Multi-objective electronic communication frequency optimization
Methods and systems are provided for improved electronic communication campaign technologies, which can automatically balance objectives or goals of an electronic communication campaign against an overall opt-out rate for the electronic communication campaign. An electronic communications frequency optimizer can generate individual contact frequencies for individual email recipients. Embodiments can avoid unnecessary or counterproductive communications while achieving overall campaign goals, and can use processes to improve the efficiency of systems. In some cases, embodiments cluster communication recipients into different groups based on their past actions, then optimizes the communication contact frequency on different groups, to avoid performing optimization directly on millions of recipients. Some embodiments automatically self-update, for example with recipients' recent responses, to generate and/or implement campaign communication schedules on an individual level.
1. A method for scheduling electronic communications to be sent to a first individual, the method comprising:
obtaining a first electronic communication campaign goal, for an electronic communication campaign, to maximize an amount of electronic communications opened by electronic communication recipients, the first electronic communication campaign goal received by a system comprising an electronic communication campaign interface and a content module accessed over a network, wherein the content module comprises one or more datasets;
obtaining an opt-out rate tolerance for the electronic communication campaign;
for a plurality of electronic communication recipients, determining a group frequency optimization based on the first electronic communication campaign goal and the opt-out rate tolerance for the electronic communication campaign, the group frequency optimization indicating a frequency, determined convexly, at which to communicate electronic communications to the plurality of electronic communication recipients, wherein the group frequency optimization includes a constraint based on responses to a first set of emails, wherein the constraint incorporates a first relevance score associated with a plurality of subject lines in the first set of emails and a second relevance score associated with a plurality of email bodies in the first set of emails;
determining an individual electronic communication frequency for an individual of the plurality of electronic communication recipients, the individual electronic communication frequency determined by applying the group frequency optimization to a prior electronic communication frequency associated with the individual;
receiving a customization of a value associated with the first electronic campaign goal, wherein the customization is input at a remote client device via the electronic communication campaign interface;
automatically determining an updated group frequency optimization, based at least in part on the value associated with the first electronic campaign goal;
automatically updating the individual electronic communication frequency based on the updated group frequency optimization; and
initiating communication, via the network, of an electronic communication in accordance with the updated individual electronic communication frequency.
2. The method of claim 1 , wherein the first electronic communication campaign goal is a total electronic communication open count maximization.
3. The method of claim 1 , wherein the first electronic communication campaign goal is a total electronic communication click count maximization.
4. The method of claim 1 , wherein the first electronic communication campaign goal is a total conversion maximization.
5. The method of claim 1 , wherein determining a group frequency optimization includes considering input from a plurality of probability models.
6. The method of claim 5 , wherein the plurality of probability models includes an electronic communication open probability model.
7. The method of claim 6 , wherein the plurality of probability models further includes an opt-out probability model.
8. The method of claim 1 , wherein at least one of the plurality of subject lines in the first set of emails relates to a purchase and at least one of the plurality of email bodies in the first set of emails does not relate to the purchase.
9. The method of claim 1 , wherein the first relevance score and the second relevance score indicate a consistency between the plurality of subject lines in the first set of emails and the plurality of email bodies in the first set of emails.
10. The method of claim 5 , wherein at least one of the plurality of probability models comprises a machine learning model.
11. A method for scheduling electronic communications to be sent to a recipient, the method comprising:
obtaining a first electronic communication campaign goal, the first electronic communication campaign goal received by a system comprising a content module accessed over the network, wherein the content module comprises a first dataset;
obtaining an opt-out rate for the electronic communication campaign;
for a plurality of electronic communication recipients, determining, at least in part based on a convex calculation to reduce computational resource utilization, a group frequency optimization and a frequency capping, the frequency capping based on responses to a first set of emails that contain hyperlink descriptions, and wherein the frequency capping is for a certain amount of time;
determining an individual electronic communication frequency for an individual of the plurality of electronic communication recipients, the individual electronic communication frequency determined by applying the group frequency optimization to a prior electronic communication frequency associated with the individual;
receiving a customization of a value associated with the first electronic campaign goal, wherein the customization is input at a remote client device via the electronic communication campaign interface;
automatically determining an updated group frequency optimization, based at least in part on the value associated with the first electronic campaign goal;
automatically updating the individual electronic communication frequency based on the updated group frequency optimization; and
initiating communication, via the network, of an electronic communication in accordance with the updated individual electronic communication frequency.
12. A method for scheduling electronic communications to be sent to a first individual, the method comprising:
obtaining a first electronic communication campaign goal, for an electronic communication campaign, to maximize an amount of electronic communications opened by electronic communication recipients, the first electronic communication campaign goal received by a system comprising an electronic communication campaign interface and a content module accessed over a network, wherein the content module comprises one or more datasets;
obtaining an opt-out rate tolerance for the electronic communication campaign;
for a plurality of electronic communication recipients, determining a group frequency optimization based on the first electronic communication campaign goal and the opt-out rate tolerance for the electronic communication campaign, the group frequency optimization determined at least in part based on a convex optimization comprising minimizing a plurality of convex functions, and indicating a frequency at which to communicate electronic communications to the plurality of electronic communication recipients, wherein the group frequency optimization includes a cap that is applied based on responses to a set of sent emails and a comparison of one or more subject lines of the set of sent emails to one or more email bodies of the set of sent emails;
determining an individual electronic communication frequency for an individual of the plurality of electronic communication recipients, the individual electronic communication frequency determined by applying the group frequency optimization to a prior electronic communication frequency associated with the individual;
receiving a customization of a value associated with the first electronic campaign goal, wherein the customization is input at a remote client device via the electronic communication campaign interface;
automatically determining an updated group frequency optimization, based at least in part on the value associated with the first electronic campaign goal;
automatically updating the individual electronic communication frequency based on the updated group frequency optimization; and
initiating communication, via the network, of an electronic communication in accordance with the updated individual electronic communication frequency.
13. The method of claim 1 , wherein the first relevance score and the second relevance score are determined from machine learning models trained on the first set of emails.