One definition of revenue
Advances, own-account transfers, refunds and returned items come out before the division, on every file, without anyone remembering to do it.
For lenders
MCA underwriting software: every submission read the same way
Bank statement analysis for MCA lenders. One definition of revenue, one definition of a position, document checks that return the same result every run, and an audit trail on every figure. Your credit policy stays yours.
- One definition of revenue across every file
- Positions typed by cadence, not analyst judgement
- Deterministic checks, repeatable on the same file
- Every number opens onto its source transactions
The problem underneath
Two analysts, one file, two different numbers
Ask two people to read the same statement and they will disagree about what counts as revenue. One counts the owner’s transfer from savings. One does not. One spots the advance that funded on the fourteenth. One reads it as a large customer payment.
Neither is careless. There is no shared definition, so the same merchant reads at 19.2% burden on one desk and 23.1% on another, and nobody notices until a file defaults and gets pulled apart.
A policy you cannot apply consistently is a policy in name only. The gap is not in the credit box, it is in what goes into the ratio before the box is applied.
What you get
Consistency you can point at
The same rules applied to every submission, whoever is on the file and whatever week it arrives.
One definition of a position
Cadence decides. Daily and weekly count toward stacking, monthly fixed counts as debt service. Not a judgement call that varies by desk.
Deterministic document checks
Metadata, balance chain, timestamps, fonts, round-amount deposits and duplicate rows. The same PDF returns the same result every time.
An audit trail by default
Every figure opens onto the transactions that produced it, so a file pulled apart months later still explains itself.
Seats for the whole desk
Analysts and processors work in one private workspace, so every deal that leaves the office was screened against the same standard.
Your data stays yours
Statements sit in private storage for your workspace. No owner names in reports, and nothing is shared between organisations.
How it works
From submission to a comparable read
Statements come in
Text-based or scanned, from any major US bank. Parsed to transactions and daily balances in minutes rather than an afternoon.
The same rules run
Revenue cleaned the same way, positions typed the same way, the same six checks on the document. No file gets a different treatment because it landed on a different desk.
Your policy applies
Check the read against criteria you define and see which sit inside or outside your range. The thresholds stay yours and stay visible.
Where the line is
We surface the picture. You set the policy
There is no approve, no decline, no risk grade, and no model deciding a file in the background. Nothing here overrides your credit box, because nothing here knows what your credit box is.
What we can do is show whether each computed figure sits inside or outside the range you set, without turning that into a verdict. The threshold is yours and stays on screen next to the number.
A note on scope. MetrikData reads bank statements. It does not verify the bank account, check company registration, or price a deal. Underwriting criteria, exposure limits and the funding decision stay with the lender.
FAQ
Common questions
Does this replace our underwriters?
No. It replaces the part where somebody types a statement into a spreadsheet. The read, the exceptions and the decision stay with your team.
Where does merchant data go?
Statements are held in private storage for your workspace. No owner names appear anywhere in a report, and files are not shared between organisations.
Can it run inside our own system?
There is an API for lenders who want the analysis inside an existing pipeline rather than another tab. Talk to us about the shape of the integration.
How do we know a number is right?
Open it. Every figure resolves to the specific transactions it came from, with dates and amounts. Nothing is derived from a source you cannot inspect.
Does it change what ISOs send us?
It tends to. When a broker can see burden and stacking before submitting, the quality of what reaches you moves without a conversation about submission standards.
What does it cost to try?
Send one deal you have already funded. You know what the file contains, so you can check the read against your own conclusion before anything else is discussed.
Keep exploring
Related
Start with a deal you have already funded
You already know the answer on that file. Compare it against the read and judge from there, before we talk about volume or pricing.