Trampoline Park Revenue by Size Band: 106 Parks Compared
Write down two numbers. The 25,001–30,000 sq ft size band in Sky Zone’s 2026 franchise disclosure averaged $2,485,250 in gross sales per park. The band above it, 30,001–58,004 sq ft (2,787–5,389 m²), averaged $2,551,115.
By midpoint the larger group carries 60% more floor, and 93% more at the top of each range. It earned 2.6% more money. If floor area drove revenue the way the equipment catalogs imply, that gap would be enormous. It is almost nothing.
This is not an argument that size does not matter. Every band earns more per park than the one below it, and the smallest band is clearly the worst business in the table. The point is narrower and more useful: the second thousand square meters is worth far less than the first, and the published data lets you put a number on how much less.
We ran that arithmetic on 106 franchised Sky Zone parks, 64 franchised Altitude parks and 243 Dave & Buster’s and Main Event venues. Every figure below is either quoted from a filing or derived from one, with the derivation shown. Where a number cannot survive its own error bar, we say so instead of rounding it into a conclusion.
1. What the disclosure actually contains
Sky Zone’s Item 19 is unusually granular for this industry. Instead of a single system average, the franchisor splits 106 reporting franchised parks into four bands by square footage and reports average gross sales, median gross sales and average EBITDA for each. That structure is the reason this article is possible at all: a size curve needs both a size and a revenue, and most operators publish neither.
| Size band (sq ft) | Size band (m²) | Parks | Avg gross sales | Median gross sales | Avg EBITDA |
|---|---|---|---|---|---|
| 16,000–22,500 | 1,487–2,090 | 25 | $1,828,750 | $1,800,570 | $396,719 |
| 22,501–25,000 | 2,090–2,323 | 21 | $2,015,495 | $2,120,549 | $415,712 |
| 25,001–30,000 | 2,323–2,787 | 30 | $2,485,250 | $2,231,719 | $578,290 |
| 30,001–58,004 | 2,787–5,389 | 30 | $2,551,115 | $2,291,994 | $553,487 |
| All reporting parks | — | 106 | $2,255,992 | $2,140,768 | $496,683 |
Source: Sky Zone 2026 FDD, Item 19, 2025 reporting year, with band counts and medians as transcribed by Capital Advisors’ filing teardown. The franchisor excludes 16 of 122 franchised parks from the table, mainly for partial-year trading, openings and transfers.
Two details are easy to miss and both matter later. First, the mean sits below the median in the 22,501–25,000 band, and again in the aggregate. A mean under the median means a tail of weak parks is dragging the average down, so “the average park” is not the typical park. Second, gross sales across the whole population run from $601,841 to $7,971,315, a spread of roughly 13 times, and only 41.5% of parks met or beat the aggregate average. Any decision built on a single average is a decision built on the right tail.
The bands also explain themselves. A franchisor bands by size because size is the variable it can screen at site approval. Rent, build cost and equipment count all follow from it. Revenue, apparently, follows much more weakly.
2. Revenue per square foot, and why the number fights back
The obvious next move is to divide. Take each band’s average gross sales, take a representative area for that band, divide one by the other, and you have revenue per square foot, which converts to revenue per square meter at 10.764 sq ft per m².
The problem is the word “representative”. A band spans a range, and the average park inside the band is not necessarily at the midpoint. So the honest version of the calculation is a range, not a point.
| Group | Area used (sq ft) | Area (m²) | Revenue per sq ft | Revenue per m² | EBITDA per sq ft |
|---|---|---|---|---|---|
| Band 1 midpoint | 19,250 | 1,788 | $95 | $1,023 | $20.6 |
| Band 1 assuming 16,000 sq ft | 16,000 | 1,487 | $114 | $1,230 | $24.8 |
| Band 1 assuming 22,500 sq ft | 22,500 | 2,090 | $81 | $875 | $17.6 |
| Band 2 midpoint | 23,750 | 2,206 | $85 | $914 | $17.5 |
| Band 3 midpoint | 27,500 | 2,555 | $90 | $973 | $21.0 |
| Band 4 midpoint | 44,000 | 4,087 | $58 | $624 | $12.6 |
| Altitude, about 25,000 sq ft | 25,000 | 2,323 | $82 | $880 | $20.1 |
Every row is our arithmetic: disclosed average gross sales or EBITDA divided by the area shown. Altitude publishes one reference size, “about 25,000 sq ft”, against a $2,105,000–$3,477,500 investment range, so that row carries a wider error than the others.
Read the first three data rows together. The same cohort of 25 parks produces $114, $95 or $81 per sq ft depending only on where inside the band you assume the average park sits. That gap of roughly 40% is wider than the gap between any two band midpoints in the table. The denominator swallows the signal.
This is why the “$95 per square foot” figure quoted for trampoline parks should be treated as a sanity check rather than a benchmark. It is a real division of real numbers, but it cannot rank two parks, and it cannot support a claim that one size band outperforms another. The uncertainty around it is larger than the effect being measured.
What survives is the level. Across four bands in one brand and one comparable brand, revenue per square foot clusters between roughly $81 and $95 in the 1,500–2,800 m² range, and it declines sharply above 2,787 m². A trampoline park floor, whatever its size, produces on the order of $900 to $1,000 per square meter per year in gross sales. That level is stable enough to build a budget on. The differences between bands are not.
3. The marginal square foot is the number that decides a lease
Leasing decisions are not made about averages. They are made about the next 1,000 square meters: whether to take the 2,800 m² unit or the 4,000 m² one down the road. Rent on that extra floor is what closes the case, and rent per square foot varies more between markets than the revenue curve varies between bands. For the revenue question the bands read as a ladder, and the rungs can be subtracted.
| Step between band midpoints | Added area (sq ft) | Added gross sales | Marginal per sq ft | Marginal per m² |
|---|---|---|---|---|
| Band 1 to Band 2 | +4,500 | +$186,745 | $41 | $446 |
| Band 2 to Band 3 | +3,750 | +$469,755 | $125 | $1,348 |
| Band 3 to Band 4 | +16,500 | +$65,865 | $4 | $43 |
| Band 1 to Band 3 | +8,250 | +$656,500 | $80 | $856 |
| Band 1 to Band 4 | +24,750 | +$722,365 | $29 | $314 |
Our arithmetic again: differences between band averages, divided by differences between band midpoints. The sequence is not monotonic, which is itself informative. Band 2 underperforms on the mean, probably because a cluster of young parks sits inside it, so the step from Band 2 to Band 3 looks strong. The step that matters for a lease is the last one.
Above roughly 2,800 m², the marginal square foot returned about $4 of annual gross sales across this population. The step from Band 3 to Band 4 adds 16,500 sq ft and $65,865. If that floor earned at the same rate as the Band 1 average, it would be expected to produce around $1.57 million. It produced four cents on the dollar.
Now put that against what the floor costs. Sky Zone’s Item 7 pairs its investment estimate with its size range: $3,246,160 to $4,782,210 for 16,000–27,000 sq ft, and $4,189,160 to $5,683,710 for 27,000–39,000 sq ft. Divided at the low ends, that is about $203 and $155 per sq ft. Divided at the high ends, about $177 and $146. Item 7 totals include the franchise fee, equipment, signage, training and working capital, so these are upper bounds on marginal build cost, not construction estimates.
The asymmetry that matters: in the top band, the marginal square foot earned around $4 of gross sales a year while carrying a build cost in the $146 to $203 range. Even at a 22% EBITDA margin, the marginal floor does not repay its own build cost inside a normal lease term. The floor you take on day one is a capital decision, and the last 1,000 m² is the most expensive part of it.
This is the reversal most feasibility models get wrong. They start with the available unit, subtract rent, and ask whether the numbers work at that size. The disclosure suggests the reverse order: fix the revenue layers you intend to run, derive the floor those layers need, and only then look at units. A 4,000 m² box holding a single-layer offer is a worse asset than a 2,400 m² box holding two layers, and the difference is visible in the band averages before a single site visit.
4. Why the curve bends: the floor is a carrier, not an engine
Revenue per square meter has an upper bound that comes from physics and pricing, not from design. A jump court holds a number of people, they stay for a session of fixed length, and each session sells at a ticket price. Multiply sessions by capacity by price and you have the admission layer’s ceiling for that floor. Adding floor replicates the same ceiling in parallel.
A second revenue layer works differently. It sells to the same visitor again, inside the same visit, without extra floor. That is the whole difference between a trampoline park at $85 per sq ft and a family entertainment center at $210.
Dave & Buster’s files the clearest version of this because it discloses both square footage and a revenue split. In fiscal 2025, the year ended February 3, 2026, it operated 179 Dave & Buster’s stores averaging 37,000 sq ft and 64 Main Event stores averaging 53,000 sq ft. Multiply the two out and the estate is about 10.0 million sq ft, or 930,000 m².
| Line, fiscal 2025 | Amount | Share of revenue | Per sq ft | Per m² |
|---|---|---|---|---|
| Entertainment revenue (games, attractions, parties) | $1,323.5M | 62.9% | $132 | $1,423 |
| Food and beverage revenue | $779.3M | 37.1% | $78 | $838 |
| Total revenue | $2,102.8M | 100% | $210 | $2,260 |
| Four-wall contribution after products, payroll and other store costs | $541.6M | 25.8% | $54 | $582 |
| Capital expenditure | $391.4M | 18.6% | $39 | $421 |
Derived from the 10-K: revenue lines as reported, estate area as store count multiplied by disclosed average store size, per-unit figures our division. Four-wall contribution is total revenue less cost of products, operating payroll and benefits, and other store operating expenses.
Two comparisons stand out. The food and beverage line alone, at $78 per sq ft, is roughly nine-tenths of an entire trampoline park’s total revenue per square foot. And the entertainment line, at $132, exceeds the whole trampoline park average even though a trampoline park’s entertainment offer is its entire business.
The reason is not that arcades are magic. It is that the Dave & Buster’s visitor spends across three registers in one visit: games, food, drinks. A trampoline park visitor typically spends across one and a half, because the café is an afterthought and the party room is booked by a minority on any given Saturday. Same floor, same family, half the transactions.
One more difference is structural and worth stating plainly. A restaurant and a bar need seats, a kitchen and a service path, and those consume the floor that the jump court would otherwise use. Dave & Buster’s is not a trampoline park that added food. It is a food and games business whose floor happens to be large. Copying its per-square-foot output means copying its cost structure, including a kitchen brigade and a bar license, neither of which appears in a trampoline park’s payroll line.
5. Where the peak sits, and what it is made of
If revenue per park rises with size but revenue per square foot falls, the operating answer should sit somewhere in the middle. The disclosed EBITDA makes that visible.
| Band | Avg gross sales | Avg EBITDA | Margin | Per sq ft |
|---|---|---|---|---|
| 16,000–22,500 sq ft | $1,828,750 | $396,719 | 21.7% | $20.6 |
| 22,501–25,000 sq ft | $2,015,495 | $415,712 | 20.6% | $17.5 |
| 25,001–30,000 sq ft | $2,485,250 | $578,290 | 23.3% | $21.0 |
| 30,001–58,004 sq ft | $2,551,115 | $553,487 | 21.7% | $12.6 |
Margins computed as average EBITDA divided by average gross sales. The 22,501–25,000 band is the weak rung on all three measures, which suggests the problem is not area but what was put in it: parks of that size are small enough to lack a full attraction mix and large enough to carry a full building’s fixed costs.
The 25,001–30,000 sq ft band leads on EBITDA in absolute terms ($578,290), on margin (23.3%) and on EBITDA per square foot ($21.0). The band above it has more than double the floor at its midpoint and earns less EBITDA. That is the shape of a business whose fixed layers are paid for by about 2,500 m² and whose extra area then dilutes rather than compounds.
A separate cohort in the same filing points at the other lever. Sky Zone isolates 34 “Model Parks” defined by three conditions at once: at least 25,000 sq ft, four or more private party rooms, and a full year of trading. Those 34 averaged $2,847,069 in gross sales and $710,790 in EBITDA at an average size of 31,843 sq ft, which sits inside the weakest band once size alone is considered.
Read carefully, that is the most useful row in the document and also the most misused. The cohort is screened, not random: the franchisor selected the parks on the outcome being measured, so the premium cannot be attributed to party rooms, to size, or to any single cause. What it does show is that parks of roughly 2,960 m² can earn 55% more than the smallest band while carrying the same floor area as the band that underperforms. Something other than area separates them, and room count is the most plausible candidate the filing offers.
6. What the data does and does not show
This section decides whether the rest of the article is worth anything, so it is deliberately blunt.
What it does show A franchisor with 122 franchised parks reports revenue by size band, with enough parks in each band (21 to 30) for the averages to carry weight. A second brand at a fixed reference size reports a comparable level. A 243-venue listed operator reports revenue, floor area and a three-layer mix, which allows a like-for-like comparison across formats. The direction of the size curve is consistent in all three: revenue per park rises with area, revenue per square foot falls.
What it does not show The Item 19 figures are franchisee-reported and unaudited by the franchisor, and 16 of 122 parks are excluded. Gross sales is not net revenue and not profit: it sits before royalties, marketing, rent, payroll and debt service. Square footage is gross leased area, including back of house, party rooms, café and circulation, so the playable court is a smaller and unpublished number. Every per-square-foot figure here inherits all of that.
Where the arithmetic is fragile Dividing a band average by a band midpoint assumes the average park sits mid-band, which nothing confirms. As Section 2 shows, that assumption moves the answer by 40% inside a single band. The marginal figures in Section 3 are differences of averages, so they inherit the noise of both ends; the $4 per sq ft result is robust in direction but is not a precise value.
What the standards do not answer The equipment rules that govern the court itself, ASTM F2970 in the United States and EN ISO 23659 in Europe, set out how a court and its surroundings must be built and operated. They do not define area ratios, revenue density or a minimum viable footprint, so none of this article can be read as a compliance argument. Our summary of the standards that apply in each market covers that question separately.
Where the disclosure contradicts itself The Model Parks margin is published as 23.6%, but dividing the two disclosed averages ($710,790 by $2,847,069) gives 25.0%. We report both and adopt neither. A filing that cannot reconcile a margin stated in the same paragraph is a filing to read with the PDF open, not through a summary.
What cannot be computed at all Dave & Buster’s reports one segment, so the 37,000 sq ft format and the 53,000 sq ft format cannot be separated in the filings; only their blended result is disclosed. No trampoline operator publishes playable floor, visitor counts or revenue by attraction, so revenue per visit and revenue per playable square meter are not obtainable from public data. Altitude discloses cost of goods sold at 9.1% of gross sales and payroll at 18.7%, with EBITDA at 24.6% for 29 accounting parks, but it does not pair those lines with floor area, so its per-square-foot margin cannot be built.
And the oldest caveat of all This is a correlation across parks that chose different markets, buildings and offers. Size bands are not an experiment. A 30,000 sq ft park is not a 25,000 sq ft park plus 5,000 sq ft; it is a different site, usually in a different trade area, often with a different landlord and rent. The curve describes the population. It does not license the claim that adding floor to an existing park will move its revenue by the amount shown.
7. Running the same arithmetic on your own numbers
The value of a published curve is the method, not the values. The method is four lines long and needs no database.
Revenue per sq ft = gross sales ÷ gross leased area Marginal revenue per sq ft = (sales at size B − sales at size A) ÷ (area B − area A) EBITDA per sq ft = EBITDA ÷ gross leased area Years to repay = build cost per sq ft ÷ EBITDA per sq ft
The first line is the one everybody quotes and the least useful alone, because it hides both the error bar and the marginal question. The second and fourth lines change a lease decision. Run the fourth before signing, not after: if EBITDA per square foot is around $20 and the build costs $150 per square foot, the floor repays itself in about seven and a half years at the EBITDA line, before debt service and before any refresh capital. If projected EBITDA per square foot is $12, the same build takes over twelve years, which is longer than most first lease terms.
Three questions separate a real benchmark from a brochure. Ask how many parks sit behind each average and whether the figure is a mean or a median: a mean over nine parks says very little. Ask whether the area quoted is gross lease area or playable area, because the two differ by the width of the back of house and the café. Ask whether the comparison group is a size tier or a screened cohort, because a size tier is a benchmark and a screened cohort is a highlight reel.
By venue size Under about 1,500 m², per-square-foot revenue is capped by throughput and the binding constraint is the number of revenue layers, not the court count. Between 2,300 and 2,800 m², the disclosed data shows the best combination of absolute EBITDA and EBITDA per square foot, provided the mix includes party rooms rather than extra open jump. Above 2,800 m², the marginal floor needs its own layer or it dilutes the blended return, and the lease should be priced accordingly.
For the cost side of the same square meter, our comparison of five construction filings sets out what a build costs per square foot. For how the floor should be divided once the size is fixed, the zone mix analysis covers that allocation. This article is the revenue side of the same arithmetic.
8. Summary
Revenue per park rises with size, and it rises far more slowly than area does. Across 106 franchised parks, moving from the smallest band to the largest added about 129% more floor at the midpoints and 39% more gross sales. Above roughly 2,800 m², the marginal square foot returned around $4 of annual gross sales against a build cost near $150.
Revenue per square foot inside the 1,500–2,800 m² range is stable at roughly $900 to $1,000 per square meter per year, and the differences between bands are smaller than the error introduced by assuming where inside a band the average park sits. That makes per-square-foot rankings unreliable and per-square-foot levels useful.
The peak of the disclosed curve is the 25,001–30,000 sq ft band, which leads on EBITDA, on margin and on EBITDA per square foot. The band above it earns less EBITDA on more than double the floor. The lever that appears to separate the strongest parks from the weakest is not area but the number of times the same visitor spends inside the same visit.
9. Three Things You Can Do Today
1. Compute your marginal cost per square meter before your average. Take the two units you are actually choosing between, subtract the areas, subtract the rents, and divide. The decision is the difference, not the average.
2. Count the revenue layers in your plan, not the attractions. Admission, party, food, retail, membership. A fourth attraction is another product in layer one. A party room is a second layer on floor you already pay for. If you want a floor plan checked against a layer count for your target size, the design-to-delivery process starts with the area and the offer together.
3. Ask for the median and the count. Any benchmark without both is a marketing number. The gap between a mean of $2.26M over 106 parks and a median of $2.14M is not academic; 58.5% of those parks sat below the average.
FAQ
How much revenue does a 25,000 sq ft trampoline park make?
Between $2.0M and $2.5M in gross sales a year, based on the two franchises that disclose at that size. Altitude reports a $2,045,001 average and a $1,884,486 median across 64 reporting parks at about 25,000 sq ft. Sky Zone’s 25,001–30,000 sq ft band averages $2,485,250 with a $2,231,719 median across 30 parks. Both figures are gross sales before royalties, rent, payroll and debt service.
Is a bigger trampoline park more profitable?
Up to a point, and then it stops. The disclosed bands show EBITDA rising from $396,719 in the smallest band to $578,290 in the 25,001–30,000 sq ft band, then falling to $553,487 in the 30,001–58,004 sq ft band despite more than double the floor at the midpoint. Above roughly 2,800 m², additional area appears to dilute EBITDA per square foot rather than add to it.
What is revenue per square foot for a trampoline park?
Around $81 to $95 per sq ft, or roughly $880 to $1,023 per m², in the 1,500–2,800 m² range. Treat that as a level, not a ranking: the same cohort of parks divides out anywhere between $81 and $114 per sq ft depending on the area assumed for the average park, a wider spread than the difference between size bands.
How many party rooms should a trampoline park have?
The disclosure supports four or more as a description of the strongest cohort, not as a proven threshold. Sky Zone defines its 34 Model Parks as at least 25,000 sq ft with four or more private party rooms and a full year of trading, and that group averaged $2,847,069 in gross sales and $710,790 in EBITDA at an average size of 31,843 sq ft. Because the cohort is screened on outcomes, no published figure isolates what a fourth room adds.
What share of revenue comes from food and beverage?
No trampoline operator in this comparison discloses a split, so there is no industry figure to quote. The nearest hard reference is Dave & Buster’s, where food and beverage produced 37.1% of revenue in fiscal 2025, or $78 per sq ft against $132 per sq ft for entertainment. Altitude’s reported cost of goods sold of 9.1% of gross sales implies a comparatively light food operation, but the filing does not break out that revenue line.
Sources
- Dave & Buster’s Entertainment, Inc., Form 10-K for the fiscal year ended February 3, 2026, filed March 31, 2026: store counts, average store size, revenue mix, cost lines and capital expenditure — sec.gov
- Capital Advisors, “Sky Zone franchise unit economics”, 2026 FDD Item 19 size bands, park counts, medians and EBITDA — capadvisors.co
- FranchisesBiz, “Sky Zone franchise pros and cons”, analysis of the FDD issued May 6, 2026, Item 7 pages 17–22 and Item 19 pages 62–66 — franchisesbiz.com
- FDD Exchange, “Altitude Trampoline Park Franchise FDD (2026)”, 64 reporting parks, average and median gross sales, accounting-park cost ratios — fddexchange.com
- Altitude Trampoline Park, franchise performance summary citing the 2026 FDD: average unit volume $2.045M, EBITDA 24.6%, cost of goods sold 9.1% — altitudetrampolinepark.com
- ResearchForSale, Sky Zone FDD filing index: issue date May 2026, Item 7 initial investment range and page reference — researchforsale.com
- ASTM International, F2970-22, standard practice for trampoline courts: equipment and operational scope. It does not define area ratios or revenue metrics — astm.org
