Sample park page: real numbers from a Texas park (name hidden). This is the dashboard the owner keeps; the Monday report is one of its tabs. Yours is built 7 days after one export.
Sample Park · your park page
Updated Wed 30 Sep · sales to Tue 29 Sep · ✓ Built from the booking-system export: 15,182 sale lines, 1 Jun – 29 Sep
See what each plan shows:Free 30 days: everything your data allows.
Joined on this page:✓ Booking system 15,182 sale lines, 1 Jun – 29 Sep✓ School calendars + holidays public✓ Website ↔ till joined by order number (97% matched) · 31 May – 27 Jun✓ Google + Meta ads ↔ till each sale counted once · 1 Jun – 7 Jul
Flagged for you
Something lifted sales about 70% from 20 July, and it wasn't the price. What happened? Did it work? →
1 in 6 entries since the price cut paid $0, almost all let in at the front desk. Pass holders, staff or comps? Money & guests →
School days off: Thu 8, Fri 9, Mon 12 Oct. You're usually closed Mon–Thu; Labor Day took $11.8k. Days & hours →
Website alert6 in 7 website visitors never reached the booking site, and most were on a phone. Visit to booking →
Ads alertOne ad channel cost 16 times more per sale, and got two-thirds of the money.Cost per customer →
Revenue
from your booking system
Sales last week (21–27 Sep)
$32.8k
▼ 8% vs week before ($35.8k)
Money per paying guest
$38
▼ from $50 before the price cut
Let in for $0 since the cut
1 in 6
up from 1 in 23 in June
Best day of the week now
Saturday
$15.5k on a school-season Saturday
not in your POS
Sales each week, with changes marked
Labelled by the week's last day. 1 = price cut, 2 = 20 July jump, 3 = schools back.
Money on a typical open day more →
Guests per hour at each activity, weekend more →
Money per paying guest, by what they bought more →
How far guests travel more →
Group bookings, year on year more →
May 2024 – Apr 2025 vs May 2025 – Apr 2026
Revenue pages: Money & guests · Days & hours · Did it work? · Where guests come from
Web traffic
your website joined to your till
Visits that reached the booking site
1 in 7
6 in 7 never got there
Visits that saw a ticket
2 in 100
1,425 of 60,246
Visits that ended in a paid online booking
1 in 110
counted in your till
On a phone
86%
of website visitors
Web traffic pages: Visit to booking · Website ↔ till· 4 weeks in June
In Plan 2 · + Website 1 alert waiting about your website. Opens in Plan 2; included in your free 30 days.
Ads performance
each sale counted once, against your till
Cost per sale · Google
$37
$4.10 back per $1
Cost per sale · Meta (Facebook)
$615
27¢ back per $1
Ad money that went to Meta
2 in 3
$28,886 of $44,755
Sales the platforms claimed
519
vs 394 online ticket orders in the till
Ads pages: Cost per customer · Platform vs till · Where the budget went· 1 Jun – 7 Jul
In Plan 3 · + Ads 1 alert waiting about your ad spend. Opens in Plan 3; included in your free 30 days.
Did the late-June price cut work?
Since the late-June price cut: sales down about 12% a day, and each guest brings $12 less.
Money per paying guest
$50 → $38
▼ −24%
Upgrades per 100 guests
~20 → 7
▼ −65%
Sales a day
$7.4k → $6.6k
▼ −12%
Weekends fell most: −37%. A booking-appointment system also started in late June; on your page we mark every change like this and ask you about it.
The catch: at $12 less per guest, the park needs about 31% more paying guests just to break even. Test it: one weekend at the old price. If Saturday beats $16.7k, the old price wins.
We also found a second change: from 20 July sales a day jumped about 70% (to $11.2k) while money per guest stayed flat, so it was not the price. Your page flags changes like this and asks you what happened.
$37 vs $615what one paying customer cost on two ad channels at this park (see Ads)
1,781 vs 1,216sales the website counted vs sales actually paid (see Web traffic)
413 giveawayssweepstakes guests, counted apart from pass and comp visits
Want this page for your park? Free for 30 days.
1 Tap the button (it opens an email that says yes)2 We tell you the one export to send3 Your page in 7 days
No card. No passwords. Nothing renews on its own: on day 25 we show you what your page found, and you decide. A real analyst builds your page and answers your reply, usually the same day (US hours).
Test the old price on one weekend. If Saturday beats $16.7k at the old price, the old price wins.
Bring upgrades back. About 7 in 100 paying guests upgrade now, down from about 20. Offer it at checkout and at the front desk.
Thu 8, Fri 9 and Mon 12 Oct are school days off locally. You're usually closed Mon–Thu: open Thu 8 and Mon 12 (Labor Day, Mon 7 Sep, took $11.8k) and staff all three like a Sunday.
See why: two changes, before and after
not in your POS
What changed, and when
Weekly, labelled by the week's last day. Line 1 = price cut (24 June). Line 2 = second change (20 July).
Weekend dayWeekdayaverage sales a day
Money per paying guest
Upgrades per 100 paying guests (weekly; before = median of 3 weeks)
Show as a table
See why: next 7 days, plus school days off
not in your POS
Likely sales each day
From the last 4 school weeks, joined with local school calendars.
School days off come from the local districts' calendars; we confirm your district's dates when we build your page.
How the numbers work
Data: sales lines 1 Jun – 29 Sep; admissions 1 Jun – 2 Aug. Before = 1–23 June. After the cut = 25 June – 19 July. Second change = 20 July – 9 August (schools went back 10 Aug). Sales = every booking-system sale line minus refunds. Paying guests = entry tickets on paid bookings (activity sessions not counted as extra guests). Upgrades = median weekly upgrades per 100 paying guests. Break-even: $50 ÷ $38 = 1.31. Likely sales = the middle range of recent days like it; days off use summer days of the same weekday and Labor Day.
What does a guest really bring in?
Since the price cut, 1 in 6 people through the gate paid $0, up from about 1 in 23 in June (not counting the June giveaway). Almost all were let in at the front desk. Who are they?
Sho's note: 791 free entries in the 25 days after the cut. If they had paid the day-pass average ($38), that's about $30k: an "if", not a loss, because pass holders, staff and comps may be among them. Tell us which, and this tab splits them. Paying day-pass guests also bring $38 now, down from $52.
Dates
to
Compare to
not in your POS
Free tickets each week
$0 entry tickets (not counting upgrades and activity sessions). Labelled by the week's last day. No admissions data 3–9 Aug; June is partly covered.
$0 passes and compsSweepstakes
not in your POS
Money per paying guest, by what they bought
Each booking's money ÷ its entry tickets. Bookings 1 Jun – 16 Aug.
Group bookings: more guests, less money each
From your group and event bookings, May 2024 – Apr 2025 vs May 2025 – Apr 2026. Different dates from the rest of this page.
School-trip guests nearly tripled (2,229 → 6,029), but each brought in $26.77, down from $43.70. Youth groups: 321 → 2,687 guests, $33.70 each, down from $48.64.
2024–252025–26, less per guestmore per guest
Money per guest = the group's booking money ÷ its guests. Not shown: unlabelled event bookings (8,452 guests in 2025–26) and small new types such as sports parties and team building.
Sho's note: the extra school trips and youth groups added about $139k, but at lower prices per guest. Worth checking which group prices or discounts changed, before next spring's school-trip season.
Gate or online, by type of day
Share of money taken at the gate (the rest is online).
School-season Mon–Thu: almost no gate sales, so the park looks closed; the online sales on those days are tickets for later visits.
How the numbers workSales = every sale line minus refunds, by the day it was paid. Paying guests and money per guest come from the admissions report (1 Jun – 16 Aug): a booking's money ÷ its entry tickets; activity sessions are not extra guests. Free tickets = $0 entry tickets. Upgrades = upgrade tickets per 100 paying guests. Check: in June the admissions report covers only about 55% of sales (from 80–90% from July), so we don't compare guest counts across June. Money per guest is shown because it held at $47–51 every June week whatever the cover. Parties: only parties booked in this system (36 bookings); parties booked by phone elsewhere are not in it yet. Group bookings (school trips, youth groups, corporate, birthdays) come from the group and event bookings report, two 12-month periods (May 2024 – Apr 2025 and May 2025 – Apr 2026), so they are not on the same dates as the charts above.
When do I need staff, and where?
In school season the till only takes money Fri–Sun, and Saturday is the big day. On summer weekends the Ropes Course averages about 24 guests an hour from noon to 6pm; Archery averages 5 or fewer.
Sho's note: on weekends, Archery could run in set sessions (say 1pm and 3pm) instead of all day; on summer weekdays it's busier (7–8 an hour), so keep it staffed then. Open on the school days off below: Labor Day (Mon 7 Sep) took $11.8k.
not in your POS
Money on a typical open day, by type of day
Average sales a day. School days from the local district calendars.
not in your POS
Guests at each activity, by hour
Average guests booked per hour on a weekend day (24 Jun – 16 Aug, since the booking-appointment system started).
Darker = more guests. Based on 12–14 weekend days (Archery 10) and 21–28 summer weekdays. We don't know each activity's capacity yet: tell us, and this shows how full each one is.
Money by hour
Average sales in each hour, by weekday.
not in your POS
Next 2 weeks
Likely sales on open days, from recent days like it; school days off from local district calendars.
How the numbers workDay types: summer = 1 Jun – 9 Aug; schools went back 10–13 Aug (left out); school season = 14 Aug – 29 Sep. Holidays: 3–4 Jul, Labor Day. Activity guests = booked session tickets by session hour, averaged over days that activity had bookings (Archery bookings in this system stop on 29 Jul, so it covers fewer days). Money by hour = sale lines by the hour they were paid. Likely sales = middle range of the last 4 weeks of the same weekday; days off use summer days of the same weekday and Labor Day.
Is my change working?
Two changes moved sales: the 24 June price cut (sales −12% a day) and something on 20 July (sales a day up about 70%, price per guest flat). We need you to tell us what happened on 20 July.
Sho's note: each change shows sales a day before and after, holidays left out; when the open days differ (like school season), we compare the same weekdays only. Add your own below (a new price, a new ad, a new website) and it's checked here.
not in your POS
Sales every day, with the changes marked
Line = 7-day average. Dots = each day. Numbers = changes in the list below.
Add a change
Pick the date and what you changed. We compare the 3 weeks before with the 3 weeks after, using only weekdays the park was open on both sides, holidays out.
How the numbers workBefore / after = average sales a day over the periods shown on each card, holidays (3–4 Jul, Labor Day) left out. School season compares Fri–Sun with Fri–Sun. Changes you add use the 21 days each side, only weekdays the park was open (took gate sales) on both sides. Changes we spotted in the data are marked "Spotted"; ones you told us are marked "You told us".
Which towns fill my park?
1 in 3 bookings comes from within 20 miles, but 1 in 4 comes from over 60 miles away. Two towns 30–45 miles north-east grew from about 1% to 2–3% of bookings each after June.
Sho's note: guests come from far: 1 in 4 bookings are from over 60 miles, about half of them from outside Texas (visitors on a trip). If you didn't change ads in July, the two growing north-eastern towns found you on their own: a cheap place to test an ad. Small numbers (13–14 June bookings each), so watch it, don't bet on it yet. On your page the towns have their real names.
not in your POS
Where bookings come from
Each dot = one zip code, size = bookings. Rings at 10, 20, 35 and 60 miles. North is up.
How far guests travel
Top 10 towns
Share of bookings in June vs July–mid Aug. Names hidden on this sample.
How the numbers workZip codes from the admissions report, one per booking (1 Jun – 16 Aug). 64% of bookings have a zip code; 36% are blank, so these shares are of bookings with a zip. Distance = straight line from the park to the middle of each zip code. 13% of zip-coded bookings are from outside Texas.
What does a paying customer cost me?
Counted once, a Google sale cost about $37. A Meta (Facebook) sale cost about $615: 16 times more.
Sho's note: before cutting anything, remember Meta may also bring walk-ins at the gate, which online counting can't see. But at 27¢ back per $1 for five weeks, and the same pattern over 14 weeks, it's the first place to test a smaller budget.
not in your POS
Each online ticket sale credited to one channel
1 Jun – 7 Jul 2026 (5 weeks). Spend from the ad accounts; sales are online ticket orders, each counted once.
Same pattern over 14 weeks
1 Jun – 6 Sep 2026, all purchases.
How the numbers workEach sale is credited to one channel using Google Analytics' de-duplicated report, cross-checked with a second, independent tool (last click): the two agreed within two points (about 85% Google, 15% Meta). Spend = the ad accounts' cost. Cost per sale = spend ÷ sales counted once. Money back per $1 = sales value ÷ spend. Walk-in (gate) sales that started with an ad are not counted here.
Are my ad reports telling the truth?
The ad platforms claimed 519 sales. The booking system recorded 394 online ticket orders.
Sho's note: no one is lying: every platform's tag fires on the same checkout, so each one claims the same order. Judge every channel against your till, once.
not in your POS
What each source says (1 Jun – 7 Jul)
Why they over-claim
Every ad tag fires on the same checkout page, so each platform claims the same order.
Meta also counts people who only saw an ad (1-day view) and some in-store sales.
Google's totals can include modelled visits; we kept its "Purchases" only. Its own report said 484 sales; counted once, it had 293 ($23 a sale on its numbers, $37 counted once).
Is my ad money going to the right place?
Meta got about two-thirds of the ad money ($28,886 of $44,755) while bringing in about 1 in 7 of the sales counted once.
Sho's note: over 14 weeks it was the same: Meta took 58% of the budget for 11% of the ad-driven purchases. A test worth running: move part of it to Google for four weeks and watch "Did it work?".
not in your POS
Share of ad money vs share of sales
1 Jun – 7 Jul 2026.
Shares are of the 340 online ticket orders credited to an ad. The other 54 of the 394 online orders weren't credited to Google or Meta.
With your daily ad data, this page also shows
Each campaign's daily budget: used up or not.
Gate sales on days ads run vs days they pause (marked as an estimate).
How many website visitors book, and where do the rest go?
6 in 7 website visitors never reached the booking site. About 1 in 110 visits ended in a paid online booking.
Sho's note: most visitors leave before they see a price: 6 in 7 at the jump from your website to the separate booking site, then 5 in 6 of the rest before a ticket page. 86% were on a phone. Before paying for more visitors, shorten that jump: tickets that open on your own site, big buttons, few taps. Any site change you make is then checked in "Did it work?". How your website is joined to your till is on the next page.
not in your POS
From website visit to paid booking
4 weeks from 31 May. Visits, booking-site and ticket-page counts from the park's weekly website sheet; paid bookings from the booking system.
The first paid online booking in the sales export is on 3 Jun, so the first week is short; "1 in 110" uses the three full weeks (47,047 visits, 438 paid online bookings). Walk-ins at the gate who saw the website first can't be counted, so the website does more than this shows.
not in your POS
Week by week
Google Analytics: different counts, same story
28 days (4 Jun – 1 Jul). Analytics counts sessions and page views its own way, so its numbers are bigger than the weekly sheet; the drop-off looks the same.
86% of these visitors were on a phone.
How the numbers workSee "Website ↔ till" for sources and windows.
Can I trust my website numbers?
Analytics records almost every online sale, but can trace only about 4 in 10 back to the visit or ad that started them.
Sho's note: that's why this page counts paid bookings from your till, never Analytics' own rate. Your booking site sits on a separate address, so the trail breaks at the jump. Joining by order number fixes the count, but not the trail.
not in your POS
How your website is joined to your till
Website Google Analytics
→
Booking site separate address
→
Booking system the till
Analytics sees almost every online sale: 1,183 of 1,216 paid online orders were found in it by order number (97%), 1 Jun – 26 Jul.
But it can trace only about 4 in 10 back to the visit or ad that started them (262 of 684, 15 Jun – 15 Jul): the trail breaks at the jump to the booking site.
It also counted 1,781 "sales" in those 8 weeks; most of the extra are free-ticket orders at $0.
So the sale is recorded, but which visit (and which ad) led to it mostly isn't. That's why we count paid bookings from your till, never Analytics' own rate.
No Google Analytics on your site?
Two ways to get this tab:
We set it up free, in your name (early-bird parks): you own it, we're added as a user, and you can remove us any time.
Or count one week by hand: visits to your website, and visits to your booking page (your booking company can give you that). Divide the second by the first. This park got about 1 in 7.
How the numbers workWebsite visits, booking-site visits and ticket-page visits: the park's own weekly website sheet, weeks starting 31 May, 7, 14 and 21 Jun 2026. Paid online bookings: online sales in the booking system with money taken, counted once per booking. Visits per paid booking = visits ÷ paid online bookings for the same weeks. Google Analytics figures are its own counts (sessions and steps), read on 2 Jul 2026. The 1,781 / 1,216 / 1,183 figures come from our 29 Sep check (1 Jun – 26 Jul); the 262 of 684 from the purchase-journey report (15 Jun – 15 Jul).
Mainnov · one page that joins your booking system with your calendar, ads and website · mainnov.com