Determining propensities to drive website target user activity

1. A method, comprising:

receiving, by one or more processors, activity data indicative of use of a set of user paths by users, wherein individual user paths, of the set of user paths, comprise at least one of:

one or more digital channels,

one or more digital systems, or

one or more non-digital systems;

identifying, by the one or more processors and based on the activity data, instances of a target user activity occurring in association with the individual user paths;

determining, by the one or more processors and based on the activity data, respective propensities of the individual user paths to drive the target user activity;

determining, by the one or more processors and based on the respective propensities, that a first propensity of a first user path to drive the target user activity is lower than a second propensity of a second user path to drive the target user activity, the first user path having a first attribute, and the second user path having a second attribute different from the first attribute; and

based on determining that the first propensity is lower than the second propensity, revising, by the one or more processors, the first user path based on the second attribute of the second user path.

2. The method of claim 1 , wherein the one or more digital systems comprise at least one of a website or a mobile application.

3. The method of claim 1 , wherein the one or more non-digital systems comprise at least one of phone calls or in-person meetings.

4. The method of claim 1 , wherein the set of user paths is associated with an insurance company, and the target user activity comprises an initiation of a new insurance quote via the one or more digital systems or the one or more non-digital systems.

5. The method of claim 1 , wherein determining the respective propensities comprises determining, by the one or more processors using a logistic regression model, relative contributions of the individual user paths to drive the target user activity.

6. The method of claim 5 , wherein determining the respective propensities further comprises determining, by the one or more processors using the logistic regression model, the relative contributions of instances of the one or more digital channels, the one or more digital systems, or the one or more non-digital systems, associated with the individual user paths, to drive the target user activity.

7. The method of claim 1 , wherein the one or more digital channels comprise at least one of a direct load channel type, a display advertisement channel type, an email advertisement channel type, a search engine optimization channel type, a search engine marketing channel type, a social media channel type, a text advertisement channel type, or a print media channel type.

8. The method of claim 1 , wherein revising the first user path comprises adjusting the one or more digital channels associated with the first user path to direct future users from the one or more digital channels associated with the first user path to the one or more digital systems or the one or more non-digital systems associated with the second user path.

9. The method of claim 1 , wherein:

the activity data indicates user classifications of the users,

the one or more processors identify the instances of the target user activity occurring in association with the individual user paths and individual user classifications, of the user classifications, and

the one or more processors determine the respective propensities of the individual user paths to drive the target user activity in association with the individual user classifications.

10. One or more computing devices, comprising:

one or more processors;

memory storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:

receiving activity data indicative of use of a set of user paths by users, wherein individual user paths, of the set of user paths, comprise at least one of:

one or more digital channels,

one or more digital systems, or

one or more non-digital systems;

identifying, based on the activity data, instances of a target user activity occurring in association with the individual user paths;

determining, based on the activity data, respective propensities of the individual user paths to drive the target user activity;

determining, based on the respective propensities, that a first propensity of a first user path to drive the target user activity is lower than a second propensity of a second user path to drive the target user activity, the first user path having a first attribute, and the second user path having a second attribute different from the first attribute; and

based on determining that the first propensity is lower than the second propensity, revising the first user path based on the second attribute of the second user path.

11. The one or more computing devices of claim 10 , wherein the set of user paths is associated with an insurance company, and the target user activity comprises an initiation of a new insurance quote via the one or more digital systems or the one or more non-digital systems.

12. The one or more computing devices of claim 10 , wherein the first propensity and the second propensity are determined by identifying, using a logistic regression model, relative contributions of the first user path and the second user path to the instances of the target user activity.

13. The one or more computing devices of claim 10 , wherein revising the first user path comprises adjusting the one or more digital channels associated with the first user path to direct future users from the one or more digital channels associated with the first user path to one or more digital systems or one or more non-digital systems associated with the second user path.

14. The one or more computing devices of claim 10 , wherein:

the activity data indicates user classifications of the users,

the instances of the target user activity occurring are identified in association with the individual user paths and individual user classifications, of the user classifications, and

the respective propensities of the individual user paths to drive the target user activity are determined in association with the individual user classifications.

15. One or more non-transitory computer-readable media storing computer-executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations, comprising:

receiving activity data indicative of use of a set of user paths by users, wherein the activity data indicates user classifications of the users, and individual user paths, of the set of user paths, comprise at least one of:

one or more digital channels,

one or more digital systems, or

one or more non-digital systems;

identifying, based on the activity data, instances of a target user activity occurring in association with the individual user paths and individual user classifications of the user classifications;

determining, based on the activity data, propensities of the individual user paths to drive the target user activity, in association with the individual user classifications;

determining a particular user classification associated with a particular user;

selecting a particular user path, of the set of user paths, associated with a highest propensity, of the propensities, to drive the target user activity in association with the particular user classification; and

recommending usage of the particular user path in association with the particular user.

16. The one or more non-transitory computer-readable media of claim 15 , wherein determining the propensities comprises determining, using a logistic regression model, contributions of the individual user paths to lead to the target user activity.

17. The one or more non-transitory computer-readable media of claim 16 , wherein determining the propensities further comprises determining, using the logistic regression model, the relative contributions of instances of the one or more digital channels, the one or more digital systems, or the one or more non-digital systems, associated with the individual user paths, to drive the target user activity.

18. The one or more non-transitory computer-readable media of claim 15 , wherein the user classifications are based on ages of the users.

19. The one or more non-transitory computer-readable media of claim 15 , wherein:

the one or more digital systems comprise a website hosted by a server, and

recommending usage of the particular user path in association with the particular user causes the server to serve, to a user device associated with the particular user, a webpage that presents content that directs the particular user to a digital system or to a non-digital system associated with the particular user path.

20. The one or more non-transitory computer-readable media of claim 15 , wherein recommending usage of the particular user path in association with the particular user causes one or more of the digital channels to direct the particular user to a digital system or to a non-digital system associated with the particular user path.