Most declines were knowable before the deal went out. A submission fails for one of two reasons: a credit problem, or a mismatch between the file and the lender's rules — and the second kind is preventable with about ten minutes of checking. These are the nine checks, in the order they kill deals.
The checklists that rank for this question are document lists — bring the application, bring the bank statements. Documents are the easy half. What follows is a decision checklist: the questions that decide whether this file belongs at this lender at all, asked before the credit pull, not after the decline.
A lender's score floor only means something on the bureau they pull, and the same borrower can sit forty points apart across the three. If your desk quotes a 690 from one bureau and the lender's floor is a 680 on another, you do not know whether this deal passes — you know it passes on the wrong report. Find out which bureau the lender pulls, get the score on that bureau, and compare against the floor. A file declined this way comes back marked "credit" when nothing about the credit was wrong; the number was measured with a different ruler.
Three lenders will compute three different TIB figures from the same file, because they measure from different dates: the state filing, the first business bank activity, or industry experience under a prior entity. A borrower who "has been in trucking eleven years" may be fourteen months old to a lender that counts from the LLC filing. Compute the number the way this lender computes it, and if it lands under their minimum, check whether they run a startup tier underneath — and what that tier trades away in deal cap and conditions.
Exclusions are hard stops, not negotiating positions. Long-haul trucking over a certain fleet age, restaurants, gyms, anything with a title problem — every lender keeps a list, and a file that touches it is dead on arrival no matter how strong the credit is. The list is written down on the program sheet. The check is thirty seconds; the decline is a week.
Private-party sales, older equipment, high mileage, refinances, and sale-leasebacks are all structures that some programs allow and others refuse outright. A lender that funds dealer sales all day may not touch the identical asset sold by a private party, because there is no dealer standing behind the title and the invoice. If the deal has any structure beyond "dealer sells new-ish equipment to established business," confirm the program allows that structure before anything else.
Every program has a size band, and every app-only program has a dollar ceiling above which the file needs a full package. Submitting $180,000 app-only to a desk whose app-only ceiling is $150,000 does not get you a decline — it gets you a stip list you were not ready for, a stalled file, and a borrower who wonders why the fast approval you promised is on day six. If the amount is above the ceiling, submit the full package on day one or pick a lender whose ceiling clears it.
If the program requires bank statements, read them the way the credit desk will: NSFs in the last ninety days, average daily balances against the proposed payment, and existing daily or weekly withdrawals that signal an MCA already feeding on the account. A payment the account visibly cannot carry is a decline you could have read yourself. This is also the check that most often converts a decline into a restructure — a longer term, a bigger down payment, or a co-signer — before the file has a fresh decline stamped on it.
Many programs want to see that the borrower has previously paid something comparable — an installment obligation in the neighborhood of the proposed payment. A borrower whose largest tradeline is a $400 car payment asking for a $3,000 monthly equipment payment is a comparable-credit decline at a desk that checks for it, whatever the score says. If the comp is not there, that points the file toward programs that do not require one, usually at a smaller size or with more money down.
Geography is the dullest check on the list and it still kills files. Some lenders are not licensed in every state; some have state-specific rules on rate or structure; a few will not fund certain states at all. It is one line on the program sheet, and it is a check that fails silently — nothing about the file looks wrong, the borrower just lives somewhere the lender does not do business.
An approval is not money — the stips are the distance between the two. Before submitting, ask whether this borrower will sign a personal guarantee, can show insurance, and can produce the down payment without draining the account the lender just evaluated. A borrower who balks at the PG after approval wastes the approval, the lender's patience, and your credibility at that desk. Ten seconds of asking up front is cheaper than an approval that dies in funding.
A failed check is a placement signal, not a dead deal. Every check that fails points somewhere: a different bureau points to a different lender; a TIB miss points to a startup tier; a structure refusal points to a program that allows it; thin bank statements point to a restructure before submission. The deals that die are the ones submitted first and checked afterward — each decline stamps the file, and by the third one, desks that would have funded it on day one now want to know why two others passed.
None of this is secret. Every answer lives on a rate sheet, a program sheet, or in a credit policy the lender will happily explain — the work is collecting those rules once, keeping them current, and actually checking the file against them every time, rather than finding out on the decline.
Dealbrace keeps the written rules for every desk you send to and checks a deal against them before it names anyone. Out is out.
Open the live demo