Before You Sign

1 · How to read this guide

This guide is a screening method. It exists to help you find out, as cheaply and as early as possible, whether a building deserves more of your time and money. It is not engineering, planning, legal or financial advice, and nothing in it approves a site. The full disclaimer is at §35, and it names the four professionals whose sign-off this method routes you towards.

Every question in this guide resolves to one of three states.

Every number shows where it came from. Figures carry a source tag and a date: published for government and statutory sources, operator for first-party experience from gyms we have built and fitted out, modelled for our arithmetic on stated inputs, vendor for a seller’s figure, and case study for figures from the anonymised project that runs through this guide. Ranges, never point estimates. Where a number could not be verified, we say so, and §32 lists every one.

2 · The site that died on a phone call

A team we know spent months planning a gym.

A 2,700 m² site in regional NSW. Full strength floor, boxing, a sixteen-station rig, a sprint track, group fitness rooms, a creche. A wellness centre with saunas, ice baths, red light, oxygen therapy. The kind of facility that would have been the biggest thing in its region by a distance.

The work behind it was not lazy. There was a catchment model covering more than 100,000 people, built suburb by suburb, every population figure sourced to the census. A construction budget itemised across nineteen trades. An equipment schedule costed line by line. Three financial scenarios, conservative, projected and stress-tested, modelled quarter by quarter for three years. Over two million dollars, planned to the dollar.

The project ended on the electrical question.

Not the catchment. Not the rent. Not the competition, or the demographics, or the interest rate. The building could not be confirmed to carry the load the facility was designed to draw, and getting that confirmation, or closing whatever gap it revealed, was not viable inside the timeline and budget the project had left.

Here is the part that matters. That question can be opened with a free phone call, before you do anything else. It is the first check in this guide and the cheapest check in the whole process.

It was asked last.

3 · Why this keeps happening

Site selection tends to get done in the order things are interesting, not the order they are cheap.

Catchment analysis is interesting. You look at population figures and drive times and growth corridors, and a picture of a thriving gym starts to form. Layouts are interesting. Budgets feel like progress. All of it is real work, and all of it is work you can fall in love with.

Meanwhile the questions that actually disqualify a building are boring, technical, and take a phone call each. So they get left. And because they get left, they get asked at the point where you have already spent months and thousands of dollars, and where you are least willing to hear a bad answer.

This guide fixes the ordering. Six gates, arranged so that the cheapest question that can kill the site is always the next one you ask.

If a site stops at gate 1, you have lost a morning. If it stops at gate 6, you have lost a year.

4 · The six gates, in cost order

You do not move to the next gate until the current one reads SCREEN CLEAR.

That is not a suggestion about tidiness. It is the whole method. A site whose gate 1 questions are UNVERIFIED does not get a catchment model. Not a quick one, not a rough one. The moment you allow yourself to model the catchment of a building you have not checked the power on, you have started falling in love, and everything after that is you looking for reasons to say yes.

Before the gates, two pieces of desk work set everything up: the brief, and the candidate list.

5 · Before the gates: write the brief

You cannot assess a building until you know what is going into it. Most people skip this because they feel like they already know. Write it down anyway, because the specifics are what every gate tests against, and vague specifics produce vague answers.

Write down, in numbers:

• Total floor area you need, and how it splits: strength floor, cardio, functional, group fitness rooms, studio, changerooms, reception, offices, creche, wellness.
• Every powered system you intend to run. Not “some recovery stuff”, the actual list. Saunas, and how many. Ice baths, and whether they are chilled. Red light. Hyperbaric. Oxygen therapy. Air conditioning, and across what area. This list is the single most important input in the guide, for reasons that become obvious at gate 1.
• Your target membership, and the weekly price you intend to charge.
• Your opening date, and how much runway you have to reach it.
• Stage two. What you would add in year three if it works. You will screen the power supply against this, not against opening day, because upgrading later costs multiples of specifying it now.

If you do not know some of these yet, that is normal. Write down the range instead. “Somewhere between 600 and 900 square metres.” “Two saunas if the budget stretches, otherwise none.” A range you have written down is workable; a number you have not thought about is not. What you cannot do is leave the powered-equipment list blank, because that is the input gate 1 tests against, and a blank there becomes an expensive surprise later.

Then write down what you will not compromise on. If a recovery room is the reason your gym is different from the one down the road, then a building that cannot carry it is not a cheaper version of your gym. It is a different business. Knowing which items are load-bearing to the concept stops you talking yourself into a site by quietly deleting the things that made it worth doing.

6 · Before the gates: choose your market, then your buildings

Most people skip straight to looking at buildings, usually near where they live or where they already work. That is how you end up doing brilliant analysis on a site in an area that was never going to support you.

So screen the area first. Four questions, an afternoon each:

• Are there enough people? A rough population count within a 10-minute drive of where you would be looking. You will do this properly at gate 4. Right now you only need to know whether the number is plausible.
• Who is already there, and how many of them? Count the gyms. An area with no gym at all is a warning as often as an opportunity; ask why nobody is there. An area with six is not automatically closed, but you had better be different.
• What does space cost? Get a feel for asking rents per square metre. Rent is the line you can least change later.
• Is it growing or shrinking? New housing estates, approved subdivisions, a council with a growth plan, versus an area losing people.

Pick one or two areas. More than that and you will spread yourself too thin to do the real work.

Then build the candidate list. Three candidate buildings is the minimum for this method to be worth anything, because the method’s value is comparative. With one site, every finding becomes something to work around. With three, findings become reasons to choose.

Where to look beyond the listings:

• Ex-industrial and warehouse stock. Height, floor plate and column spacing that retail cannot match, usually cheaper per square metre.
• Failed big-box retail. Often already has the parking, the loading dock and, occasionally, a supply upgrade left behind by the previous tenant.
• Buildings not on the market. Owners of half-empty industrial stock will often talk. Agents do not list what they have not been asked to.
• Anywhere a gym has closed. Not a bad omen by itself; sometimes the operator failed rather than the site, and the plumbing and power may already be right.

7 · Worksheet 1: The candidate-site list

8 · Why gyms fail on power specifically

Here is the thing almost nobody accounts for: your electrical load is not driven by your gym equipment.

Treadmills and lights are trivial. What actually draws power is climate control across a large floor plate, and the wellness equipment that has become the modern gym’s main point of difference.

Saunas run hot and continuously. Ice bath chillers run continuously, because holding water cold is a constant job, not an occasional one. Hyperbaric and oxygen therapy run compressors. Red light arrays are meaningful in aggregate. And air conditioning across two or three thousand square metres will usually be your single largest load before you have plugged in one piece of gym equipment.

Which produces the trap that catches good operators:

The buildings available at gym rents, older retail, warehouses, light industrial, were supplied for a use that drew a fraction of what you are about to draw. That mismatch does not show up on a floor plan, a rent schedule or a walkthrough. It shows up at the switchboard, and only if you look.

Three different facts, and you must not confuse them. Everything in this gate rests on keeping these separate:

1. What you can observe: the main switch rating. The amps per phase printed at the switchboard, and whether the supply is single or three phase. You can read this yourself on a walkthrough. It is a screening input and nothing more.
2. What only the network can confirm: the connection capacity. What the building’s connection is actually approved and able to deliver. The main switch rating is not proof of this. Only the distribution network service provider (DNSP), or a licensed electrician working with them, can confirm it, and the answer you rely on must be in writing.
3. What only a professional can calculate: your maximum demand. What your fit-out will actually draw at its worst simultaneous moment. That is a maximum-demand assessment by a licensed electrician, done against your equipment schedule, not a number you estimate off this guide.

This guide helps you do a screening comparison between fact 1 and a rough version of fact 3, so you know whether facts 2 and 3 are worth paying for on this site. It never substitutes for them.

9 · What gyms actually pay for power

Real monthly bills from trading Australian sites, not estimates.

operatorOur own sites and sites we work with, as at August 2026. Australian rates; your tariff and climate will move these.

Read the bottom two rows again. Roughly the same floor area. The cheaper one is running two saunas and four chillers. The dearer one is running ducted air.

That is the most useful thing on this page, and it cuts both ways:

For your power bill, air conditioning the whole floor is the decision that matters. It can more than double your energy cost at the same square meterage. Wellness equipment is real but plannable; a fully loaded recovery offering sits inside a bill most operators can absorb. Conditioning two thousand square metres of open gym floor is a different order of commitment, and it deserves to be a deliberate decision rather than something you inherit from a plan or a landlord’s existing fit-out.

For your supply capacity, the opposite is true. Saunas and chillers are what force your amps up, because they draw continuously and they draw at the same time as everything else. Your bill measures what you use over a month. Your connection has to survive the worst minute. That is why the 200 amp rule at §11 is driven by saunas and aircon together, not by the annual bill.

10 · Estimating your load

You are not producing an engineer’s calculation here. You are producing a screening number, good enough to know whether you have a question worth paying a professional to answer, before you pay anyone.

Find what the building has (the observed rating).

Go to the main switchboard and read the main switch rating, the amps per phase, and whether the supply is single or three phase. Photograph it. If you cannot get to the board, the agent, the owner or the outgoing tenant can tell you, and what they tell you stays UNVERIFIED until you have seen it or have it in writing. Remember what this number is: the rating of the switch, observed by you. It is not confirmation of what the connection can deliver. Only the DNSP or an electrician confirms that.

Convert it to indicative capacity.

At 400V three phase, capacity in kVA is roughly amps × 0.693:

modelledArithmetic: √3 × 400V × amps ÷ 1000. This conversion and your building’s actual supply configuration must be confirmed by a licensed electrician; supply arrangements vary, and this table has not yet completed its electrician review (see §34).

Estimate what you need.

Add up the nameplate ratings of every system on your brief. This guide deliberately does not publish per-equipment draw figures. Estimates in circulation are unreliable, and the real numbers are free: manufacturers publish nameplate data, and your suppliers will give it to you in writing if you ask. Ask. A supplier who cannot produce a nameplate rating for a sauna heater or a chiller is telling you something.

Then apply a diversity factor, because not everything runs flat out at once. For screening, 0.7 to 0.85 of connected load is a reasonable range, and you should use the top of it, because this is exactly the number an operator who wants the site will quietly set too low.

modelledScreening range only. Your electrician sets the real one in the maximum-demand assessment.

Then compare like with like: capacity is in kVA, load is in kW, and kVA × 0.9 ≈ usable kW is close enough for screening. Skipping that step flatters the site by 10%.

modelledElectrician confirmation pending, see §34.

Finally, add headroom for stage two. Screening for opening day and upgrading in year three costs multiples of specifying it now.

Where the load actually sits, in rough order of size: air conditioning across the floor plate, saunas, water chillers, then lighting, fans, compressors and equipment. If your brief has three or more continuous wellness loads plus full HVAC, treat the supply question as open until the professionals close it.

Reading your result:

• Estimated load comfortably inside indicative capacity: SCREEN CLEAR on this question. Not an approval. The electrician and the DNSP still confirm the real numbers before you sign anything.
• You could not read the board, or a number came from someone else’s say-so: UNVERIFIED. Go and verify it.
• Estimated load close to or above indicative capacity: STOP AND ESCALATE. Indicative shortfall. Verify before acting. The named professionals are a licensed electrician (maximum demand) and the DNSP (connection capacity, in writing). Spend nothing else on this site until they have answered.

11 · The 200 amp rule

That is roughly 139 kVA, or about 125 kW of usable load, and once you have covered HVAC across a real floor plate plus a couple of sauna heaters, you have spent most of it before the lights go on.

Use it as a screening threshold, not a target, and not an engineering figure. Walk into a building with a 100A main and a brief that includes climate control and a sauna, and you already know the conversation you are about to have. You have not sized anything yet, but you know the site needs an upgrade conversation, and you can go and price that before you fall in love with the floor plan.

And 200A is a floor, not a spec. A large facility with several wellness modalities, a full HVAC system and a creche will need considerably more. The number that gets you past the screen is not the number that runs your gym.

The rule is about your brief, not about gyms in general. A modest 24/7 site with cardio, weights and split-system air conditioning is a different electrical proposition to one with saunas and plunge pools. If your brief has no continuous heating or chilling loads in it, you have more room. But read this page again the moment you start sketching a recovery room, because that is the decision that changes the answer.

operatorSource: our own experience building and fitting out gyms, as at August 2026. Not a published standard, not an engineering figure. A site under the threshold reads STOP AND ESCALATE, not condemned; a site over it reads SCREEN CLEAR on this question only. Your electrician sets the real figure either way.

12 · The solar trap

Operators get this wrong constantly, and it is worth being blunt about.

Solar reduces your electricity bill. It does not raise the import capacity your building is approved to draw from the network.

Be precise about this, because the sloppy version is also wrong. A grid-connected inverter is a supply at your switchboard, and it does feed your load while it is generating. Switchboards have to be assessed for the combined current from the grid and the inverter, which is a real engineering consideration. What solar is not, on its own, is firm capacity. At night, in winter, and on any day the array underperforms, you are back on the connection you were always limited to. And your approved import limit does not move because you put panels on the roof.

So: a large array is an excellent answer to the operating cost problem, and given the size of that line it belongs in your plan. Solar plus properly designed storage and controls can also shave peaks and genuinely change what a site needs. But that is an engineered demand-management design that your electrician specifies and the connection rules have to approve. It is not something you assume on the strength of a solar quote.

Two different problems, consumption and capacity, one more time. Do not let a panel salesperson talk you out of the supply question.

13 · The call to make

If your estimated load looks close to or above the building’s indicative capacity, contact the network distributor before you do anything else. In NSW that is Ausgrid, Endeavour Energy or Essential Energy depending on where you are; every state has its equivalent, and the building’s electricity bill will name it.

Be realistic about what a phone call gets you. It gets you triage: what supply the site has, whether spare network capacity is conceivable in that area, and which pathway your request falls into. That is genuinely valuable and it is free.

What it does not get you is a costed, timed answer. The simple connection pathways distributors offer top out far below gym scale. A real answer on capacity, cost and timing comes from a formal connection enquiry or application, and in NSW an Accredited Service Provider usually assesses the site, designs the work and deals with the distributor on your behalf.

This guide publishes no upgrade costs and no upgrade timelines, deliberately. Connection costs are site-specific and network-specific, and any figure we printed would be a guess dressed as data. What you need is not our number. It is a written DNSP assessment for your site. Get it before you sign.

So run it in two moves:

1. The call. What is the existing supply? Is additional capacity plausible here, or is this area constrained? What process would an upgrade go through?
2. The enquiry. Lodge it. Expect weeks, not minutes, and possibly a fee. Ask for capacity, cost and timeline in writing, and if a substation is required, where does it physically go? A padmount needs space, access and usually an easement. That is a change to your site plan, not just a line in the budget.

Weeks sounds slow. Set it against the alternative: the team at §2 spent months building a catchment model and a financial structure for a building whose supply was never confirmed. The slow version of this step is still the fastest way to find out.

Get the substantive answer in writing and keep it with the lease documents. A verbal “should be fine” is not an answer, and the agent giving it carries none of the risk. Until the written answer exists, this question is UNVERIFIED, whatever anyone has told you on the phone.

14 · The rest of gate 1

Same discipline, less detail, and each of these has ended a gym.

Water and sewer. Showers at your peak hour, plus ice baths, plus a laundry. Ask about incoming water main size and pressure, and about the sewer connection. If you are discharging anything unusual, ask the water authority about trade waste approval. Recovery facilities move real volumes of water and it is not always a simple connection.

Floor loading. Dropped weights are an impact load, not a static one. A full rig is a concentrated load. If you are planning anything above ground level, a mezzanine or a first-floor strength area, this is a structural engineer’s question, not a judgement call, and it reads STOP AND ESCALATE until the engineer has answered. In the case study, the fit-out budget carried $110,000 of fabrication for a mezzanine (case study), which is exactly the scale of spend that makes the structural question worth answering first.

Also worth asking now, while you are on the phone: gas availability if you need it, internet and fibre.

15 · Gate 2: Planning and permitted use

A phone call and a search, and it can end a site as surely as the switchboard can.

Permitted use. Ask the council whether a gym is permissible in that zone. If it needs a development application, ask how long a comparable one has taken recently. A DA that runs longer than your runway is a blocked site even if it would eventually be approved.

Conditions. Ask whether there are hours-of-operation restrictions on the zoning. A 24/7 model dies quietly on a condition of consent.

Who verifies this gate: a town planner or building certifier. The council counter gives you screening answers; the planner gives you the answer you can rely on, including the realistic DA pathway and the conditions that will attach. If the council answer is anything other than a clean “permissible without consent”, or the pathway matters to your runway, this gate reads STOP AND ESCALATE until the planner has looked at it.

16 · Worksheet 2: The phone-call gate

17 · What to measure when you are standing in it

Only sites whose phone-call gate reads SCREEN CLEAR get visited. Bring a tape measure and the worksheet.

Ceiling height at the lowest point, not the highest. Ducting, sprinklers and beams set your real clearance. A rig needs height for pull-ups with a tall member on the bar; rope climbs and any overhead work need considerably more. Measure under the obstruction.

The column grid. Columns are the most common reason a floor plan does not work. Measure the spacing and mark them on your layout before you fall in love with a zone plan. A column in the middle of where the platforms go is a real constraint, not a detail.

The floor plate shape. Long and narrow behaves very differently to square. Consider sightlines from reception; a shape you cannot supervise costs you staff hours forever.

Parking. Count the spaces and compare against your peak hour, not your average. Then check whether they are exclusive to the building or shared with a whole centre. Shared parking at 6pm is not parking.

Loading and access. How does equipment get in? Roller door dimensions, truck access, forklift room. Equipment that has to be craned or disassembled costs real money on install day.

Amenities and shell condition. What exists already, and what you are building from scratch. Existing changerooms in the right place are worth a lot; in the wrong place they are worth less than nothing.

Read the switchboard yourself. This is where the observed main switch rating from §10 stops being someone else’s say-so. Photograph the board. And note where it is, because if you need an upgrade, the run from the board to your equipment is part of the cost.

The floor itself. You cannot verify floor loading by standing on it. What you can do is note the construction (slab on ground, suspended slab, timber), and whether your plans put weights anywhere other than slab on ground. If they do, that line stays with the structural engineer from Worksheet 2.

18 · Worksheet 3: The walkthrough checklist

19 · Drive time, not distance

Now, and only now, you find out whether enough people live near the site.

Measure your catchment in drive time, not in kilometres. Ten kilometres up a highway and ten kilometres through congested suburban streets are not the same catchment, and your members experience the drive, not the map. A gym competes on convenience above almost everything else.

Then tier it. This is the part most people get wrong. They pick one radius and apply one penetration rate across it, which assumes someone twenty minutes away is as likely to join as someone three minutes away. They are not, because they have something closer.

The method: three rings, declining penetration.

operatorOne experienced operator’s working rates, not a published standard. A starting point to be argued with, and gate 5 is how you argue.

Build the table. List every suburb inside each ring. Get each population from the census: search the suburb name plus “census QuickStats” and use the official figure. Write the source next to every row. That sourcing is what makes the number defensible later to a landlord, a bank or an investor, and it takes seconds at the time and hours to reconstruct afterwards.

Make your rings mutually exclusive, 0–10, over 10 to 15, over 15 to 20, so no suburb is counted twice.

Multiply, sum by ring, then sum the rings. Do it for both columns.

20 · Worked example

The case study’s catchment, built exactly this way:

Every suburb population sourced to the 2021 Census, row by row.published(populations)case study(rates and totals)

This is good work. It is the right method, executed properly, and the output is a defensible range rather than a single optimistic number.

It is also not yet an answer, and §21 explains why. Remember where this model sits in the story: it was finished, and defensible, on a site whose gate 1 questions were still unverified.

21 · There is no national number

Here is the honest problem with every catchment model, including the good one at §20: the penetration rate is asserted.

Where does 3% come from? Why not 2%, or 6%? A number with no basis cannot tell you whether your projection is conservative or fantasy, and the gap between those two is the whole business.

You might expect to fix this by comparing against a national gym membership rate. We went looking for one.

Which is a good reason not to borrow anyone’s national average. Count your own catchment’s paying market directly. That is gate 5, and it is what turns the catchment model from an assumption into an answer.

22 · Two adjustments

Growth, counted separately. If suburbs in your rings have committed residential development, that matters, but count it as a separate line, not blended into current population. Check the council’s DA register and land-release maps, and your state’s population projections. Committed dwellings under construction are a reasonable bet; rezoning proposals are not the same thing and must not be counted as though they were.

Traffic is branding, not catchment. A busy road past your door is genuine marketing value. It is not members, and it must not enter the membership arithmetic. The case study handled this correctly and it is worth copying: the road carried 20,000 to 25,000 vehicles a day (case study), and the model recorded it as visibility upside, explicitly excluded from the member count. Treating passing traffic as catchment is one of the most common ways operators talk themselves into a site. Write the number down. Keep it out of the total.

23 · Worksheet 4: The catchment rings

24 · Count who already pays

List every gym inside your rings. All of them: the big-box franchises, the independents, the boutiques, the CrossFit boxes, the ladies-only studios, the council leisure centre. Anywhere a person in your catchment currently pays to train.

Estimate each one’s membership yourself, from field observation. You will not get exact numbers and you do not need them. These are your estimates, and this guide never publishes membership figures for named gyms; the number is always yours, always a range, and always labelled as your estimate.

Start from footprint, then adjust for the model:

operatorDensity ranges from gyms we have built and fitted out, as at August 2026. A 700 m² 24/7 franchise running around 900 members and an 1,800 m² full-service site running around 800 are both normal; the difference is the model, not the market.

Then correct your estimate with what you can see. Class timetables tell you a lot. Car park counts at 6pm on a Tuesday tell you more. Staff talk, and so do members: walk in, take a trial, ask how busy it gets.

Enter every estimate three times: low, base and high. Five confident guesses summed together still produce a guess, and the honest way to carry that is a scenario range, not a single number. The worksheet at §27 is built for exactly this.

Sum each scenario. The base total is the number of people in your catchment who already pay someone to train. Not a national average. Your own local market, counted.

Then do the division that matters:

Be precise about the denominator. It is the locally served paying market, the people inside your rings who currently pay a gym, not the ring population, and not a national statistic. Dividing by population flatters every projection; dividing by the served market is what makes the number mean something.

Now it can be argued with. If the rings already support 4,000 paying members across six facilities and you are projecting 2,300, you are planning to take more than half of everyone in the area who currently pays anyone, while charging more than most of them. That is a claim that has to be defended, not assumed.

One honesty about this method. It assumes you grow no new demand, which is conservative by design: a genuinely differentiated facility does pull in people who were not paying anyone. Run the division anyway. If the answer is only survivable on demand you have to conjure, you should know that before you sign, not after.

25 · The ceiling check

AusPlay’s 36% is worth one comparison. Split your projected membership across your rings, convert each ring’s population to adults 15 and over (QuickStats publishes the age breakdown), and check whether any ring’s share approaches 36% of them. If it does, you are projecting close to every person in that area who does any gym-type activity at all, including the ones running in parks and doing home workouts.

Treat that as a signal to go back over your assumptions, not as a pass or fail. It is a national survey estimate with sampling error, your local mix can differ from it, and it measures activity, not willingness to buy your membership. It is a ceiling, never a base: a projection under the ceiling proves nothing, and a projection over it is almost certainly wrong.

publishedASC AusPlay, survey period July 2023 to December 2024. Refresh from the January to December 2025 tables before print, see §33.

26 · The price gap question

While you have the competitor list open, write down what each one charges per week. Then write down what you intend to charge.

Look at the difference and answer, in a sentence you would be willing to say to a bank: what does a member get for that difference, and how many of them will pay it?

“We’re much bigger and we have wellness” is a reasonable answer. It is not automatically the right answer, because the person deciding is comparing a weekly price against a gym they can already see from their house.

27 · Worksheet 5: The competitor scenario table

28 · The money is a link, and one piece of arithmetic

Gate 6 is a week of work with advisors, and most of it is already built for you. The full cost stack, fit-out, landed equipment, soft costs, contingency and runway, lives in the cost-to-open calculator, with every rate sourced and dated. The freight and customs line that surprises direct importers lives in the landed cost calculator. This guide does not duplicate them; Worksheet 7 hands you off.

But one calculation belongs here, because it should be done before the lease is signed and usually is not:

Then ask two questions of that number.

How long will it take you to sign that many? Be honest about the ramp. Members arrive over months, not on opening day, and the curve is slowest at the start when nobody has heard of you. Your presale is the only real evidence you will have before you commit: run it, and treat what it actually delivers as your forecast, not what you hoped it would deliver.

Can your runway survive that long? Budget for six to twelve months before you are profitable, longer for a large site, because the cost base lands on day one and the members do not.

operatorOur experience across real openings, as at August 2026.

What this looked like in the case study. The plan’s own figures: quarterly operating costs at opening of $350,000 against revenue of $341,250 at 750 members paying $35 a week, so breakeven sat around 770 members. In the stress scenario, ramping 600 to 800 members across the first half-year at $32.50 a week, the site burned $148,500 across six months while it filled.

case studyBudgeted figures from a facility that never opened, not trading results.

None of that is a red flag on its own; losing money while a big site fills is the normal shape of opening a gym. The question the plan has to answer is simply whether that burn is sitting in the bank on opening day, and for how many months. A number you have not turned into a sentence is a number you have not understood.

Who verifies this gate: your accountant or advisor on the model, and a commercial lease lawyer on the lease itself, before anything is signed. Make the lease conditional on the gate 1 and gate 2 answers you are still waiting for; your lawyer drafts that condition, and it is the single cheapest insurance in this entire guide.

29 · Worksheet 6: The site scorecard

30 · Self-audit

The site you are already looking at. Honestly. Tick only what is actually done.

Any box unticked above the line you are currently working on means you are working on the wrong thing. An unticked box is an UNVERIFIED state, and an unverified question is not a cleared one.

31 · Worksheet 7: The breakeven handoff

32 · What we could not verify

This page is a feature, not an apology. A method is only as honest as its gaps, and these are ours.

Australian gym membership penetration. No credible published figure exists. Membership claims in circulation (3.4 million to 8.5 million, a 2.5 times spread) trace to unattributed vendor blogs and auto-generated statistics sites; the most cited industry estimate is a paid report whose methodology is not public. This guide therefore derives the served market locally (§24) and uses AusPlay’s properly sourced participation figure as a ceiling only (§25). We will not reintroduce a national membership percentage without a named primary source.

Network upgrade costs and timelines. We hold no real quotes and no published connection-charge schedule we can stand behind, so this guide publishes none. The method is the substitute: a written DNSP assessment for your specific site (§13).

Per-equipment electrical draw figures. Circulating figures are estimates of unknown quality, so this guide does not print a load table. The method is the substitute: supplier nameplate data, in writing, summed by you and verified by a licensed electrician (§10).

The case study’s measured electrical supply. We do not have a verified measurement of what the failed site’s connection could deliver, which is why §2 says the building could not be confirmed to carry the load, rather than stating a figure. The lesson survives the gap: the confirmation was never obtained, and it was the cheapest item on the entire project plan.

We publish where our numbers come from. Where a number does not reliably exist, we say that too, rather than inventing one.

33 · Sources and provenance

Every figure in this guide, its tier, its source and its date. Rates as at 2026-08-03.

34 · Review status

This guide is written to be safe before these reviews complete: it publishes screening methods and questions, not engineering figures, and every verdict is a screening state. The reviews harden it. Each is listed with exactly what its professional must sign off.

35 · The site-selection disclaimer

This is a site-selection disclaimer, specific to this guide’s method. It also appears in short form beside the load estimate (§10) and the scorecard (§29).

36 · Run your shortlist through the gates

Want a second pair of eyes on a shortlist? Get the guide by email below, then reply to that email with the addresses and where you are up to. We read every reply.

We already publish the two calculators that cover the money end of this:

• gymfactories.com/tools/cost-to-open
• gymfactories.com/tools/landed-cost-calculator

The interactive version of this method is live: run your shortlist through the gates in the site selector. Two to five candidate sites, the same three screening states, every assumption sourced — and everything stays in your browser until you ask us for a review.

gymfactories.com

We publish where our numbers come from. Where a number does not reliably exist, like Australian gym membership penetration, we say that too, rather than inventing one.