The Permission Layer
Find landlords who'll actually say yes.
Revenue data is everywhere now. The scarce, valuable signal in rental arbitrage is knowing which landlords permit subleasing before you spend a week cold-calling. Stay Stax generates that as living, transparent, compliance-checked data, with the basis for every signal shown in the open.
The problem
Permission is where arbitrage operators stall out.
Finding a unit with good numbers is the easy part. The wall most operators hit is the landlord conversation: most leases are silent or hostile on subleasing, and you can burn weeks of calls to find the few owners open to it. Stay Stax's view is that the permission graph, who allows this and who does not, is the real moat in this business, and it is exactly the part no public revenue dataset covers.
The tiers
Five tiers, and the honesty is in the basis.
Every landlord carries a permission tier. What makes it trustworthy is that each tier shows whether it is Observed (a real outcome on record) or Modeled (an estimate from owner signals). A modeled LIKELY is a lead to work, never a confirmed yes, and the interface never lets one stand in for the other.
- CONFIRMEDObserved
A real sublet outcome on record, with the date it was verified.
- LIKELYModeled
Modeled from owner signals. A lead worth working, never sold as a yes.
- UNKNOWNNo data
No signal yet. Worth a call, with eyes open.
- UNLIKELYModeled
Owner signals lean no. Deprioritize before you spend the week.
- PROHIBITEDObserved
A documented no: a lease clause, a stated policy, or a prior refusal.
Observed means it is on record (a real outcome, with a date). Modeled is an estimate from owner signals, never presented as a confirmed yes.
Living, not static
Permission decays, so the signal has to stay alive.
Owners sell, management companies change policy, a building flips from permissive to prohibited overnight. A list assembled once and sold behind a paywall is stale by the time you buy it, and you have no way to tell which entries still hold. Stay Stax's signal ages on purpose: a CONFIRMED carries the date it was verified and loses confidence over time until it is re-verified, so freshness is part of the data rather than an assumption.
Compliance built in
A yes is worthless if the parcel can't legally host it.
Permission and legality are two different gates, and you need both. Stay Stax checks parcel-level zoning and tracks ordinance changes per market alongside the permission signal, so a willing landlord in a banned zone gets flagged before you waste the conversation. This is the product side of the same shift we lay out in our market thesis: the durable edge is in commercial and mid-term models, in zones and stay-lengths that often sit outside residential STR bans, verified market by market.
The flywheel
Every outcome sharpens the signal, without sharing your contacts.
When you log how a landlord conversation went, that outcome improves the model for everyone. The boundary is firm: your contacts stay yours. Only a de-identified permission signal is shared, never the landlord's identity, their contact details, or your relationship with them. The graph gets smarter; your book stays private.
Work the landlords most likely to say yes.
Stay Stax points you at permission you can act on, screens the deal with conservative math, and writes the pitch. Start mapping permission in your market.