Will our data be valuable to customers?

How precise are ticket lines from bars and restaurants? What we learned in our first year.

When we started Tea Time Data, one question mattered more than any other: is the data from bar and restaurant tickets precise enough to be valuable to drink brands?

Is a ticket line precise enough?

When we started this business about a year ago, we had a series of business checks to perform. The main one was simple: is the data we receive from ticket lines precise enough to be valuable to customers?

If most of the anonymised bars and restaurants recorded their sales as “soft drink”, “beer” or “alcohol”, the data would tell a brand very little. You can't measure a brand's market share, prices or distribution if you don't know which brand was sold.

Every bar and restaurant decides for itself how to record products in its point-of-sale system.

That is the real challenge. One venue types “Coca-Cola 33cl”, another types “coke”, a third just types “soft”. If the recording isn't specific, the data isn't good enough. So we spent months analysing ticket lines, category by category.

Three levels of accuracy

The share of ticket lines we can link to a specific brand depends strongly on the product category.

01
High accuracy 92%
The data is good enough to link a line to a specific product, a specific brand and, in most cases, a specific quantity.
Soft drinks, beer and alcohol (excluding wine).
02
Mid-level accuracy ~50%
We always know the type of product, but in about half of the cases there isn't enough detail to link it to a brand.
Example: water. We know whether it is still or sparkling, but can't always tell which brand.
03
Low accuracy <20%
We know what was ordered, but fewer than one line in five tells us the brand behind it.
Example: coffee. We know the type (espresso, cappuccino…), but not the coffee brand.

Share of ticket lines that can be linked to a brand.

Valuable where it matters most

Brand data is not available at the same level in every category. For coffee, for example, ticket data alone can't tell you which brand was served.

Luckily for us, our main customers are soft drink, beer and alcohol producers. For them, the level of accuracy is very high. That answered our first and most important business check: yes, our data is valuable to customers.

Key takeaways

High for soft drinks, beer and alcohol (excluding wine): 92% of lines can be linked to a specific product and brand.
Mid for products like water: we know the type, but can link only about half to a brand.
Low for products like coffee: we know the type, but can link fewer than 20% to a brand.
Our main customers are soft drink, beer and alcohol producers, and that is exactly where accuracy is highest.

Want to know what ticket data says about your brand?

Let's talk data →