How to Price Tour Packages: Per-Person and Group Rates With AI

Work out what one departure costs, split into a fixed part and a per-guest part, then divide the fixed part by what each guest contributes to find the break-even group size. Set the per-person rate so break-even sits near two-thirds of capacity, quote private groups from cost, and let AI replay last season at the new prices.

Break-even is not one number, because the vehicle changes with the group. The same tour that breaks even at nine guests in a minibus can lose money at seven in a minivan, and a flat $95 price can look healthy on revenue while leaving almost nothing. In the example below, a season of 24 departures brings in $22,420 and keeps $200 after the cost of running each one.

Follow me on Instagram@sagnikteaches

A small operator, one tour and a flat $95

The operator here is invented: a small company running one guided full-day tour with a driver and a separate guide, in whichever vehicle the group size needs. Last season it ran 24 departures at a flat $95 a person and counted the season a success because it was busy. Revenue was $22,420. The owner had never added up what each departure cost, only watched the bank balance.

Connect on LinkedInSagnik Bhattacharya

This is last season's booking record, grouped by time of year, with the guests booked on each departure. The largest vehicle seats 14.

Subscribe on YouTube@sagnikteaches

Everything that follows is built on this record. It is the one piece of data an assistant cannot supply, and the reason a replay of your own season beats any generic calculator.

Cost per departure, in steps

Split every cost into two kinds. Fixed costs belong to the departure whether it carries four guests or fourteen. Variable costs belong to each guest. The fixed side is not one figure here, because the vehicle, and so the bill, steps up with the group.

The variable side is $32 a guest: a $16 entry ticket, a $14 lunch and $2 for water and headsets. The card fee is 3% of the price, which makes it a percentage cost, so it comes off the price and not out of the $32. At $95, each guest contributes $95 x 0.97 - $32 = $60.15 towards the fixed cost.

A single fixed cost and a single break-even figure would hide the steps at 7 and at 9 guests, and those steps are where this tour's money is won and lost, so they have to be in the sheet. If your own costs are quoted per vehicle size, ask each supplier for the price at every size you run, and keep the invoices: the assistant will need them as numbers, not memories.

Break-even group size at each vehicle, and the dip at nine guests

Break-even for a vehicle is its fixed cost divided by the contribution per guest, rounded up, because you cannot carry 8.4 people. Here it is at the current price and at two others:

At $95 the minibus needs 11 of its 14 seats full, 79%, and the two smaller vehicles can never earn their keep. Now look at what happens departure by departure at $110, where the result is contribution from all guests minus the fixed cost of the vehicle that group size needs:

The row has a kink. Going from eight guests to nine adds $74.70 of contribution and an $85 step for the bigger vehicle, so the ninth booking lowers the departure's result from $58 to $47. A ninth guest is worth having only if a tenth follows. That one fact changes how late bookings and group requests should be handled, and no single break-even figure would have shown it.

Week one: ask the assistant for the grid, and check its rounding

Building this table by hand can take an hour. An assistant does it in seconds, provided the prompt spells out the steps.

My tour has fixed costs per departure that depend on group size: 1 to 6 guests $480, 7 to 8 guests $540, 9 to 14 guests $625. Each guest costs $32 (ticket, lunch, extras) and card fees are 3% of the price. For a price of $110 per person, make a table of the result per departure for 4 to 14 guests, using the fixed cost that matches each group size. Then give the break-even group size for each vehicle, rounded UP to whole guests, and say if any vehicle can never break even. Show one row's working.

A first answer with the prompt left looser than this often comes back like so, and the slips are worth knowing:

• "Contribution per guest is $74.70. Break-even is $625 / $74.70 = 8.4, so about 8 guests." Rounded down. At 8.4 you are still short; the answer is 9.
• A results table that uses $625 for every row. Eight guests would travel in the minivan at $540, so the table hides the cheaper vehicles and the dip at nine. The fixed cost has to follow the group size.

Fix both by saying "round up" and "use the fixed cost that matches each group size", as the prompt above does, then check one row by hand every time. Ten guests at $110: 10 x $74.70 = $747, minus $625, is $122. If the assistant's row says anything else, find out why before you go further. The break-even analysis tutorial covers the single-product version of this sum if you want the basics first.

Week two: set the standard price from break-even and the street

A rule of thumb the operator adopts: the price should put break-even in the biggest vehicle at about two-thirds of its seats, so an ordinary departure, not only a sell-out, earns something. Two-thirds of 14 is 9.3. At $95 the break-even is 11 guests; at $110 it is 9, which fits; at $125 it is comfortably lower. That is a starting rule and not a law, so loosen it if your tours fill reliably and tighten it if they do not.

The rule gives a floor, not a price. The other half is what comparable tours charge, and for that the owner reads the published prices of half a dozen similar day tours and writes them down. An assistant cannot reliably look them up, and a figure it offers without a page you can open is a guess. If comparable days turn out to run from $99 to $135, $110 sits inside the range and clears the break-even test, so it becomes the standard price.

One more input belongs in the grid before anyone relies on it: the channel. If part of the business comes through a marketplace that charges commission, each of those guests contributes less. At an assumed 20% commission, a marketplace guest at $110 contributes $110 x (1 - 0.20 - 0.03) - $32 = $52.70, against $74.70 direct, and a minibus departure sold entirely that way would need 12 guests to break even, not 9. If the marketplace takes the payment and you pay no card fee on those bookings, drop the 0.03: the guest contributes $56 and the answer is still 12. The 20% is a placeholder: read the commission in your own contract and add a line for the share of guests who arrive that way. The tutorial on rewriting marketplace listings covers the listing side, and the booking software comparison lists the rules to test a system against, from minimum numbers to private hires and seasonal prices.

Week three: group rates that cover their own cost

Group rates come in two kinds, and they follow different logic. A private departure is a quote from cost. A discount for a party that joins a public departure is a concession with a ceiling.

For a private group, work from the vehicle the group needs: the cost is the fixed cost plus $32 a guest, and the price divides that by one minus your margin and the card fee. With a 25% margin and the 3% fee, the divisor is 0.72.

A private group of six pays $156 a head and a group of fourteen pays $106, below the $110 public price. Neither is a discount or a surcharge chosen by taste. It is what the cost structure says: small private groups are expensive to serve and big ones are efficient. An assistant can build the quote table for you from the tiers once you give it the margin, and the owner's job is to pick that margin and read each number against a calculator once.

Using my cost tiers (1 to 6 guests $480, 7 to 8 guests $540, 9 to 14 guests $625), $32 per guest and a 3% card fee, build a quote table for private groups of 6 to 14 guests. Price = (fixed cost + $32 x guests) / (1 - 0.25 - 0.03). Show the price per person and flag any group whose price per person is lower than my public price of $110.

For a party joining a public departure, the discount ceiling is the departure's result at list price minus the result you need, divided by the guests getting the discount (and by 0.97, because the card fee shrinks with the price). Suppose the owner wants at least $150 from every departure as a share of overheads. A party of eight joins four other guests, so the departure has 12 guests and a result of $271 at list. The ceiling is ($271 - $150) / (8 x 0.97) = $15.60 a head. A 10% discount ($11) fits and leaves $186; a 20% discount ($22) does not, because it leaves $101.

The ninth-seat rule belongs here too. A party that tips a departure from eight guests to nine has to pay for the $85 step, so it gets no discount at all unless more bookings are already in sight. The cancellation policy tutorial covers drafting the policy wording, which is where a minimum group size belongs; that rule is the other half of controlling small departures.

Week four: a peak uplift backed by sell-out evidence

Peak departures last season averaged 13.1 guests on 14 seats, a 94% load. That is the evidence for a seasonal uplift, and the booking dates add to it: if peak departures were full well ahead of the date, the price was too low for those dates, and if the last seats only went in the final days, it was not. The owner checks the booking dates and finds peak departures were full three to four weeks ahead, which supports a rise of about 14%, from $110 to $125.

At $125 the minibus contributes $89.25 a guest and breaks even at just over seven guests, which a peak departure clears with room to spare. Off-peak is a different problem. With an average of six guests in the $480 car, a departure needs about $116 a head just to break even; at $110 it loses $32, and at $95 it loses $119. A discount cannot rescue it, and a price high enough to cover it would be hard to sell off-peak. What the owner changes instead is a rule: shoulder and off-peak departures run only with eight or more guests booked, and anyone booked on a departure that does not reach eight is offered the next date or a private departure at the table's price.

Replaying last season at the new prices

Now the test that makes AI worth using. The assistant takes the 24 real departures and recalculates the season under each price option. The prompt lists the data, the options and the assumptions in plain words.

Here are last season's 24 departures with guests on each: peak 14, 13, 14, 12, 14, 11, 14, 13; shoulder 9, 8, 11, 10, 12, 7, 9, 10, 11, 8; off-peak 6, 5, 8, 4, 7, 6. Fixed cost per departure: 1 to 6 guests $480, 7 to 8 guests $540, 9 to 14 guests $625. Variable cost $32 per guest. Card fee 3% of price. Price options: A, $95 flat. B, $110 flat. C, $125 in peak and $110 otherwise, with shoulder and off-peak departures under 8 guests not running. Assume the same guests book at every price, and count guests on cancelled departures as lost. For each option give departures run, guests carried, revenue and the result by season and for the year. Show the working for two departures.

This is the illustrative result:

Three things to read in it. Option C earns the most while taking less revenue than B, which is the point: revenue was never the target. Option A shows why the owner felt busy and was not paid. And the shoulder and off-peak figures show where the old price did its damage, since peak departures covered for the rest.

Check the totals before you believe the table. The assistant should report 24 departures and 236 guests for A and B, and 18 departures and 201 guests for C, which means 35 guests were on departures that did not run. If the figures differ from your own count, the assistant has dropped or merged rows.

Then stress the assumption that demand stays put. Ask for option C again with 10% fewer guests on every departure, then 25% fewer, rounding each departure down to whole guests and applying the vehicle steps and the eight-guest rule again:

• With 10% fewer guests the season result is about $3,370.
• With 25% fewer guests it is about $1,940.

Both still beat the old $200. That answers the question your numbers can answer, which is how much demand you could lose before the new prices are no better than the old. It says nothing about whether guests will actually accept $125, and only the bookings can answer that. The replay also leaves out your overheads (booking system, website, your own time), so the contribution has to cover those too.

Weeks five and six: publish a batch, then change one thing at a time

The owner does not republish everything at once. The new prices go on the next eight dates, peak dates first, and the shoulder and off-peak changes follow once peak has shown how bookings behave. Changing the price, the minimum group and the channel mix together would leave the owner unable to tell which one moved the numbers.

1. Record pace, not just totals. For each published departure, note the guests booked each week against the same dates last season. A departure that used to be full three weeks out and is still two seats short a week out is telling you the rise was too big for that date.
2. Use a simple rule for the response. If pace is behind on the first batch, halve the rise on the next batch rather than reversing it. If pace holds or improves, extend the rise to the next batch.
3. Ask the assistant for the weekly comparison. Paste the counts and ask it to flag departures behind last season's pace by more than a couple of guests. It can build the table and sort it. The decision stays with the owner.
4. Re-run the grid when a cost changes. A new vehicle quote or a guide's rate rise changes the break-even in minutes once the sheet exists, so no price should outlive an invoice that has moved.

The same approach works for other kinds of tour. The bike tour pricing tutorial shows a start-up version with routes and bookings, where the vehicle steps do not exist but the break-even logic does.

What the operator would change on a second pass

The first version of any model like this is wrong in a few places, and the useful habit is knowing which. Here are the ones this operator would fix next.

• Costs that are not fixed by tier. A guide paid by the day is fixed, but one paid by the hour is not, and a departure that runs long costs more. If your costs flex, give them their own line in the sheet and not a flat figure.
• Children and concessions. A child takes a seat and may take a ticket and a lunch, so price the child from the cost lines: if a child's ticket and meal cost $22 against $32, the price can fall by about $10 and the contribution stays level.
• Weather and cancellations. A departure called off on the day still has some costs. The weather cancellation tutorial covers the messages, and the cost of a cancelled departure belongs in next season's replay as its own line.
• Demand that does not hold still. The replay assumed the same guests at every price, so run it again at the end of the season with the real bookings, and compare what actually happened with what option C predicted.

The pattern to keep is the order: cost per departure, break-even by vehicle, price from the break-even and the street, group rates from cost, replay, and then change prices a batch at a time. The assistant does the arithmetic at each step. You provide the invoices, the booking record and the judgement about what your guests will pay.

Questions about pricing tours

Should I price per person or per vehicle?

Price public departures per person, because guests join one at a time. Price private hires per group, because the cost follows the vehicle and the guide. Quote a private group from cost plus a margin, then show the per-person figure beside it so the group can compare it with your public rate.

Should I publish a group rate on my website?

Publish the public per-person price and a line inviting private groups to ask for a quote. A private price depends on the vehicle the group needs, so one published group rate either overcharges large groups or loses money on small ones. If you do show a figure, label it as a starting price for groups above a stated size.

Can AI predict how many guests will book at a new price?

Not reliably, because it has no view of your demand. It can replay your own booking history at new prices and show how much demand you could lose before profit falls, which is a question your numbers can answer. Whether guests accept the price you learn from watching bookings.

How often should I review tour prices?

At the end of each season, and whenever a vehicle, guide, ticket or meal cost changes. Replay last season's departures at the new costs before you publish next season's dates, so a rise in one supplier's price does not quietly turn a profitable departure into a loss.

Further reads

• Tour Guide Scheduling Software: What to Pick for Freelance Guides — Guide cost is the second-largest line; plan rotas for freelance guides.
• How Small Tour Operators Use AI for Bookings and Itineraries — More jobs AI handles for a small operator, from bookings to itineraries.
• 8 Examples of AI in Tourism From Guided Tours to Group Bookings — More examples of AI in tourism, from guided tours to group bookings.
• AI for Travel Companies: Track Supplier Payment and Release Deadlines — Track supplier payment and release dates for group departures.
• AI Marketing Booklet Creator: Print a Tour Brochure From Itineraries — Turn three tour itineraries into a print-ready booklet: AI-drafted copy, a pricing page, Canva layout, print specs and a proof checklist.
• Walking Tour Software: Ticketing, Guide Rotas and AI Enquiries — How a small walking-tour company sets group caps, guide rotas, weather rules and private-group quotes, and which writing jobs AI can take on.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: no outside data is used. The operator, costs and guest counts are invented for illustration; replace them with your own invoices, booking records and contract terms.