Implicit Personalization scoring explained
In Umbraco Engage you can personalize the website experience of any visitor based on implicit scoring.
Implicit personalization is based on gaining confidence that a visitor shows behavior that can be mapped to a persona or a customer journey step. To gain this confidence it is possible to assign points to specific actions within your website. If a certain threshold of points is reached Umbraco Engage assumes the visitor is this persona or in a specific customer journey step. As soon as that point is reached, you can use that information to personalize the website experience of your visitor.
There are four ways to score the behavior of your visitors:
Score the content that a visitor is viewing. This can be done per node.
Score the campaigns that a visitor is part of.
Implement your own scoring. In this way, the sky is the limit, because you can hook into any external data source you have or behavior that you want to score.
Collecting Points
The points of all these different sources are added and this reaches a certain amount of points per persona. Once a persona or journey step reaches the set threshold, the algorithm assigns you to that persona or step.
In the example, the visitor collected 40 points for the Data & Privacy officer, 30 points for the Marketer, and 0 points for the developer persona:
The threshold in this specific case was set to 25 points. As soon as the Data & Privacy officer reached 25 points Umbraco Engage assumed that this visitor was a Data & Privacy officer.
In this example the Think customer journey step is assumed based on the collected amount of points:
Tweaking the Scoring
The threshold value and the expected difference between two personas or journey steps can be set in the customer journey group and persona group.
Setting up a deviation of at least 35 points between two personas the cockpit will show a different visualization in the previous example:
You can see that the "Data and privacy officer" still has 40 points and the marketer 30 points. Both have also reached the threshold of 25 points, but there is not a minimal deviation of 35 points. The Umbraco Engage algorithm waits for the deviation to reach the set threshold before assuming a persona. For example: the Data & privacy officer reaches 65 points (30 points of the marketer + a minimal deviation of 35 points).
Absolute versus Percentage deviation
Minimal deviation can be measured in two ways: Absolute or Percentage. Learn more about each approach below.
Absolute
When set to Absolute, the deviation is measured in points. For example, if the highest score is 40 and the next highest score is 35, the deviation is 5 points.
Absolute deviation table:
The table below uses the same Data & Privacy officer / Marketer example as above:
40
35
5
10
Not activated. Gap too small
45
30
15
10
Activated. Gap exceeds 10 points
Percentage
When set to Percentage, the deviation is measured as a percentage of the second-highest score. This can be useful when you want the required gap between profiles to scale with the overall score.
Percentage deviation table:
The table below uses the same Data & Privacy officer / Marketer example as above:
34
30
4
25%
7.5 points (25% of 30)
Not activated — gap is only ~13%
40
30
10
25%
7.5 points (25% of 30)
Activated — gap is ~33%
68
60
8
25%
15 points (25% of 60)
Not activated — same 8-point gap that would pass at a lower score fails here
Last updated
Was this helpful?