Klever UI: People and Devices

How to Read the People and Devices Chart in Segment Summary Screen:

 

Key

  • Blue bar: count of unique devices that meet the segment criteria
  • Orange bar: count of uniques  that are mapped to at least 1 device that meets the segment criteria
      • Dark orange: deterministic count of people that is gathered through onboarding
      • Light orange: predictive count of people through Salesforce DMP's proprietary identity graph

Tips

  • Hovering over the people and devices icons will show a descriptive label that clarifies the definition of the iconography
  • Hovering over the light orange people bar will display the predictive tooltip:

In the above example, a total of 127.62K people that meet the segment criteria map to at least 1 device. ~57% (or 73.17K) of total people were mapped using Salesforce DMP's proprietary identity graph. Of that ~57% (or 73.17K), ~50K map to only a single device, whereas ~23K map to more than one device.

  • Hovering over the dark orange people bar will display the deterministic tooltip:

The total number of devices for this segment (~270K) were mapped to ~127K people. If you divide the number of devices by the number of people who meet this segment criteria, you get 2.1 devices per person. Notice the total devices mapped is less than the segment's device population of ~473K. This difference is due to the fact that not all devices appear on Salesforce DMP's Identity Graph. A device must have a sufficient amount of information tied to it to inform a prediction. If there is insufficient information, that device will not appear on the device graph. For this reason, the user should not divide total devices with total people to estimate devices per person, but use the mapped devices instead.

This particular example shows that the ~73K people that have devices that meet the segment criteria, are related to an additional ~150K devices that did not directly meet the segment criteria. In other words, the number of people (~73K) determined by predictive CDIM that meet the segment criteria have a total of ~270K related devices, or 3.7 devices per person. Only 1.6 of their devices directly met the segment criteria.

  • The remainder of the People & Devices chart describes how devices and people breakout to corresponding device categories:

 

  • Device categories include:

          ~Web = Counts IDs associated with a browser (desktop or mobile)

          ~App = Counts IDs associated with a MAID (mobile or tablet)

          ~Safari = Counts IDs associated with Safari and/or DNT. These IDs do not appear on the identity                   graph, so the predictive people stack will never appear for this device category.

 

  • People populations for device categories are not additive. The same person may appear in more than one category if they deterministically or predictively map to more than one device category. For example, the same person may map to a browser and a MAID.
  • Device populations are additive.

 

People and Devices Chart in Segment Summary Screen FAQs:

How do I see the number of people and devices under Insights in the segment details?

If you hover over the bars in the chart, a pop up will show the actual number of people and devices.

 

When does data appear in these charts?

Data will appear in this chart as soon as the segment is computed.

 

What confidence level is the predictive population within people and devices?

Within People & Devices you will see the predictive population of people and devices. This number represents our medium confidence equivalent calculation.

 

Will device numbers reconcile between the Konsole and Klever UIs?

Device populations will match, but people populations will not match. Unlike the  Klever UI, the Konsole UI counts all devices not mapped to a person deterministically as an individual person. The Klever UI leverages Salesforce DMP’s proprietary Identity Graph to predict which devices relate to a common person. Devices that are not present on the Identity Graph are not included in the people count.

 

How do I calculate devices per person?

Devices per person should not be calculated by dividing total devices by total people. See above for the methodology that should be used.

 

Do all devices map to at least one person on the device graph?

No, the device graph only contains devices that have enough information to provide a reasonable prediction about their association with other devices.

 

Are the breakouts of populations by device category additive?

The device populations are additive, but the people populations are not. The same person may map to both a browser and a MAID. See above for more details.

 

Is mobile web counted in the web or apps device category?

It is counted in the web category. App IDs correspond to MAIDs only. All browser IDs, web or mobile, are counted as web IDs.

 

 

Why does Safari not show people data?

Safari does not allow 3rd party cookies, therefore deterministic or probabilistic matching is not supported in Safari.

 

 

How to Read the Device Population Trend Chart in Segment Summary Screen:

  • Key
    • Blue line: count of devices that meet the segment criteria. This data corresponds to the left axis.
    • Green line: count of devices added to the segment. This data corresponds to the right axis.
    • Gray line: count of devices removed from the segment. This data corresponds to the right axis.
  • The chart looks back up to 90 days starting from the Klever UI Release. A new data point will appear every time the data is processed. If the segment is set to process on-demand, the chart may be more sparse than a segment that processes daily.
  • Devices added/removed will only appear after the segment processes twice.
  • Sudden changes to the segment population may be caused by a variety of reasons.
    • Examples include:
      • The segment's rules were changed
      • There was an issue with a corresponding first-party data import
      • An interaction defined in the segment criteria has stopped or started. For example, if a segment rule is tied to an action/attribute that is no longer tracked or a section that no longer exists

 

Device Population Trend Chart in Segment Summary Screen FAQs:

 

When does data appear in these charts?

The blue line representing total devices will take shape as soon as the segment is processed the first time. The green and grey lines representing day-to-day change in device counts will render after the segment has processed a second time.

 

Why are no devices being removed?

Devices may not be removed if none of the rules within the segment use recency criteria. Devices may not be removed if the segment's age is less than the organization's licensed lookback window.

 

Why is there a sudden change to the segment population?

A segment’s population can experience volatility for a variety of reasons. A few examples are as follows:

  • The segment's rules were changed
  • There was an issue with a corresponding first-party data import
  • An interaction defined in the segment criteria has stopped or started. For example, if a segment rule is tied to an action/attribute that is no longer tracked.

 

 

Have more questions? Submit a request

0 Comments

Please sign in to leave a comment.