For the complete documentation index, see llms.txt. This page is also available as Markdown.

Explore a user profile

Look one user up and read what the platform actually holds about them.

Your goal

Open a single user and read their raw data - to check that collection works as designed, to investigate a support question, or to answer an access request.

Expected outcome

You can see a user's identifiers, profiles, activities, segments, and choices, and compare them against what you expected to collect.

Why this is the fastest debugging tool you have

Aggregate reports tell you that something is wrong; one user tells you what. A property landing at the wrong path, an identifier that never arrives, an event rule doing something unexpected, an activity analyzer stripping a field - all of these are visible in seconds on a real UserPoint, and invisible in a count.

Open the User lookup screen

In Navigator, open User lookup, under the Audience section. It gives you a 360° view of one UserPoint, gathering everything the platform holds about that user.

By default, the screen shows the UserPoint associated with the user agent ID - the vector ID - from your own browser. In other words, it shows you.

Look up someone else

Use the User lookup button at the top right of the view. You can find a UserPoint by any of these:

Search on
Notes

User Account ID

The identifier of a registered user, in one of your compartments.

User Agent ID

The vector ID.

Email hash

The hashed form of the user's email address.

UserPoint ID

The identifier the platform created for that user, shown at the top of the screen, for example ab8e2bba-6b6a-46ab-bd6a-a212063b8e10.

What each block shows

Block
What it holds

Profile

Everything on the UserPoint's profiles: data captured from a CRM such as name or birth date, and computed data such as scores. Which fields exist is decided by your datamart schema, and compartments keep profiles from different sources apart.

Segments

Every segment the UserPoint belongs to.

Timeline

The user's activities and events, newest first.

User Account Id

The UserAccount information. A user can hold several account IDs when your datamart has several compartments.

User Device

Every UserDevice known for the user, with the UserDeviceTechnicalIds attached to each: mediarithmics cookie ID, mobile ID, network ID, and so on.

User Email

The UserEmail: the hash, and the raw email where it is available.

User Choices

What the user has accepted or refused, per processing activity.

Read the timeline

The timeline represents all of the user's activities and events, from the most recent to the oldest.

Each block on it is an activity, gathering the events of one user session. Browsing an e-commerce site, for instance, produces events such as product pages, basket views and purchase actions, all grouped into the activity for that visit.

Click Details on an event and expand it to see everything attached to it - item, category, price, and any other property. Which properties exist is decided by your tag integration, so this is the fastest way to check that the tracking plan you agreed on is the one actually running.

You can also view the complete JSON for each activity. That raw JSON is the point: it shows the activity as stored, after event rules and activity analyzers have run - which is what your queries and your schema see.

Switch to the technical timeline

A switcher turns the activity timeline into the technical timeline, which shows the identifier lifecycle of the UserPoint: when an identifier was added, when UserPoints merged, and when an identifier was removed. These event types exist:

Event
Triggered when
Fields

Add identifier

An identifier is added to the UserPoint.

Subject UserPoint - normally the one you are looking at; the identifier type; the identifier itself; optionally the cause type and the activity that triggered it.

Merge UserPoint

The UserPoint is merged with one or more others.

Target UserPoint - the one kept after the merge; the list of merged UserPoints; optionally the cause type and the activity that triggered it.

Remove identifier

An identifier is removed from the UserPoint.

Subject UserPoint; the identifier type; the identifier itself; optionally the cause type and the activity that triggered it.

No merge UserPoint

The UserPoint should have merged with one or more others, and did not, because of a specific cause or a technical rule.

Target UserPoint; the UserPoints that were not merged; optionally the cause type and the activity that triggered it.

When to reach for it. The technical timeline is the answer to "why are these two users still separate?" and "where did this cookie ID come from?". A No merge UserPoint event is the explicit trace of a merge the platform decided against - see Resolve identities.

Look a user up by API

Any identifier can select the UserPoint, and any document type can be read from it.

Selector
Form

UserPoint ID

<USER_POINT_ID> or user_point_id=<USER_POINT_ID>

User agent ID

user_agent_id=<USER_AGENT_ID>

Account

compartment_id=<COMPARTMENT_ID>,user_account_id=<USER_ACCOUNT_ID>

Email hash

email_hash=<EMAIL_HASH>

<DOCUMENT_TYPE> designates what you read:

  • user_identifiers

  • user_profiles

  • user_activities

  • user_choices

  • user_scenarios

  • user_segments

<DOCUMENT_SELECTOR> is optional - use it to get, update, or delete one particular document. Which selector applies depends on the document type:

The UserPoint API allows you to retrieve, create, update, or delete all data related to a single UserPoint - so the same endpoints serve investigation and correction.

For a single request returning several document types at once, use GraphQL instead - bearing in mind its 100-activity limit.

What to check, and what it tells you

Look at
It tells you

The identifiers on the UserPoint

Whether your tag and imports send what you think. A missing account ID explains an unconsolidated user.

The number of UserPoints for one real person

Whether identity resolution is working. See Resolve identities.

The raw activity JSON

Whether properties are where the schema expects them, and whether event rules and analyzers did what you intended.

Event names

Whether predefined names are used, and whether $page_view events are being relied on - they are dropped at session close.

The profiles, per compartment

Whether imports landed in the right compartment, and whether $user_account_id was filled inside the profile object.

The user choices

Whether consent exists for the processing activities you are about to enforce.

The segments

Whether membership matches what the segment's query should produce.

Use it in these situations

1

After implementing tracking

Take one known test user and read their activities. Faster and more conclusive than any aggregate.

2

When a query returns nothing

Check whether the data is absent, or present but not where the query looks. This distinguishes a collection problem from a schema problem.

3

Before writing a computed field

The activities a plugin receives are raw data - you cannot rely on your schema's structure to know their shape. Read a real activity first. See Build a computed field.

4

When answering an access request

Retrieve before any erasure, and look the person up by every identifier they gave you. See Exercise rights of the user.

Last updated

Was this helpful?