Indoor Play Expansion 2026: Repeat Units, Not New Markets
On 27 October 2026, JumpYard said it will open a 4,000 m² (43,056 sq ft) park in Frankfurt-Griesheim before the end of the year, inside a former cinema. It will be the Swedish operator’s second park in Germany, eight months after Nuremberg, and its 30th in Europe.
The same day, Jump 4U said its seventh trampoline park will open in Świdnica, in western Poland, at 1,200 m² (12,917 sq ft). Two days later, 360 Play said its eighth UK centre will open at Broadway Plaza in Birmingham on a 20-year lease, at 20,000 sq ft (1,858 m²). On 6 November, Prison Island said its second London site opens at Westfield London on 26 November, with 35 cells.
Four announcements in ten days, three markets, four operators. None of them is a new format. Every one is a unit number greater than one.
That is the claim this article tests: the current wave of indoor play expansion is built out of repeat units inside networks that already know how to run the format, and the unit number now predicts the floor area, the equipment order and the schedule. Capital is not going into invention. It is going into the second, seventh and eighth versions of something that already works.
It is not a claim about market size. None of these operators publishes a pipeline, and we are not going to estimate one. What follows is six dated openings, one counter-example, and four observations that would break the argument if they turn out to be true.
1. The six openings, with dates and sources
The evidence base is small and specific. Each row below comes from an announcement with a date, and in four of the six cases the operator’s own website can be checked against the trade report. That check matters more than it sounds, because the trade reports undercount.
| Operator | Market | Unit number | Floor area | Opening | Source and date |
|---|---|---|---|---|---|
| JumpYard | Germany | Second in Germany, 30th in Europe | 4,000 m² / 43,056 sq ft | Before end of 2026 | Amusements & Attractions, 27 Oct 2026 |
| Jump 4U | Poland | Seventh trampoline park | 1,200 m² / 12,917 sq ft | Fourth quarter 2026 | Amusements & Attractions citing Parkmag, 27 Oct 2026 |
| 360 Play | United Kingdom | Eighth centre | 20,000 sq ft / 1,858 m² | Before end of 2026 | Amusements & Attractions, 29 Oct 2026 |
| Hello Park | Peru and India | First in each country | Not published | Fit-out phase | Amusements & Attractions, 23 Oct 2026 |
| Sould Park | Spain | New outdoor park | 21,000 m² / 226,042 sq ft | Before Christmas 2026 | Amusements & Attractions, 30 Oct 2026 |
| Prison Island | United Kingdom | Second in London | 35 cells | 26 Nov 2026 | Amusements & Attractions, 6 Nov 2026 |
Read the third column as a group and the pattern is hard to miss. Two of the six are second units. One is seventh. One is eighth. The two that are not repeats by number are repeats by another route. Hello Park is adding its brand to two countries through local operating partners, and Sould Park is adding an outdoor format in a market where it already runs family entertainment centres.
There is a mundane explanation for this, and it is probably the right one. A second unit costs less to think about than a first. The layout exists, the equipment list exists, the staffing model exists, and the opening has already been debugged once. What the operator buys on unit two is not a concept. It is a repeat with corrections.
What we did not do: we did not add these floor areas together to produce an industry figure. Six openings are six decisions, not a market total, and the sample leans toward announcements that named a size.
2. Unit two is where a format gets tested twice
JumpYard is the cleanest case because the operator publishes its own park list, and the list can be counted.
The Swedish company was founded in 2017. Its website lists 29 parks as of 9 November 2026: 18 in Sweden, 6 in Spain, 2 in Portugal, 1 in Norway, 1 in Denmark and 1 in Germany. Frankfurt would be the 30th, which matches the number the company gave the trade press. Two independent counts, one number.
That count contains the most useful statistic in this article. Eighteen of 29 parks sit in the home market, or 62.1%. The European footprint is not an even spread across six countries. It is a dense Swedish base with single-digit footholds elsewhere. A second German park is therefore not a step into the unknown. It is a decision to convert a foothold into a market.
Parks per country, from the operator list (9 Nov 2026) Sweden 18 (62.1% of 29) Spain 6 Portugal 2 Norway 1 Denmark 1 Germany 1 Average per country: 29 / 6 = 4.83
The German sequence is what makes unit two interesting. Nuremberg opened in April 2026. Frankfurt is scheduled before the end of 2026. That is two parks in one country inside eight months, after nine years spent building the other 28. The second unit is the one that tells the operator whether the first was a market or an accident.
The Frankfurt site adds a constraint that has nothing to do with trampolines. It is a former cinema, which means the ceiling, the column grid and the floor slab were fixed by somebody else years before the equipment list was written. Auditorium volumes create a different design problem from a warehouse shell, and the difference is not cosmetic. Clear height under a raked ceiling and the load paths around it set what can be stacked, what can be hung and where the high-energy zones can sit.
The company’s description of the mix matters for purchasing. The park is listed with trampoline areas, climbing, slides, interactive games, a SkyRider and a toddler area. That is a multi-attraction list rather than a trampoline court with extras, and every item on it has its own lead time, installation crew and maintenance contract. Coordinating that many suppliers into one opening date inside a leased shell is the real project. The trampolines are the easy part.
One limit needs stating. JumpYard has not published the Frankfurt floor plan, the ceiling height, the number of trampoline beds or the capital cost, and this article will not invent them. What is public is the area, the building type, the mix and the date.
3. Unit seven and eight buy a smaller box, not a bigger one
The opposite pattern shows up in Poland. Jump 4U publishes a dated project list, and the areas on it move in one direction.
| Opened | Site | Floor area | Format |
|---|---|---|---|
| 2016 | Wrocław Tarnogaj | About 1,700 m² | Trampoline park |
| 2022 | Dzierżoniów | Over 1,000 m² | Trampoline park |
| 2022 | Oława | Over 1,000 m² | Trampoline park |
| 2023 | Wrocław TGG Szczepin | 1,400 m² | Trampoline park |
| 2025 | Ostrołęka | About 1,000 m² | Trampoline park |
| Jan 2026 | Strzelin | Not stated | Soft play room |
| Jun 2026 | Nowy Tomyśl | About 480 m² | Trampoline park |
| Q4 2026 | Świdnica | 1,200 m² | Trampoline park |
The first park was about 1,700 m². The median of the seven trampoline parks on the list is about 1,000 m², which is 41% below the first unit. The two most recent are 480 m² and 1,200 m², an average of 840 m², or 49% of the 2016 floor. This is a chain that got smaller as it got better, which runs against what most first-time investors assume.
Geography explains part of it. Six of the eight sites are in Lower Silesia: two in Wrocław, plus Dzierżoniów, Oława, Strzelin and Świdnica. Ostrołęka is in Masovia and Nowy Tomyśl in Greater Poland. A 480 m² box works in a smaller town, because drive time does the segmentation that a bigger building would otherwise have to do. Put the same 480 m² in a city where you already trade and it takes guests from your own site.
There is a counting discrepancy worth flagging, because it tests how carefully these numbers travel. Amusements & Attractions called Świdnica the operator’s seventh entertainment destination. The operator’s own list shows six trampoline parks trading before it, plus a soft play room in Strzelin opened in January 2026. So Świdnica is the seventh trampoline park and the eighth site. Both numbers are right, depending on whether a play room counts.
360 Play gives the same signal in the UK at a different scale. The company’s own centres page lists seven open sites (Basildon, Farnborough, Leicester, Milton Keynes, Redditch, Rushden Lakes and Stevenage) plus Birmingham marked as coming soon. That is the eighth, as announced. The trade report named three of the seven; the operator’s page names all seven. At 20,000 sq ft, the new unit is 1,858 m², about 33% larger than the biggest park the Polish chain has published, the 1,400 m² site in Wrocław.
The 20-year lease is the part an equipment buyer should read twice. A 20-year commitment on 1,858 m² tells you the operator expects the format to hold through a full replacement cycle of the attraction mix. Equipment chosen for a 20-year box has to be serviceable in year 12, which is a different specification from equipment chosen for a five-year lease.
4. The building decides the schedule, not the brand
Two of the six openings are fit-outs inside buildings that already exist: a former cinema in Frankfurt and a unit at Westfield London. A third sits inside an existing retail scheme at Broadway Plaza. That ratio is the operational core of this trend, because a retrofit and a new build fail in different places.
A retrofit fails on survey. The ceiling is where it is, the columns are where they are, and the services run where the previous tenant left them. If design starts before those three things are measured, the equipment list changes after the steel is ordered, and the change order lands on the tenant. This is a scheduling argument rather than a safety one, and it explains why a repeat unit can still miss its date when the layout is a copy.
A new build fails on budget. Sould Park in Malaga is the counter-example in this set: 21,000 m² outdoors next to the Plaza Mayor shopping centre, with two roller coasters, bumper cars, a boat ride, 20 lanes of tenpin bowling and a mini-golf course, opening before Christmas 2026 and employing up to 70 staff. Its budget rose from €15m to €24m, according to SUR, an increase of 60%.
Set the Sould Park number beside the retrofit cases and the contrast is the point. The €9m increase is 60% of a budget that was already agreed, and it arrived during construction rather than before it. Ground-up builds carry that risk because every element is new at once: planning, groundworks, structure, services and fit-out, each with its own approval clock. None of the six operators publishes a capital figure for the indoor units above, so this is not a cost comparison. It is a comparison of what moves.
Prison Island shows what a retrofit schedule looks like when it is compressed. The Westfield London site opens on 26 November 2026 with 35 cells, twelve of which are new challenges not seen at the Islington site that opened in September 2025. That is 34% new content on top of a format that already trades. A cell-based product suits a mall unit, because the constraint is the number of doors off a corridor rather than the clear height over a jump area.
The distinction matters for a mixed box. Formats that consume volume, such as trampoline courts and climbing, are constrained by the shell. Formats that consume floor plan, such as escape rooms, game rooms and soft play, are constrained by the corridor. A landlord with an awkward basement can host the second kind and not the first.
5. New countries arrive on somebody else’s balance sheet
The two genuine first-in-country entries in this set do not look like the others, and the difference is who pays.
Hello Park, which lists more than 50 locations globally, is building in Lima and Hyderabad. Neither is being built by Hello Park on its own. The Lima site, inside Real Plaza Puruchuco, will be operated by Grupo Diverticenter. The Hyderabad site is being developed with Imagicaaworld Entertainment and the Malpani Group, and will be operated by Imagicaaworld. Both projects are described as entering an intensive construction and fit-out phase.
Prison Island works the same way in the UK. Its site lists Maidstone as open, with Leeds and London Westfield marked as coming soon, and a franchise page inviting operators to open their own. The company started in an old factory in Västerås, and it now sells the format rather than the building.
For an equipment supplier this changes the counterparty. In a franchise or licence deal, the order does not come from the brand head office. It comes from a local operator who has just signed a development agreement, usually with a deadline attached, and who may be building a venue for the first time. The brand supplies the specification; the partner supplies the capital and absorbs the overrun.
That structure has a predictable failure mode. The partner underestimates the fit-out, the opening slips, and the equipment arrives before the building is ready. Storage costs then land on the operator, and the warranty clock starts before the first guest walks in. It is worth writing the delivery trigger as a site-readiness condition rather than a calendar date, precisely because the capital sits with the party that has the least experience.
6. What the six have in common, and what they do not
Three properties hold across all six openings. Every operator already runs the format somewhere. Every opening has a date tied to a lease or a season rather than to a construction milestone, which is why a year-end date appears four times in the table above. And every one is a variation on an existing template, including the two new-country entries, which repeat a format that already trades elsewhere.
Three properties do not hold, and pretending otherwise would be the easy mistake. Floor area ranges from 480 m² to 21,000 m², a spread of 44 times, so there is no such thing as a standard unit. The shells differ completely: cinema, mall unit, retail scheme, open air. And the ownership models differ too, because company-operated, leased, franchised and licensed all appear in a six-row table.
| Unit number | What the operator already owns | What the purchase looks like |
|---|---|---|
| First | A site and a business plan | Full design, full equipment list, full risk |
| Second | One template and one opening behind it | Repeat specification, site corrections, tighter date |
| Seventh or eighth | A kit, a spares pool, an install crew | Smaller box, standard components, schedule-led order |
| First in a new country | The brand, no local property | Partner carries capital, head office approves specification |
The table is the practical output of this article. It says that “what does a park cost” is the wrong question at unit seven. The right question is what can be reordered without redesign, and what has to be re-measured because the building is different.
7. What would prove this wrong
A trend claim is worth reading only if it can lose. Four observations would overturn this one, and each is specific enough to check against next quarter’s announcements.
If the 2027 pipeline turns out to be mostly first units. The argument depends on repeat units being the majority of visible activity. If the next twelve months of openings are dominated by operators entering their first market, the pattern here is a scheduling artefact of late 2026 rather than a direction.
If the retrofit units slip past the first quarter of 2027. Four of these six carry a date tied to the year end. If a majority open in the second quarter of 2027 instead, the constraint is not the template but the shell, and the lesson changes from cheap replication to replication that only works when the box cooperates.
If unit counts stall while revenue per unit rises. Replication assumes the format still has room. If JumpYard holds at 30 European parks and grows revenue per park instead, or if Jump 4U stops at eight Polish sites and invests in the ones it has, the growth is coming from yield rather than units. A supplier’s opportunity would then be retrofit work instead of new-build kits.
If a new format enters without an incumbent. Every opening here belongs to an operator with prior sites. A genuinely new format reaching ten sites in a year, from a team with no existing parks, would show the barrier to invention is lower than this article assumes.
One caveat about the evidence itself. Five of the six rows come from a single trade publication, checked against operator websites where those exist. That is enough to establish what happened on a date. It is not enough to establish that these six are typical of the whole market, and we do not claim they are.
8. What this changes in a purchase order
If the argument holds, four things change for anyone buying equipment for a second or later unit.
Measure before you design. The repeat unit’s advantage disappears if the shell survey happens after the layout. Clear height, column positions, slab capacity and the route a 6 m steel section takes from the loading bay to the floor are the four inputs that decide whether the copy fits. On a ceiling-height check, the number that matters is the lowest obstruction in the jump zone, not the highest point of the roof.
Standardise components, not layouts. A chain that repeats a full layout across different shells ends up with bespoke everything. A chain that repeats components (bed modules, pad profiles, spring sets, frame sections) can re-measure the shell each time and still order from the same catalogue. Common spares across seven parks are worth more than a perfect plan in one.
Put the date on the building, not the calendar. A lease commencement date is fixed; site readiness is not. Tying delivery to a signed-off survey and a completed slab protects both sides when the fit-out runs late, which on retrofit projects it usually does.
Budget the retrofit premium as its own line. A fit-out inside an existing building usually costs more per square metre than the same equipment in a clear shell, because access, protection and out-of-hours work add cost without adding attractions. The build-out cost per park varies enough between programmes that a single benchmark misleads; the useful comparison is your own first unit against your own second.
For a first-time investor reading this while planning unit one, the trend is still useful, just in reverse. The operators above are not competing on novelty. They compete on repetition, which means the design decisions you make now will be copied six more times if the first site works. Specifying for the seventh unit at unit one costs less than retrofitting the standard later. A reference layout for a mid-size park is a reasonable place to start that conversation, and the design-to-delivery process is where the shell survey belongs.
Two related pieces are worth reading alongside this one. The change in how mall landlords recruit family entertainment explains why so many of these units sit inside retail schemes, and the conversion of a former bingo hall is a worked example of what a retrofit constrains. If you are specifying a mixed box, the trampoline range and the materials specification answer questions this article raises but does not settle.
Frequently asked questions
Is a second trampoline park cheaper to open than the first?
Usually, but not for the reason people expect. Design and specification costs fall because the template exists, and the operator already owns a spares pool and an installation relationship. Construction does not fall. A retrofit inside an existing building can cost more per square metre than a clear shell because of access, protection and out-of-hours working. Budget the design saving and the retrofit premium separately.
What size is a typical second-unit trampoline park?
There is no typical size. The six openings in this article range from a 480 m² park in Nowy Tomyśl to a 21,000 m² outdoor site in Malaga. What is consistent is that later units in a dense network tend to be smaller, because drive time segments the market that a larger building would otherwise have to capture on its own.
Can a trampoline park go into a former cinema?
Yes, and JumpYard is doing that in Frankfurt-Griesheim with a 4,000 m² site. The constraints are the ones a cinema creates: raked floors, auditorium volumes, and structural elements positioned for seating rather than for jumping. Clear height is fixed by the building, so the equipment list is written around it rather than the other way round.
Do franchise partners order equipment through the brand or directly?
It varies by agreement. In the licensed openings described here, the brand supplies the specification while the local operator carries the capital and the build. In practice the purchase order comes from the partner and the technical requirements come from head office, which is why delivery should be tied to a site-readiness condition rather than a calendar date.
How much does a repeat unit cost compared with a new build?
No operator in this set publishes its capital cost, so any single figure would be invented. The one public budget comparison here is Sould Park in Malaga, which rose from €15m to €24m during construction on a 21,000 m² outdoor project. Use your own first unit as the baseline and treat retrofit costs as a separate line rather than a percentage of the new-build number.
Planning a second unit, or a first one that has to survive being copied? Send us the shell dimensions and the target opening date, and we will come back with a layout that can be reordered without redesign.
Sources
- Amusements & Attractions, “European trampoline park brand to continue German expansion”, 27 October 2026. amusementsandattractions.com
- JumpYard, park list and company page, accessed 9 November 2026. jumpyard.se
- Amusements & Attractions, “Polish trampoline park to open seventh location”, 27 October 2026. amusementsandattractions.com
- Jump 4U, completed projects list and location pages, accessed 9 November 2026. jump4u.pl
- Amusements & Attractions, “DP Leisure to launch eighth 360 Play location”, 29 October 2026. amusementsandattractions.com
- 360 Play, centres page, accessed 9 November 2026. 360play.co.uk
- Amusements & Attractions, “Second London Prison Island to launch this month”, 6 November 2026. amusementsandattractions.com
- Prison Island, England locations page, accessed 9 November 2026. prisonisland.com
- Amusements & Attractions, “Construction on two Hello Park locations underway”, 23 October 2026. amusementsandattractions.com
- Amusements & Attractions, “Sould Park Malaga set to open before Christmas”, 30 October 2026, citing SUR. amusementsandattractions.com
