Using media information for improving direct marketing response rate

• US 10,460,334 B2
• Filed: 02/22/2013
• Issued: 10/29/2019
• Est. Priority Date: 02/22/2013
• Status: Active Grant

• Current Assignee: DoorDash, Inc.
• Original Assignee: International Business Machines Corporation
• Inventors: Haber, Eben M., Mahmud, Jalal U.
• Primary Examiner(s): Boyce, Andre D
• Application Number:
  US13/774,226
• Publication Number:
  US 20140244359A1
• Time in Patent Office: 2,440 Days
• Field of Search: 705 733
• CPC Class Codes: G06Q 30/0204 Market segmentation

Abstract

Embodiments of the invention relate to improving direct marketing response rate through use of media information. One embodiment includes extracting samples of language usage in one or more social media activities. Language usage features comprising relationships to products are derived by analyzing the extracted samples for relevant language terms. The language usage features are mapped to one or more of personality traits, emotional state and personal features. The mapped language usage features and personal features are stored in a repository. Customers are segmented for direct marketing using the mapped language usage features and the personal features.

Claims

1. A method comprising:

at a hardware processor configured to execute computer program instructions stored on a non-transitory computer readable medium;

training a prediction model to compute a probability of customer response to direct marketing, wherein the training comprises;

extracting samples of language usage in one or more social media activities performed by customers and potential customers on one or more online social media platforms;

deriving language usage features indicative of product interest of one or more of the customers and potential customers in one or more products by searching text of the extracted samples for mentions of the one or more products, and converting speech responses of the extracted samples to text using speech recognition to search for mentions of the one or more products;

mapping the language usage features to personal features including one or more personality traits and emotional states;

storing the mapped language usage features and the personal features in a repository comprising a storage device;

extracting time and date information for the one or more social media activities;

deriving temporal activity patterns based on statistics for the extracted time and date information, wherein the temporal activity patterns are indicative of particular days of week and times of day that one or more of the customers and potential customers are most likely to be using the one or more online social media platforms, and the temporal activity patterns are stored in the repository; and

training the prediction model based on the temporal activity patterns; and

improving rate of customer response to direct marketing, wherein the improving comprises;

determining, based on the trained prediction model, probabilities of the customers and potential customers receiving and viewing direct marketing information at the particular days of week and times of day;

identifying at least one customer or potential customer of the customers and potential customers who is not ready to receiving direct marketing information by determining, based on the probabilities, at least one of;

device readiness of at least one customer or potential customer, device connectivity of the at least one customer or potential customer, or whether the at least one customer or potential customer is engaged in a task that is interruptible;

filtering out the at least one customer or potential customer who is not ready to receive direct marketing information; and

timing one or more electronic offers of the one or more products to one or more remaining customers or potential customers of the customers and potential customers who are ready to receive direct marketing information via the one or more online social media platforms in accordance with the particular days of week and times of day.

2. The method of claim 1, wherein the one or more social media activities comprise one or more of writing activities and speech interactions.

3. The method of claim 2, wherein the one or more writing activities comprise one or more of social media written interactions, personality questionnaires, or surveys.

4. The method of claim 3, wherein the one or more speech interactions comprise speech used in one or more interactions with a call center.

5. The method of claim 4, wherein the language usage features are mapped to the personal features using one or more of Big 5 and linguistic inquiry and word count (LIWC) techniques.

6. The method of claim 5, wherein the extracted samples comprise:

language usage comprising language related at least one of product names of the one or more products and competing product names, and relationships to the one or more products comprising at least one of an indication for liking or disliking the one or more products, indication of ownership of the one or more products, and indication of intent to purchase the one or more products;

language usage recently used, on the one or more online social media platforms, by the one or more remaining customers or potential customers who are ready to receive direct marketing information including language indicative of personality type and emotional state of the one or more remaining customers or potential customers in relation to the one or more products; and

the language usage recently used, on the one or more online social media platforms, by the at least one customer or potential customer who is not ready to receive direct marketing information including at least one of;

a status message indicating a device connectivity problem associated with a device utilized by the at least one customer or potential customer, a status message indicating a device readiness problem associated with the device utilized by the at least one customer or potential customer, or a status message indicating that the at least one customer or potential customer is busy.