PouchBook a demo
Industries · Business banking lending teams

Flag trouble six weeks early, from data you already hold

Pouch turns the categorised transactions you already hold into a live free-cashflow position and a daily forward forecast for every business borrower.

Arden Consulting Ltd

working capital · £40,000
£0

Flag · fired on the forecast

Projected free cash crosses £0 in 6 weeks.

Current position still positive · no payment missed yet

01

The problem

You hold the accounts, and still lend off snapshots

You hold your business customers' accounts, the richest data in the lending market, and still underwrite off snapshots and annual reviews.

Nov 4 · 09:12 · the first you hear of it

Direct debit returned · unpaid

Arden Consulting Ltd

−£3,600

loan repayment 7 of 24 · insufficient funds

The same borrower · your own core banking

the decline began in July · and it was all in the account

02

How it works

Static at origination, Live for the life of the loan

Step 01

Push the transactions you already hold

Send categorised transactions to Pouch's endpoints: a fixed history at origination, then continuously as they land.

CLIENT PAYMENT · INV-311+£4,850.00

SUPPLIER DD · MERIDIAN−£980.00

STUDIO RENT−£1,900.00

PAYROLL RUN−£3,200.00

POST /v1/transactions · 202

38,000 transactions a year · already in your core banking

zero new data to collect

Step 02

The engine resolves the position

Pouch detects every recurring inflow and outflow, excludes internal transfers, and resolves true free cash: income minus commitments, on any day behind the borrower or ahead of them.

Arden Consulting Ltd · 12-month average

Deposits into the account£22,400 / mo
Transfers · director's own account name-matched−£2,100
VAT refund · one-off HMRC · Jun 12−£1,450
Loan drawdown Iwoca · Sep 3−£1,250
Trading revenue£17,600 / mo

Committed · detected cycles

Payroll · 28th · monthly−£8,400
Rent & rates · 1st · monthly−£2,300
Suppliers · recurring−£2,700
VAT & PAYE · quarterly, averaged−£1,900

True free cash

£2,300 / mo

every line resolves to its source transactions · auditable

Step 03

Static output at origination

You get the itemised picture: true income, mandatory and discretionary outflows with internal transfers stripped out, historical daily balances, and a forward daily forecast across the repayment schedule.

Twelve repayments · one seasonal business

A flat schedule

fails in the off-season ✗

Shaped to the forecast

clears all twelve ✓

month 8 · off-season · instalment halved

not a decline · a loan that fits the business · PDF or JSON

Step 04

Defend the book, then grow it

The same rolling forecast that flags deterioration weeks before a missed payment also shows capacity: set your own criteria, and when a borrower's free cash holds above them you see the room for more lending before a competitor funds it.

Defend

Arden Consulting Ltd

Flagged on the forecast · 6 weeks before the bounce.

Restructured before the miss

flag · wk 14

bounce · wk 20

Grow

Fenwick Joinery Ltd

Forecast free cash ≥ 2× repayments for the next 4 months · your rule.

Offer a £15,000 top-up

one feed · defended and grown

03

The demo

What you will see in the demo

01A flag that fires on the forecast
02The transfer reveal
03Cash on the day repayment falls due
04Your data, in your environment

04

FAQ

The questions your team will ask.

We already have internal monitoring. What does this add?
Balance and arrears alerts fire when it is already happening. Pouch flags on the forward position, the projected dip, which is the difference between restructuring early and chasing arrears. It sits beside your existing monitoring as an additional layer, not in place of it.
Couldn't we build this ourselves?
The frequency, classification and forecasting engine is the hard part: years of iteration on real cashflows. It runs as a layer in your environment, on your categorisation, and everything your teams and customers see stays yours.
Where does this sit with compliance if we act on the signals?
Pouch provides descriptive information only: positions, trends and dates, never a score, rating or recommendation. Decisions and the governance around them stay with you. The instrument panel, not the throttle.
How accurate is the forecast?
We claim the resolution: a daily value on any future date, not a coarse multi-month band. Accuracy is yours to judge, which is what the pilot is for: it can replay a historical loan month by month, showing when the Live flag would have fired against when the trouble actually surfaced, measured on your own book.
Next step

See it on your own data.

Book a demo with your business lending team, then judge it on your own book.