Data Stream: What Is?

Data Stream provides access to raw behavioral data based on thousands of data points for billions of unique consumer profiles.

Data can enhance existing data assets to provide deeper insights into buyer profiles.

Data Stream enables data science teams to meet data needs, build own data tools, including personalized experience and messaging, recommendation systems and machine learning algorithms.


Your Current Data

Data Stream

Full User Profile

Data Stream: Key Benefits

Deep insights into visitors and customers
Tracking customer behavior outside of your reach
Implementing own machine learning methods and rules
Personalized experience and messaging for customers

Data Stream: How It Works?


Part 1:
ID Synchronization

Anonymous and secure Synchronization of IDs through code implementation. DMP installation if necessary.

Part 2:
Data Stream Implementation

Developing secure daily Data Stream to specified location and technological process verification.

Part 3:
Data Processing

Enhancing Client’s data assets to improve customer insights. Building advanced tools and models based on data.

Data Stream: Formats


Based on segments

  • user ID
  • list of segment IDs

Based on data points

  • cookie birthdate (timestamp)
  • data point IDs
    • number of occurrences for each ID
    • timestamp of last occurrence for each ID

Data Stream: Segments

Segment is a logical statement, built on specific Data Points.
For identification and reference purposes, each segment has a unique ID.
User should have required level of access to DMP to modify their own segments.
If not necessary, it is possible to provide only those segments that are available in our taxonomy.

The segments for users we share, must be cookiematched with the partner, which means to have already built a mapping table for identifying users.

9223304722221876165 and 9223255841725915589 are User IDs; the rest are the IDs of the shared segments.
The format can be extended to
User ID, date of cookie creation, last user’s activity, country, Seg ID…

Data Stream: Data Points

Data Point is our tag, which is divided by Events and Attributes with unique ID used to collect data.

‘Event’ as a datapoint is used to collect data about occurrences of the specific event (e.g. visiting site, displaying ad, clicks, etc.)
‘Attribute’ datapoint used to collect numerical or alphanumeric values that can be later used in DMP to create appropriate audiences.

As well as in case of Segments, cookie matching process should already have a place here.

Particularly said, the Data Point can be a type of the page, visited by the user.

1 – 9221671221284926901
2 – User ID (Partner ID) 160003 is Data Point ID
3 – 1 – number of occurrences
4 – ‘2018-8-19 1:35:30’ – date of last timestamp

We share data about Data Points on client’s demand.
Additionally we can also provide DMOZ and IAB2 dictionaries which will help in partial identifications of Data Points.

This data type is especially useful for a partner, who doesn’t have a domain or own system.