Transactional data

Transactional data is data recorded from a transaction: a sale, a booking, or a customer using a loyalty card. It is not simply any data an organisation happens to hold — it is the record of a transaction taking place. It can be collected by the business the customer bought from, or by a card company that handles the payment.

Each record can hold what was bought, when and where it was bought, how often that customer buys, how much was spent, and customer details. Loyalty card data is transactional data too, and because the card is linked to a named customer it can also reveal customer demographics and shopping habits.

Finding trends, patterns and demographics

Transactional data is analysed to find trends and patterns — general changes or repeated behaviour found by looking at data over time — and customer demographics, the characteristics of customers that help an organisation understand who they are.

For example, a garden centre records every till transaction. Looking back over one spring, the manager finds that Saturday mornings are busiest and need extra staff, that the lavender range sells least, that takings are highest in May, that the town-centre shop sells far more bedding plants than the out-of-town one, and that most customers come from two nearby villages. Those are patterns and demographics in action. The centre can reorder stock, cut back on the poor seller, discount quiet weekday afternoons, advertise in a village it is not yet reaching, and run a planting workshop on a quiet weekday. Displaying the figures in graphs makes these patterns easier to spot than reading a long list of records.

# Plant sales recorded at a garden centre across one Sunday
sales = ["shrub", "lavender", "shrub", "bird feeder", "shrub", "bird feeder", "shrub"]

print("Transactions recorded:", len(sales))

for item in ["shrub", "bird feeder", "lavender"]:
    print(item, "sold", sales.count(item), "times")

Output:

Transactions recorded: 7
shrub sold 4 times
bird feeder sold 2 times
lavender sold 1 times

Drawbacks to weigh up

Analysis is not free: staff may need to be employed to extract and analyse the data, which adds cost. The data must also be accurate — errors in transactional data lead to bad decisions, such as ordering 500 extra umbrellas because a mistyped quantity made them look popular.

A discount or special offer helps the customer rather than the business, so it is not a benefit of transactional data to the organisation.