Correlated consumer telephone numbers and user identifiers for advertising retargeting

A method in a computing system of targeting a consumer with advertising via an online channel based on a telephone number associated with the consumer that is identified via an interaction between the consumer and a business that did not occur via an online channel, the method comprising:

• establishing, via interactions of users with publishers, correlations between user identifiers and caller telephone numbers, wherein each user identifier is associated with a user accessed a publisher website or application over a data channel, each caller telephone number is associated with a telephone used by a user, and establishing each correlation includes:
• identifying an advertising time when a first advertisement having an advertising telephone number is shown to a consumer on a computing device associated with a correlating user identifier,
• identifying a call time when a call is received on the advertising telephone number from a caller telephone number,
• determining a numerical confidence level, characterizing a probabilistic degree of confidence that the consumer associated with the correlating user identifier is the same consumer as is associated with the telephone number, based on the amount of time elapsed between the advertising time and the call time, wherein the probabilistic degree of confidence is inversely related to the amount of time elapsed, and
• correlating the respective user identifier with the respective caller telephone number if the call time is after the advertising time;
• storing the established correlations and corresponding numerical confidence levels in a cross-channel targeting database accessible to the one or more processors;
• receiving, from a business, a telephone number associated with a consumer and a request to send an advertisement to the consumer, the telephone number derived from a prior interaction between the business and consumer that did not occur via an online channel;
• retrieving, using the one or more processors, at least one user identifier stored ion the cross-channel targeting database that corresponds to the received telephone number, wherein retrieving the at least one identifier includes searching the cross-channel targeting database for established correlations that include a caller telephone number that matches the receive telephone number, identifying one or more user identifiers identifying one or more user identifiers and one or more associated corresponding numerical confidence levels included in the established correlations that include the caller telephone number that matches the received telephone number, and selecting the at least one user identifier from the identified one or more user identifiers;
• comparing the one or more corresponding numerical confidence levels to a desired threshold confidence level;
• and
• when a numerical confidence level of the one or more corresponding numerical confidence levels exceeds the desired threshold confidence level, transmitting a request to the corresponding one or more publishers to cause a second advertisement, via a publisher website or application, to the user associated with the one user identifier, wherein the transmitted request includes the at least one user identifier.

A method in a computing system of targeting a consumer with advertising based on a telephone number associated with the consumer, the method comprising:

• maintaining a correlation between masked user identifiers and caller telephone numbers in a cross-channel targeting database that is accessible to the one or more processors, wherein:
• each masked user identifier is provided by a publisher and associated with a computing device used by a consumer to access a publisher website or application over a data channel, the masked user identifier being assigned by the publisher so that the publisher maintains a mapping between the masked user identifier and an actual identifier identifying the computing device;
• and
• each caller telephone number is associated with a telephone used by a consumer, and
• each correlation established by:
• logging an advertising time when a first advertisement having an advertising telephone number is shown to a consumer on a computing device associated with a respective masked user identifier,
• identifying, using one or more processors, a call time when a call is received on the advertising telephone number from a caller telephone number,
• comparing, using one or more processors, the advertising time to the call time,
• correlating the respective masked user identifier with the respective caller telephone number when the call time is after the advertising time, and
• calculating, using one or more processors, a numerical confidence level, characterizing the probabilistic degree of confidence that the consumer associated with the respective masked user identifier is the same consumer as is associated with the respective caller telephone number, wherein calculating the numerical confidence level includes calculating an amount of time elapsed between the advertising time and the call time, and wherein the probabilistic degree of confidence is inversely related to the amount of time elapsed;
• storing the maintained correlations and the numerical confidence levels in a cross-channel targeting database;
• receiving a telephone number associated with a consumer and a request to send an advertisement to the consumer;
• retrieving, using the one or more processors, one or more masked user identifiers stored in the cross-channel targeting database that correspond to the received telephone number, wherein the retrieving the one or more masked user identifiers includes searching the cross-channel targeting database for maintained correlations that include the received telephone number as a caller telephone number, identifying one or more masked user identifiers and one or more associated numerical confidence levels included in the maintained correlations that include the received telephone number and identifying one or more publishers associated with the one or more masked identifiers;
• selecting, using the one or more processors, at least one masked user identifier from the one or more retrieved masked user identifiers;
• comparing, using the one or more processors, the one or more associated numerical confidence levels corresponding to the selected at least one masked user identifier to a threshold probability;
• and
• when the corresponding one or more associated numerical confidence levels exceed the threshold probability, transmitting, at least in part using the one or more processors, a request to the publisher(s) corresponding to the selected at least one masked user identifier to cause a second advertisement to be presented to the consumer associated with the selected at least one masked user identifier via a publisher website or application.