Skip to content

Data flow

The data flow in Tracardi goes through the following stages:

Stage Description
Source validation Tracardi must have event source defined and enabled in the system.
Event data validation Tracardi can validate the event daa schema.
Identity resolution If an event is an idification point then this step is performed.
Event reshaping Tracardi can change the event schema if needed.
Event collection Tracardi saves the event.
Event routing Tracardi reads a rule that defines which workflow must precess event.
Processing Tracardi runs a workflow that processed event data, enhances the data,routes
it to external systems, etc.
Profile merging Tracardi checks if after processing the profile can be merged with other
profile that seems to be the same.
Profile segmentation Tracardi segments profile based on the rules provided by the system administator.

Source validation

Tracardi can collect data from many places, but all of them enter the system via two API endpoints (/track and /collect).

However, in order to transfer data to the system, the system administrator must enable it. By creating an event source, creates a token that, when attached to the event data, will allow it to be sent through the above-mentioned endpoints

Tips

Event sources can be found in Tracardi GUI under Traffic. We have inbound traffic that comes to Tracardi - we call it event sources. And there is outboud traffic sometimes called the destination. This is the external system that the profile data can be sent when profile changes.

Event data validation

Event data validation is the process of checking the accuracy, completeness, and consistency of data associated with an event. This may involve verifying that the data meets certain standards or requirements, such as being in the correct format, or containing required fields.

Event data validation is an important part of ensuring the quality and reliability of data used for analysis and decision-making. By ensuring that the data is accurate and complete, businesses can trust that the insights they derive from the data are reliable and actionable.

Event collection

Event collection refers to the process of gathering data about events within Tracardi.

Example of an event

Event ia data that is described by a type, e.g. a purchase-order event may contain data such as product name, quantity and price.

All events go to a central database in elasticsearch and can be viewed via the GUI in the Data section.

Events can be saved in the database or only processed. For each new event without an assigned user, Tracardi will create a new profile and assign upcoming events to it.

Browsing events

Event routing

Tracardi saves the collected event and then looks for the rules that define the workflow to use to process the events. A rule is a simple filtering process that checks that an event comes from a defined source and has a defined event type. For example, a simple rule might require a website to be the source of an event and a page-view for the event. Then, if an event occurs meets these criteria, the rule will run the defined flow and send the event data.

The rules connect events of a given type with the indicated workflow.

Warning

Events that at the time of their registration did not have rules created, i.e. they were not connected to the process that processes them, are not processed. They remain in the received state. Tracardi does not currently offer reprocessing of old data. More on events can be found in Event's core definitions

Tip

You can find more about the technical aspects of collecting events and processing them in Event tracking

Processing

Tracardi uses workflows to process data.

Workflow is a graph of actions that will run when an event is matched with workflow. Actions may run one after another or in parallel. Workflow is represented as a graph of nodes and connections between them. Actions are assigned to nodes in the workflow graph. Actions may perform different tasks such as copying data from the event to profile, save profile, query for additional data, send parts or whole data to another system or emit another event.

Browsing events

When the workflow ends Tracardi checks if there is a need for profile update, segmentation, and merging.

Segmentation

Segmentation is a term that refers to aggregating prospective customers into groups or segments with common needs. Segmentation enables companies to target different categories of consumers who perceive the full value of certain products and services differently from one another.

Segmentation in Tracardi is based on a segment logic that is defined in the system, e.g. apply segment frequent-customer to a profile that visited the page more than 10 times.

Segmentation runs after the workflow ends.

Common characteristics of a market segment

Common characteristics of a market segment include interests, lifestyle, age, gender, frequency, etc. Common examples of market segmentation include geographic, demographic, psychographic, and behavioral.

Merging

Profile merging is the process of finding customer profiles that belong to one person and have been saved as separate records for various reasons. Tracardi runs profile merging after the workflow ends.

How the profile can be merged?

In order to combine data into one record, it is necessary to indicate the field containing the value by which the customer data could be combined. This process will start automatically if the workflow has a merge profile action.

Summary

Questions that this documentation answers:

  • What stages does data flow through in Tracardi?
  • How does Tracardi route events to the appropriate workflow?
  • What is the role of segmentation in Tracardi?
  • How does Tracardi handle profile merging?
  • What is event reshaping in Tracardi?
  • How does Tracardi determine which workflow to use when processing an event?
  • How does segmentation in Tracardi differ from traditional market segmentation techniques?
  • Can Tracardi be used to send data to multiple external systems simultaneously? If so, how?
  • How does Tracardi validate event data?
  • What are some examples of tasks that can be performed by actions in Tracardi workflows?
  • How does Tracardi segment profiles?
  • How to merge profiles from different channels?