storefront openings tracked

Storefront risk, measured.

Restaurant, payment and public data, combined at the level of one store.

Every store, in three layers.

Public records, card sales and POS operations meet at one address.

Three layers. One dataset.

Matched by business name and address. Public layer live today; card and POS layers from 2027.

Test decisions on accurate data.

Checked on years the model never saw, in two cities.

62.6%Vancouver AUC
64.4%Toronto AUC
1.7×riskiest 10% closure rate
Request a demoBy invitation

How it works

Three kinds of data, joined at one store.

Public data · live

Openings, closures and the street around a store

Card data · from 2027

How sales move, with merchant consent

POS data · from 2027

What happens inside the store

How the score is calculated

From public records to a risk band.

1 · Inputs

What we look at

Business age and industry, staff count, nearby competition, foot traffic and transit access, and recent closures in the area.

2 · Model

How it learns

A statistical model trained on storefront openings from 1998 to 2021, then scored on 2022–2023, years it never saw.

3 · Output

What you get

Owners see break-even sales and area comparisons. Lenders see a closure-risk band for each store.

Model weights and data-cleaning rules are not published.

One engine, two screens

For owners and for lenders.

For owners

Where to open, and what you keep

Compare areas and see the sales you need to break even.

For lenders

When, how much, and on what terms

Risk bands checked on years the model never saw.

Evidence

Tested on years the model never saw.

62.6%ranking accuracy, Vancouver
64.4%ranking accuracy, Toronto
1.7×closure rate of the riskiest 10%

A storefront closure is not a loan default. These figures support decisions next to a credit file; they do not replace it.

Demo

The demo is available by invitation.

For lenders, partners and program reviewers.

Or email [email protected]

Request a demo