US 12,086,835 B2
Cross-channel personalized marketing optimization
Shike Mei, San Carlos, CA (US); Teng Wang, Millbrae, CA (US); Shawn Chen, Millbrae, CA (US); and Ganesh Venkataraman, San Jose, CA (US)
Assigned to AIRBNB, INC., San Francisco, CA (US)
Filed by Airbnb, Inc., San Francisco, CA (US)
Filed on Apr. 4, 2022, as Appl. No. 17/657,869.
Application 17/657,869 is a continuation of application No. 16/368,691, filed on Mar. 28, 2019, granted, now 11,295,345.
Prior Publication US 2022/0222713 A1, Jul. 14, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 30/02 (2023.01); G06Q 30/0241 (2023.01); G06Q 30/0242 (2023.01); G06Q 30/0251 (2023.01)
CPC G06Q 30/0271 (2013.01) [G06Q 30/0246 (2013.01); G06Q 30/0249 (2013.01); G06Q 30/0276 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A system for selecting a digital channel for delivery of advertising content to a user, the system comprising:
a memory configured to store data relating to a user's activity on a first advertising channel and a second advertising channel;
a processor coupled to the memory, the processor being configured to:
(a) create a training set based on the data relating to the user's activity on the first advertising channel and the second advertising channel;
(b) train a machine learning model for modeling user intent using the training set:
(c) collect a first plurality of user metrics relating to the user's activity on the first advertising channel and a second plurality of user metrics relating to the user's activity on the second advertising channel;
(d) generate, based on the first plurality of user metrics, a first channel budget value;
(e) generate, based on the second plurality of user metrics, a second channel budget value;
(f) determine, based on the first plurality of user metrics and the second plurality of user metrics, a channel overlap value representative of an effect of advertising on both of the first advertising channel and the second advertising channel;
(g) calculate, in real-time:
(1) using the machine learning model, an overlap weight for the channel overlap value; and
(2) a multi-channel budget value, the multi-channel budget value being a user-specific budget for delivering advertising content to the user across the first advertising channel and the second advertising channel, wherein the multi-channel budget value is a function of the first channel budget value, the second channel budget value, and the channel overlap value weighted by the overlap weight; and
(h) select, based on the calculated multi-channel budget value, a channel from among the first advertising channel and the second advertising channel, for delivery of advertising content.