Remove anything that implies you know more about the guest than the three merge fields.

The site-type lines need the same scrutiny. Asked for one line each for RV, tent and cabin guests, the AI returned something like this (illustrative):

• RV: "Full hook-up sites are waiting, so choose the same spot you had last time."
• Tent: "Bring the whole group: our shaded tent area is ready for another season."
• Cabin: "Cabins are the first to book each year, so get in early."

Every line has a problem. "Full hook-up" is not on the fact sheet, and some of your RV sites may have power only. Nothing is being held "waiting", and you cannot promise the same spot to everyone who asks. The shaded tent area is a guess, and "first to book each year" is a statistic the AI made up. Safe versions restate the sheet and nothing more: "Ask for the same site you had last time and we will do our best to match it, subject to availability", "Tent sites are open to returning guests in the same window", and "Cabins are open to returning guests in the same window too". A segment line may repeat a fact from the sheet. It may not add one.

Email 4 is where the merge fields do real work, and where their limits show. Here is a first draft (illustrative):

Subject: Same weekend as last [last stay month]?
Hi [first name],
You stayed with us in [last stay month] on a [site type]. If that weekend suits you again, the returning-guest window is open until Wednesday 20 January, and your site is being held for you until then: [link].
[your name], [campground name]
Unsubscribe: [unsubscribe link]

Two edits. Nothing in your system is holding a site for anyone, so "being held for you" must go; replace it with "you can ask for the same sort of site when you book". And "that weekend" promises more than the data holds, because the field is a month, not a weekend. Either add the arrival date to the export for these guests, so the email can say "around 4 July", or change the wording to "the same month". Catching a mismatch like that is exactly why you send a test to yourself first.

Read the result as a guest would, on a phone, then send yourself a test. Check that every merge field fills in, that the link goes to the right page, and that the footer shows your address and a working unsubscribe.

Subject lines and footers: details that decide if it lands

Ask the AI for five subject lines and choose the plainest. Compare these (illustrative): "BIG SAVINGS ON NEXT SEASON'S CAMPING!!!" against "Your site for next summer: first pick opens Thursday". The first reads like every promotion a spam filter has seen and tells the guest nothing about why it is for them. The second names the benefit, the guest's situation and the day it starts. Avoid capital-letter shouting, a string of exclamation marks and a fake "Re:" prefix, and tell the prompt to match your guests' own spelling and vocabulary.

Small details protect your sending reputation. Gmail's sender guidelines ask for a reported-spam rate under 0.10% and say to stay clear of 0.30%. On a list of 636, where perhaps 400 are Gmail users, one complaint is already 0.25% of those recipients, so every complaint deserves a look at whether that guest ever agreed. Gmail also clips messages over 102 KB, so keep images light and put the dates in text, not in a picture. Senders of more than 5,000 messages a day to Gmail accounts must support one-click unsubscribe; at your size, a clear unsubscribe link in the footer is the minimum, and an email tool adds it for you. Put your business name and postal address in the footer, and set up your sending domain in the email tool, which walks you through the records it needs.

Before the first send, run the draft through the email newsletter QA checklist. It catches the broken links and missing merge fields that make a campaign look careless.

Tracking bookings and handling the replies

You need to know which email produced which booking. Give emails 3, 4 and 5 their own tracked link or booking code (for example RETURN-A, RETURN-B and RETURN-C in your reservation system) so each booking carries the email that caused it. Do not rely on open rates. Some mail apps load images automatically, which inflates opens, and clicks and booking codes tell you far more.

Replies will arrive: "Can we move our dates?", "Is the pool open?", "Can we bring a second tent?" Keep a short list of answers (the fact sheet plus your policies) and let the AI draft replies from it, with one rule: if the answer is not in the sheet, the draft must say you will check and come back. You read and send every reply. When a guest says stop, remove them from the list the same day and note it on the suppression list.

Prices deserve a rule of their own. Link to your rates page rather than typing nightly rates into emails. One page that stays correct beats six emails that may not. If you are comparing email tools to run all this, the tutorial on email marketing tools and what a small list costs per month sets out the options. For orientation, list prices in USD at the last check were about $19 a month for up to 1,000 contacts on MailerLite's Comfort plan and $26.50 for 1,500 on Mailchimp Essentials, so confirm both on the vendors' pages.

Three numbers to count by the end of March

1. Rebooking rate against your baseline. Count how many of the 636 have booked by the end of March, and compare with the unprompted rate you worked out earlier. If the baseline was 95 guests (15%) and you reach 127 (20%), that is 32 extra rebookings, worth about $4,600 at $144 a stay, before the cost of the offer.
2. Bookings by email. Count bookings under each tracked link or code. If email 4 produced nothing, the problem may be the segment, not the copy.
3. Unsubscribes and complaints per send. Say 6 guests unsubscribed after email 3, which is 0.94% of 636. Compare each send with the last. A jump between two emails points at those emails, not at the list as a whole.

Then return to the break-even table with real figures. If you needed 5 extra bookings and got 9, the perk paid for itself comfortably. If you needed 5 and got 3, change the offer or the audience before next year, not the subject line. Set up costs about six hours the first time, split roughly into two for the list, two for the fact sheet and drafts, and two for building and testing the emails. After that each send takes about half an hour. A free ChatGPT, Claude or Gemini account can draft every email here, and the guest list never leaves your own systems.

Questions campground owners ask about rebooking emails

Can I email guests who booked through a marketplace?

Only if the marketplace's terms allow marketing use of guest details and the guest has agreed to hear from you. Many platforms restrict it, so check each one before importing anything. A better route is a small card at check-in inviting them to book direct next time, with a marketing tick box, so their next booking arrives with permission attached.

Is a paper registration card enough as consent?

It can be, if the card has a clear, separate tick box for marketing emails, left unticked, with your name and what you will send. Keep the cards or a scan, and enter the permission and date in your system. A card that only collects an email address for the booking does not show consent to marketing.

How many emails a year is too many?

For a seasonal business, six to eight a year to past guests is a sensible starting point, with most of the asking packed into one short window. Watch unsubscribes and spam complaints after each send. If either jumps, the fix is usually fewer emails or a clearer reason for each, not a new subject line.

Should I add texts to the series?

Texts need their own, separate consent, and they suit a different job: a last-minute gap or a weather change, not a six-month series. Start with email. Once guests opt in to texts at booking, keep messages short, identify the campground, and include a way to stop them. Marketing-text rules vary, so ask an adviser.

Further reads

• Holiday Let Website: Get Direct Bookings and Repeat Guests With AI — Turn marketplace guests into direct bookers who can join your list.
• Hotel Email Marketing With AI: Past-Guest Segments and Subject Lines — Past-guest segments and subject lines from the hotel side.
• How to Automate Pre-Arrival and Post-Stay Emails With AI — Automate the emails before and after each stay.
• How to Win Back Lapsed Customers With AI-Personalised Emails — Reach guests who stopped coming back, with a different tone.
• How to Build a Welcome Email Sequence With AI — Welcome new subscribers who join from your booking form.
• AI Receptionist for a Campground: After-Hours Calls, No Night Warden — Handle the booking calls the emails will generate.
• Best AI Email Marketing Tools for Small Businesses (2026) — Compare email tools before you commit to one.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Gmail email sender guidelines (spam-rate thresholds; one-click unsubscribe for senders of more than 5,000 messages a day to Gmail accounts); Mailchimp and MailerLite pricing pages (USD list prices, to confirm before buying); OpenAI and Anthropic privacy settings help pages. All campground figures in the examples are illustrative.