Determining an optimal communication channel
Examples include receiving, at a system, user data related to a user of a plurality of users, the user data including user characteristics associated with the user. The system assigns the user to a group of users based at least in part on the user sharing at least one user characteristic with users in the group of users. The system uses a machine learning model for determining, at least in part, an optimal communication channel through which to send a loan offer to the user, the machine learning model having been trained to determine optimal communication channels for providing loan offers based at least on user data associated with respective users of the group of users. The system sends the loan offer to the user via at least one communication channel selected based at least on the optimal communication channel determined at least in part using the machine learning model.
1 . A system comprising:
one or more processors; and
one or more computer-readable media storing instructions executable by the one or more processors to configure the one or more processors to:
receive user data related to a user of a plurality of users, wherein the user data comprises one or more user characteristics associated with the user;
assign the user to a first group of users based at least in part on the user sharing at least one user characteristic with users in the first group of users;
track statistics associated with conversion of a plurality of loan offers previously offered to the first group of users and communication channels used for sending the plurality of loans, the statistics including a number of clicks, a number of website visits, and a number of interactions with prior push notifications;
determine, based at least in part on a machine learning model that is trained based at least on user data associated with users of the first group of users and the statistics to determine optimal communication channels for providing loan offers, an optimal communication channel through which to send a loan offer to the user, the optimal communication channel is one of an electronic mail communication channel, an electronic text message communication channel, and a push notification on a user device communication channel; and
send the loan offer to the user via the optimal communication channel determined based at least in part on the machine learning model.
2 . The system as recited in claim 1 , wherein execution of the instructions further configure the one or more processors to determine the optimal communication channel by determining a predicted acceptance rate for respective communication channels of a plurality of candidate communication channels for users in the first group of users.
3 . The system as recited in claim 2 , wherein the machine learning model is configured to determine the optimal communication channel through which to send the loan offer to the user based at least on identifying a communication channel corresponding to a highest predicted acceptance rate for respective communication channels of a plurality of candidate communication channels for users in the first group of users.
4 . The system as recited in claim 1 , wherein the optimal communication channel further includes a social media application messaging communication channel, a user application dashboard message communication channel, or a telephone call communication channel.
5 . The system as recited in claim 1 , wherein the first group is a first subset of users of the plurality users, wherein execution of the instructions further configure the one or more processors to:
determine an additional user characteristic to use for grouping the plurality of users; and
based at least on the additional user characteristic, reassign the first user from the first group of users to a second group of users, wherein the second group of users is a second subset of users of the plurality users, comprising one or more users not included in the first subset of users.
6 . The system as recited in claim 1 , wherein execution of the instructions further configure the one or more processors to assign the user to the first group of users based at least in part on the user sharing at least one user characteristic with the users in the first group of users by:
using a grouping data model to assign respective users of the plurality of users to respective user groups based at least on respective user data associated with the respective users, the grouping data model having been trained via a machine learning mechanism.
7 . The system as recited in claim 1 , wherein execution of the instructions further configure the one or more processors to:
determine an optimal timing for sending the loan offer to the user via the selected at least one communication channel, the optimal timing being based at least in part on loan data associated with one or more loans previously offered to the users in the first group of users.
8 . A method comprising:
receiving, by one or more processors, user data related to a user of a plurality of users, wherein the user data comprises one or more user characteristics associated with the user;
assigning, by the one or more processors, the user to a first group of users based at least in part on the user sharing at least one user characteristic with users in the first group of users;
tracking statistics associated with conversion of a plurality of loan offers previously offered to the first group of users and communication channels used for sending the plurality of loans, the statistics including a number of clicks, a number of website visits, and a number of interactions with prior push notifications;
determining, by the one or more processors, based at least in part on a machine learning model that is trained based at least on user data associated with users of the first group of users and the statistics to determine optimal communication channels for providing loan offers, an optimal communication channel through which to send a loan offer to the user, the optimal communication channel is one of an electronic mail communication channel, an electronic text message communication channel, and a push notification on a user device communication channel; and
sending, by the one or more processors, the loan offer to the user via the optimal communication channel determined based at least in part on the machine learning model.
9 . The method as recited in claim 8 , wherein the determining the optimal communication channel comprises determining a predicted acceptance rate for respective communication channels of a plurality of candidate communication channels for users in the first group of users.
10 . The method as recited in claim 9 , wherein the machine learning model is configured to determine the optimal communication channel through which to send the loan offer to the user based at least on identifying a communication channel corresponding to a highest predicted acceptance rate for respective communication channels of a plurality of candidate communication channels for users in the first group of users.
11 . The method as recited in claim 8 , wherein the optimal communication channel further includes a social media application messaging communication channel, a user application dashboard message communication channel, or a telephone call communication channel.
12 . The method as recited in claim 8 , wherein the first group is a first subset of users of the plurality users, the method further comprising:
determining an additional user characteristic to use for grouping the plurality of users; and
based at least on the additional user characteristic, reassigning the first user from the first group of users to a second group of users, wherein the second group of users is a second subset of users of the plurality users, comprising one or more users not included in the first subset of users.
13 . The method as recited in claim 8 , wherein the assigning the user to the first group of users based at least in part on the user sharing at least one user characteristic with the users in the first group of users further comprises:
using a grouping data model to assign respective users of the plurality of users to respective user groups based at least on respective user data associated with the respective users, the grouping data model having been trained via a machine learning mechanism.
14 . The method as recited in claim 8 , further comprising:
determining an optimal timing for sending the loan offer to the user via the selected at least one communication channel, the optimal timing being based at least in part on loan data associated with one or more loans previously offered to the users in the first group of users.
15 . One or more non-transitory computer-readable media comprising instructions executable by one or more processors to configure the one or more processors to:
receive user data related to a user of a plurality of users, wherein the user data comprises one or more user characteristics associated with the user;
assign the user to a first group of users based at least in part on the user sharing at least one user characteristic with users in the first group of users;
track statistics associated with conversion of a plurality of loan offers previously offered to the first group of users and communication channels used for sending the plurality of loans, the statistics including a number of clicks, a number of website visits, and a number of interactions with prior push notifications;
determine, based at least in part on a machine learning model that is trained based at least on user data associated with users of the first group of users and the statistics to determine optimal communication channels for providing loan offers, an optimal communication channel through which to send a loan offer to the user, the optimal communication channel is one of an electronic mail communication channel, an electronic text message communication channel, and a push notification on a user device communication channel; and
send the loan offer to the user via the optimal communication channel determined based at least in part on the machine learning model.
16 . The one or more non-transitory computer-readable media as recited in claim 15 , wherein execution of the computer-readable instructions further cause the one or more processors to determine the optimal communication channel by determining a predicted acceptance rate for respective communication channels of a plurality of candidate communication channels for users in the first group of users.
17 . The one or more non-transitory computer-readable media as recited in claim 15 , wherein the optimal communication channel further includes a social media application messaging communication channel, a user application dashboard message communication channel, or a telephone call communication channel.
18 . The one or more non-transitory computer-readable media as recited in claim 15 , wherein the first group is a first subset of users of the plurality users, and execution of the computer-readable instructions further cause the one or more processors to:
determine an additional user characteristic to use for grouping the plurality of users; and
based at least on the additional user characteristic, reassign the first user from the first group of users to a second group of users, wherein the second group of users is a second subset of users of the plurality users, comprising one or more users not included in the first subset of users.
19 . The one or more non-transitory computer-readable media as recited in claim 15 , wherein execution of the computer-readable instructions further cause the one or more processors to assign the user to the first group of users based at least in part on the user sharing at least one user characteristic with the users in the first group of users, by:
using a grouping data model to assign respective users of the plurality of users to respective user groups based at least on respective user data associated with the respective users, the grouping data model having been trained via a machine learning mechanism.
20 . The one or more non-transitory computer-readable media as recited in claim 15 , execution of the computer-readable instructions further cause the one or more processors to:
determine an optimal timing for sending the loan offer to the user via the selected at least one communication channel, the optimal timing being based at least in part on loan data associated with one or more loans previously offered to the users in the first group of users.