Traditional screening asks three questions: what is the score, where is the pay stub, does income hit 3x rent. In a market where 84.3% of operators saw forged pay stubs and 54.5% of residents were born abroad, those three answers get less reliable every year.
Three proxies: a score, a stub and a ratio. None of them measures whether this person will pay you rent next March — they measure how closely a life resembles the shape the form expects. In Miami, most working lives do not have that shape.
Start with the ratio, because it rejects the most people. It assumes a renter who spends about a third of income on housing — a reasonable assumption somewhere. In this metro 63% of renters are already cost-burdened, the highest share in the state that leads the country. The rule is not describing risk. It is describing a population that does not live here.
Then the pay stub. It is not proof of income; it is proof of one particular employment arrangement. Miami starts more than 4,000 new businesses per 100,000 residents, roughly three times the national rate, and the metro leads every large US market in self-employment. Screening reads all of that as a missing document. The money is there. The format is not.
A credit file records borrowing inside one country's system. It says nothing about fifteen years of rent paid on time in Bogotá, São Paulo or Caracas, and nothing about the person who never borrowed at all. In a county built by arrival, a thin file usually means new here, not unreliable. The tape starts the day you land, and a short tape keeps getting read as a bad one.
None of this is an argument for approving more people on less information. It is an argument about what counts as information. Identity confirmed against a source, income confirmed where it arrives, payment behavior observed over time: all three are harder to fake than a page, and all three describe the applicant instead of the applicant's paperwork. The criteria do not move. What you can see through them does.
A screening model is a theory about who pays. When the theory was written somewhere else, for someone else, the false rejections are not an accident — they are the design working as intended.
Sources: RentCafe (Miami rental competition and occupancy, 2025); U.S. Census Bureau, ACS 2020-24 (Miami-Dade demographics and rents); CFPB (credit invisibility).
| What you are deciding | Traditional screening | Grenty |
|---|---|---|
| Who clears the filter | Credit score, W-2 pay stub and 3x income. In Miami that math demands $95,760 a year while the median renter household makes $56,328. | Verified identity, income and payment behavior. The 1099 earner and the newcomer without a US credit file get judged on evidence instead of format. |
| What the evidence proves | 84.3% of operators saw forged pay stubs last year. Document review loses to a PDF editor, and 23.8% of evictions began as fraudulent applications. | Verification happens at the source rather than on the page, which is where fraud has to be stopped: at intake, months before it turns into an eviction. |
| What happens if rent stops | The write-off line. Operators average $4.2 million a year in bad debt, and an eviction runs $3,500 with $2,540 of it rent you never see again. | The rent obligation is backed for the lease term, so unpaid rent becomes a documented claim instead of a number you explain to ownership after the fact. |
| What it costs, and when | Application fees of $100 to $150 per adult buy a report, not an outcome. The real cost of a wrong approval lands months later, on another line. | Rent Protection from $399, paid once for the entire lease. It covers rents up to $2,000 a month, above that pricing varies, and either party can pay it. |
Traditional screening does not measure risk. It measures resemblance to a form.
Count the false rejections. Every model makes two errors. Yours already measures the applicant who failed and defaulted. Almost nobody measures the applicant who was declined and would have paid.
Evidence, not format. A pay stub is a claim about income. The account it lands in is a record of one. Both describe the same money with very different reliability.
The cost lands twice. A wrong no costs days vacant and another turn. A wrong yes costs a write-off. Only one of the two ever shows up on a report with a name on it.
The question stops being hypothetical. Screening asks how likely it is that this rent arrives; a backed application answers what happens if it does not — and that second answer is the one the approval was really hesitating over.
Most declines in the middle of a funnel are not confident decisions. They are a reasonable person choosing the cheaper mistake: a unit that sits another week costs less than a resident who stops paying in month seven. That calculation is what a backing changes, because it puts something documented behind the obligation instead of leaving it as an open risk the office carries alone. The criteria do not move. The consequence of being wrong does.
At the application, not after it. Backing attached when the file is submitted is read as part of the file; backing offered after a decline is a second application carrying the first one's answer. The order is the whole difference, and it costs nothing operationally: the applicant is evaluated exactly the way you evaluate everyone, and what stands behind the rent is simply part of what you are looking at.
Rent Protection is from $399, a single payment that covers the entire lease — never monthly, never charged again mid-term — for rents up to $2,000 a month; above that the price varies with the rent, and either party can be the one who pays. It buys no approval and replaces no one's process. What it removes is the part of the decision that was never about this applicant at all: what the portfolio is exposed to if something changes later.
No. You keep every criterion and the final say on every file. What changes is the evidence available to judge the applicants your current forms cannot read, and what stands behind the rent once you have said yes.
They still pass, and nothing about their file changes. This comparison is not about the top of your funnel. It is about the middle, where a credit score reads the applicant wrong: 26 million Americans have no score at all.
Yes, on every applicant, exactly as you do now. Nothing replaces your screening and nothing asks you to stop reading a score. The report still tells you what it always told you about the people who have a file. What changes is what you can do about the ones whose report comes back almost empty.
Yes, for the people who have one. On a thick US file it is a real signal about payment behaviour and you should keep weighing it. The failure is not the score, it is what gets done with an absence: 26 million Americans have no score whatsoever, and a model that reads blank as risky is guessing.
It helps with the reading and does nothing about the exposure. A wider model still only estimates how likely the rent is to arrive. It puts nothing behind the month it does not. Keep the better model, it is the right investment, and understand that it improves your yes rate rather than what a missed payment costs you.
Keep the criteria and change what sits behind them. The owner is asking for certainty and using the only two dials they know, so hand them a third. On a median Miami rent the 3x rule asks $95,760 a year against a median renter household at $56,328 — that bar screens out most of the market it is applied to.
One conversation, no commitment and no paperwork. Tell us where your rental stands today and we will show you exactly what changes on your side of the table.
Talk to our team →