‘Transactions’ in Blueshift allow you to model journeys and update the state of the journey via multiple events. An example of a transaction would be orders for an e-commerce company. Various events in an order’s life cycle (e.g. shipped, delivered, canceled, returned etc.) can be used to update the order’s state. At any given instant of time, you will be able to fetch the entire history of a user’s order and use the information to communicate with the user. Refer to this document to learn more about transactions.
The ‘transaction replay’ feature allows you to build segments of transactions (instead of users). In other words, it allows you to query the transactions table. Using the e-commerce example above, you could use this feature to find all orders for a certain product A in the last 7 days and then reach out to the users who placed these orders.
With the introduction of the ‘triggering transaction’ filter option in the campaign journey, you can send follow up messages conditional to attributes of the transaction that triggered the journey. For example you could send a different follow up message to users who purchased product A and subsequently canceled it than those who didn’t cancel their order.