ZipF / Blog / POS data from the bill
Data
POS data from the bill
Most alternative-data panels infer what happened at a merchant. ZipF records it. POS data in this stack is the checkout itself: SKU, quantity, price, discount, basket, and sentiment, written as the merchant operates.
Card transaction panels already sell nowcasts to high-frequency, systematic, quantamental, and fundamental investors. They are valuable. They see a total at a merchant, often with lag. ZipF's panel is merchant-native. It sees the line items that summed to that total, on infrastructure the company owns.
That difference is the product. A card event is a total at a restaurant. A ZipF event is two specialty coffees, one croissant, a 12% discount, 8:14am in Hayes Valley, a repeat customer, sentiment positive - the same schema as a ₹780 bill in Indiranagar. Mix, premiumization, brand switching, and discounting-with-volume live in the second record. The buyer class already pays in dollars: Toast $13.3K ARR per US location, Square $250B GPV, global alt-data a multi-billion market with North America the largest buying region.
Who this is for
Merchant operators get analytics and an action button. Brands get share inside venues. Advertisers get closed-loop attribution to the purchase. Desks get a Commerce Signal API with latency SKUs and exclusivity. The same events feed all four. The contract is the packaging.
Coverage thesis
The US from day one and India from day one. Integrations first (Square, Toast, Clover, Lightspeed, Petpooja, Razorpay, Restroworks), then the ZipF terminal as the complete stream. The US restaurant economy is $1.55T in 2026. Worldwide installed POS is about 292 million devices on a $130.6B market. The India plan is a density beachhead, not a ceiling.
Normalized alternative data sold as institutional intelligence already supports large contract books. ZipF's book is the physical-economy version of that motion, with SKU fidelity the card panel does not have.
POS data that only records the sale is a card-panel-shaped product. ZipF's POS data includes the leak before the sale and the outcome after it. That is what makes the dataset useful to merchants (where did we lose the guest?), to brands (latent demand for a SKU), and to agents (is this merchant actually operable right now). Extra line items on the bill are not the moat. The full funnel is.
Questions
What is ZipF POS data?
Merchant-native, item-level checkout data: the bill ZipF generates on its own terminal or via POS integrations, not a reconstructed card total.
How is ZipF POS data different from card panels?
Card panels infer spend from cards. ZipF records SKU, basket, price, and sentiment from the merchant operating system.
Can trading desks buy the feed?
Yes. The Commerce Signal Network sells API credits, latency SKUs, and exclusive rights to systematic, quantamental, macro, and market-making desks.