AUDIENCE LOCATION SCORING
A method to score audience location for an advertising campaign is provided. The method includes identifying, for a target audience group, a demographic data, a visit history, or a purchase history of a consumer within the target audience group, defining a target location for an advertising campaign based on a penetration of the target audience, training, to predict an exposure to the advertising campaign by the target audience, a model that includes multiple target locations within a retailer network, generating a score for a likelihood that the target audience will have sufficient exposure at one or more locations within a geographic zone, and providing, for a display in a client device, a map indicative of the score for the target locations. A system and a memory storing instructions which, when executed by a processor, cause the system to perform the above method, and the processor, are also provided.
1 . A computer-implemented method, comprising:
identifying, for a target audience group, a demographic data, a visit history, or a purchase history of a consumer within the target audience group;
defining a target location for an advertising campaign based on a penetration of the target audience group;
training, to predict an exposure to the advertising campaign by the target audience group, a model that includes multiple target locations within a retailer network;
generating, with the model, a score for a likelihood that the target audience group will have sufficient exposure at one or more locations within a geographic zone; and
providing, for a display in a client device, a map indicative of the score for the target locations.
2 . The computer-implemented method of claim 1 , wherein identifying a demographic data for the target audience group comprises extracting a binary value for the target audience group in one of an urban category, a suburban category, and a rural category.
3 . The computer-implemented method of claim 1 , wherein validating the target location further comprises assessing a performance of a model that generates the score based on a known feature.
4 . The computer-implemented method of claim 1 , further comprising aggregating the demographic data, the visit history, or the purchase history to a specific geographic area or census block group value associated with the consumer.
5 . The computer-implemented method of claim 1 , wherein defining a target location for an advertising campaign comprises defining the target location based on information describing where the target audience group lives, a visit place for the target audience group, or a purchase item associated with the target audience group.
6 . The computer-implemented method of claim 1 , wherein training a model to predict an exposure to the advertising campaign by the target audience group comprises selecting at least one of a demographic pattern within a specific geographic area, an urbanicity level within the specific geographic area, and a visiting pattern within the specific geographic area, independent of a pre-selected target audience group.
7 . The computer-implemented method of claim 1 , wherein identifying a visit history comprises receiving a longitude and latitude information of a mobile device of a consumer from a server hosting a location application installed in the mobile device of the consumer.
8 . The computer-implemented method of claim 1 , wherein identifying a visit history comprises determining a radius threshold from a centroid that a selected portion of consumers within a demographic segment are willing to travel to purchase a type of product.
9 . The computer-implemented method of claim 1 , wherein identifying the demographic data, the visit history, or the purchase history of a consumer within the target audience group comprises selecting the consumer that is subscribed to the retailer network.
10 . The computer-implemented method of claim 1 , further comprising determining a list of demographic features for a consumer sorted by a weight factor indicative of a likelihood that the consumer purchases a type of product based on a weighted average.
11 . A system, comprising:
one or more processors; and
a memory storing instructions which, when executed by the one or more processors, cause the system to perform operations, comprising to:
identify, for a target audience group, a demographic data, a visit history, or a purchase history of a consumer within the target audience group;
define a target location for an advertising campaign based on a penetration of the target audience group;
train, to predict an exposure to the advertising campaign by the target audience group, a model that includes multiple target locations within a retailer network;
generate, with the model, a score for a likelihood that the target audience group will have sufficient exposure at one or more locations within a geographic zone; and
provide, for a display in a client device, a map indicative of the score for the target locations.
12 . The system of claim 11 , wherein to identify a demographic data for the target audience group the one or more processors execute instructions to extract a binary value for the target audience group in one of an urban category, a suburban category, and a rural category.
13 . The system of claim 11 , wherein to validate the target location the one or more processors further execute instructions to assess a performance of a model that generates the score based on a known feature.
14 . The system of claim 11 , wherein the one or more processors further execute instructions to aggregate the demographic data, the visit history, or the purchase history to a specific geographic area value associated with the consumer.
15 . The system of claim 11 , wherein to define a target location for an advertising campaign the one or more processors execute instructions to define the target location based on information describing where the target audience group lives, a visit place for the target audience group, or a purchase item associated with the target audience group.
16 . A computer-implemented method, comprising:
retrieving, from multiple consumers, at least one of an urbanicity feature, a visit history feature, and a demographic feature, the urbanicity feature associated with a population density, and the visit history feature associated with consumer visits to a store;
identifying a centroid for a common geographic area;
aggregating the urbanicity feature, the visit history feature, or the demographic feature based on the centroid;
determining an event predictor value associated with the centroid based on at least two features aggregated from the urbanicity feature, the visit history feature, and the demographic feature; and
forming a map including the common geographic area, indicative of a geographic distribution of the event predictor value.
17 . The computer-implemented method of claim 16 , wherein retrieving an urbanicity feature comprises retrieving at least one of an urban value, a suburban value, and a rural value.
18 . The computer-implemented method of claim 16 , further comprising retrieving a store location within the common geographic area, and selecting a radius from the store, wherein the visit history feature is associated with a consumer that resides within the radius from the store.
19 . The computer-implemented method of claim 16 , wherein determining an event predictor value associated with the centroid comprises correlating the urbanicity feature, the demographic feature, and the visit history feature.
20 . The computer-implemented method of claim 16 , wherein determining an event predictor value comprises reducing a dimension of the visit history feature.