It's like Supermarket Scan Data for Hospitality.

Trickle takes raw point-of-sale lines from pubs, bars and restaurants across Aotearoa and resolves them into clean, anonymised, product-level sales data — by brand, volume and vessel, every trading day.

Talk to us about data access

Received from venues

Resolved to one product

HEINEKEN 330

heiny btl

Heineken Bottle

HEINEKIN 330ML

HEIN-BTL-330

Heineken

Distributor: DB Breweries
Category: Lager

Steinlager Classic

Distributor: Lion
Category: Lager

Villa Maria Sauvignon Blanc

Distributor: Indevin
Category: White wine

Raw at one end… Reportable at the other.

Received

We read each sale from the venue's point-of-sale exactly as it was rung up — item, quantity, value and timestamps.

Reconciled

Sales are matched to our library of products using AI and every serve is normalised. Anything that doesn't resolve cleanly is flagged for a human to intervene rather than discarded.

Reported

The reconciled serves are aggregated into a reporting layer that carries no venue identifiers, and is ready to be queried or exported.

Venues share data because it can't come back to them.

Anonymity isn't a policy we apply at export. It's built into where the data is allowed to sit, which is what makes the pool licensable at all.

Venues are handles, not names

Every venue carries a stable pseudonymous handle like v-7f3a91c2. It's consistent everywhere, so a venue can be followed over time, and it identifies nobody.

The reporting layer holds no identity

No venue names, no addresses, no point-of-sale identifiers and no raw item text — only the product, the serve, the value and the trading date.

Venue details aren't decryptable at rest

Contact details sit behind a key that exists only in memory, and only while a person is actively using it. With no password entered, there is no key to steal.

Data fresh from the Sauce Source.

Each point-of-sale system is integrated on its own terms, so venues change nothing about how they work.

Kounta

Lightspeed

posBoss

SwiftPOS

Tevalis

OneTap

Revel

Bepoz

Menumate

Running something else? Tell us what it is — most systems can be integrated.

The things buyers ask first.

What's actually in the data?

One row per serve sold: the canonical product with its brand, category and volume, the vessel it was served in, the value including GST, the trading date and a basket reference that ties together everything bought at the same time. Every quantity is a serve, so a row means the same thing regardless of which venue or till produced it.

How current is it?

Sales are collected per venue per trading day, and a trading day runs 4am to 4am so that a Friday night belongs to Friday rather than splitting across midnight.

Can a venue be identified in the data?

No. The reporting layer carries a pseudonymous venue reference and nothing else that relates to the venue — no name, no address, no point-of-sale identifiers, not even the raw item text the till produced.

Can we get a custom cut?

Yes — by category, brand, period, vessel or venue type, as a one-off extract or on a recurring schedule. Tell us the question you're trying to answer and we'll tell you whether the data can answer it.

How do venues get involved?

Venues connect their point-of-sale once and get market insights back for free, with nothing to install and no change to how they take orders. If you run a venue and want in, email us and we'll organise access.

How do we get access?

Email us with what you're trying to measure. We'll walk you through the coverage, what's in scope, and what an extract or licence looks like for your use case.

Put real on-premise numbers behind the decision.

Tell us what you're trying to measure and we'll tell you, straight, whether this data can answer it.