Yep — this is the right move. Once you start listing variables, you stop “arguing a vision” and start designing something that can survive reality.

Below is a complete variable map + a practical allocation trigger based on rental income (exactly what you asked: “how many rentals before giving the first member home?”).

1) Core variables you must model
Property economics

Purchase price (P) per home (incl. fees, refurb, legal)

Rent (R) per home per month (modest, below market)

Occupancy rate (Occ) (e.g., 95–98% long-run)

Arrears/bad debt (Ar) (% of rent lost)

Rent growth (gR) (could be 0 if you cap rents; but costs still inflate)

Operating costs (per home per year)

Insurance (Ins) (buildings + liability; contents is tenant responsibility)

Routine maintenance (RM) (minor fixes, servicing)

Lifecycle reserve (LR) big-ticket items averaged: roof, boiler, windows, electrics, kitchens/bathrooms

Compliance & safety (Comp) gas certs, electrics, smoke/CO alarms, EPC work

Management/admin (Mgmt) (even if volunteer-led, you’ll still have real costs)

Void/turnover costs (Void) cleaning, re-letting, small repairs between tenants

A useful simplification is:

OPEX per home per year = Ins + RM + LR + Comp + Mgmt + Void

Capital preservation

Contingency reserve (CR) for shocks (mould litigation, structural issues, major works spikes)

Inflation of repairs (inf) (repairs inflate differently than CPI sometimes)

Concentration risk (too many same-build homes = correlated failure)

Membership dynamics

Members (M) and monthly contribution (C)

Attrition rate (d) (people leaving)

Transfer/waitlist mechanism (to replace leavers without draining the pool)

Eligibility rules (points, tenure, hardship rules)

Fraud / gaming risk (people joining just before allocation, etc.)

2) The one equation that decides when “rentals can support allocations”

The key is: once you start allocating homes to members at near-zero rent, you remove rent income. So you only do it when the remaining rentals can pay for the allocated homes too.

Define:

Net rent surplus per rental home per year

S=(R×12×Occ×(1−Ar))−OPEX
S=(R×12×Occ×(1−Ar))−OPEX

Support cost of a member-allocated home per year (because it still needs insurance/maintenance)

K=OPEXmember
K=OPEX
member
	​


(very similar to OPEX, maybe slightly lower without lettings/voids)

Then the rule for sustainability is:

To allocate 1 home, you need enough rental homes so their net surplus covers the support costs:

Nrent≥KS
N
rent
	​

≥
S
K
	​


That’s the clean “how many rentals before the first allocation?” answer.

3) A sensible allocation trigger based on “what that first allocated home used to earn”

You suggested: base it on the amount of rent the first allocated house was making.
That’s actually a good intuitive anchor, but you must base it on net surplus, not gross rent.

Trigger A (conservative, simple):
Allocate the first member home only when the portfolio has enough rentals that:

The lost net surplus from removing that home from renting, plus the ongoing support cost of that allocated home, is covered by the surplus from the remaining rentals.

In symbols, allocating one home removes roughly S (its surplus if it stayed a rental), and adds K in costs you still must pay.

So required surplus coverage becomes:

(Nrent−1)×S≥K
(N
rent
	​

−1)×S≥K

Rearranged:

Nrent≥1+KS
N
rent
	​

≥1+
S
K
	​


That “+1” is exactly your intuition: you replace what the allocated home used to contribute.

Trigger B (even safer):
Same as above, plus require a portfolio-level reserve floor:

CR ≥ X years of OPEX on allocated homes (e.g., 2 years)

So you don’t collapse if a roof fails the year after you allocate.

4) Concrete example (just to see scale)

Say modest numbers (illustrative, not “true UK values”):

Rent R = £700/mo

Occ = 0.97

Ar = 0.02

OPEX = £2,800/year (insurance + maintenance + compliance + admin + void average)

Then:

Gross effective rent ≈ 700×12×0.97×0.98 ≈ £7,980/yr

Surplus 
S≈7,980−2,800=£5,180/yr
S≈7,980−2,800=£5,180/yr

Support cost for an allocated home 
K≈£2,800/yr
K≈£2,800/yr

Then:

K/S≈0.54
K/S≈0.54

Trigger A says:

Nrent≥1+0.54⇒2
N
rent
	​

≥1+0.54⇒2 rental homes

So in this toy case, you could allocate very early.
But here’s the 5-steps-ahead catch:

That only works if OPEX assumptions are realistic and you have reserves. Real lifecycle costs can be nastier (boilers, roofs, damp remediation), so the actual K might be much higher once you model long-run capital works properly. When K rises and S falls, the required rentals jumps.

That’s why the rule matters: it scales automatically with your real data.

5) The practical policy I’d recommend (simple + robust)
Phase 1: Build-only (no allocations yet)

Buy properties, rent them all.

Build reserves until:

CR ≥ 2 years of total OPEX for the current portfolio

and you have at least (say) 20 rentals to diversify risk

Phase 2: Start allocations slowly

Allocate one member home per year only if all are true:

Rental coverage (Trigger A):

(Nrent−1)×S≥K
(N
rent
	​

−1)×S≥K

Reserve coverage:

CR≥2×(AllocatedHomes×K)
CR≥2×(AllocatedHomes×K)

Stress test (simple):
Assume Occ drops to 90% for a year and a major repair hits—still solvent.

This prevents “one bad year” collapse.

Phase 3: Ratio-based steady state

Move to a rule like:

Maintain at least X rental homes per allocated home (X derived from your real S and K)

Keep CR proportional as above

Let points/queue control who gets allocated, but the ratio rule controls when

6) What you should do next (lowest-effort high-value)

Make a spreadsheet (or I can structure it for you) with these inputs:

R, Occ, Ar

Ins, RM, LR, Comp, Mgmt, Void

CR target (years)

Allocation policy (Trigger A/B)