Method and apparatus for optimizing customer service across multiple channels

1. A computer-implemented method for containing users within a Web environment where issues can be resolved, the method comprising:

monitoring a behavioral attribute of a user on a Web page in a multi-channel system comprising each of a call channel, a chat channel, a self-service channel, and an assisted self-service channel,

wherein the user is on the Web page attempting to resolve an issue, and

wherein the behavioral attribute is based at least in part on click activities performed by the user, hover activities performed by the user for a predetermined duration, or both;

based on said monitoring, determining that the user is unable to find a solution to the issue on the Web page; and

executing, by a Web containment processor, a Web containment strategy designed to automatically retain the user on the Web page rather than initiate an interaction with a live agent associated with a call center when the Web containment strategy is successful;

wherein said retaining is effected by proactively:

inviting the user to the chat channel having an automated feature;

presenting the user with a self-service opportunity on the self-service channel; or

presenting the user with an assisted self-service opportunity on the assisted self-service channel; and

wherein said executing causes the Web containment processor to:

identify a top call driver Web page associated with a top call driver,

wherein the top call driver represents a particular issue responsible for driving a highest number of calls to the call center; and

identify Web journeys that correspond to the top call driver Web page,

wherein the Web journeys represent Web traversal footprints associated with visitors to the top call driver Web page.

2. The method of claim 1 , wherein said executing further causes the Web containment processor to:

compute Web traffic on the top call driver Web page and on each Web page included in each Web journey; and

compute an estimation of chat volume based at least in part on the Web traffic.

3. The method of claim 1 , further comprising:

executing, by a targeting processor, a targeting user strategy,

wherein executing the targeting user strategy comprises

applying a derived statistical model to compute a chat acceptance score for the user; and

making an offer to interact with the user on the chat channel when the chat acceptance score exceeds a particular threshold.

4. The method of claim 1 , wherein said executing further comprises:

determining, by a processor, voice call drivers based on customer relationship management (CRM) data,

wherein each voice call driver represents a separate issue that causes users to call the call center;

determining, by the processor, an amenability of the voice call drivers to be resolved on the Web page; and

targeting, by the processor, one or more particular issues based on their amenability to being resolved over the Web.

5. The method of claim 4 , further comprising:

identifying, by the processor, customer journeys corresponding to the one or more particular issues;

determining, by the processor, uniform resource locators (URLs) corresponding to the customer journeys; and

determining, by the processor, a hot lead rate for each URL,

wherein the hot lead rate represents a percentage of users to be targeted for retainment by the Web containment strategy, and

wherein the hot lead rate is based at least in part on a hierarchical level of each URL, customer behavioral attributes, and customer journey attributes.

6. The method of claim 1 , further comprising:

determining, by a processor, an acceptance rate from a multivariate model; and

using the acceptance rate to determine when to present the user with an invitation to the chat channel.

7. The method of claim 1 , further comprising:

identifying, by a processor, and servicing, by a servicing processor, optimal segments of URLs based at least in part on exit rates and user ratings of the URLs.

8. The method of claim 1 , further comprising:

text mining, by a processor, transcripts of chat sessions between users and live agents of the call center to identify the top call driver; and

identifying, by the processor, Web pages associated with the top call driver to target for deployment of the Web containment strategy.

9. An apparatus for effecting containment of users to a Web environment, the apparatus comprising:

a Web containment processor configured to

execute a Web containment strategy for a user on a Web page in a multi-channel system comprising each of a call channel, a chat channel, a self-service channel, and an assisted self-service channel,

wherein the user is on the Web page attempting to resolve an issue;

monitor a behavioral attribute of the user,

wherein the behavioral attribute is based at least in part on click activities performed by the user, hover activities performed by the user for a predetermined duration, or both;

determine, based on said monitoring, that the user is unable to find a solution to the issue on the Web page;

retain the user on the Web page rather than initiate an interaction with a live agent associated with a call center when the Web containment strategy is successful;

wherein the Web containment processor is configured to effect said retaining by executing any of:

inviting the user to the chat channel having an automated feature;

presenting the user with a self-service opportunity on the self-service channel; and

presenting the user with an assisted self-service opportunity on the assisted self-service channel;

identify a top call driver Web page associated with a top call driver,

wherein the top call driver represents a particular issue responsible for driving a highest number of calls to the call center; and

identify Web journeys that correspond to the top call driver Web page,

wherein the Web journeys represent Web traversal footprints associated with visitors to the top call driver Web page.

10. The apparatus of claim 9 , wherein the Web containment processor is further configured to:

determine Web traffic on the top call driver Web page and on each Web page included in each Web journey; and

based at least in part on the Web traffic, determine an estimation of chat volume.

11. The apparatus of claim 9 , further comprising:

a targeting processor configured for executing a targeting user strategy, the targeting user strategy including

applying a derived statistical model to compute a chat acceptance score for the user; and

making an offer to interact with the user on the chat channel when the chat acceptance score exceeds a particular threshold.

12. The apparatus of claim 9 , wherein the Web containment processor is further configured to:

determine voice call drivers based on customer relationship management (CRM) data, wherein each voice call driver represents a separate issue that causes users to call the call center;

determine an amenability of the voice call drivers to be resolved on the Web page; and

target one or more particular issues based on their amenability to being resolved over the Web.

13. The apparatus of claim 12 , further comprising:

a processor configured for identifying customer journeys corresponding to the one or more particular issues;

the processor configured for determining uniform resource locators (URLs) corresponding to the customer journeys; and

the processor configured for determining a hot lead rate for each URL,

wherein the hot lead rate represents a percentage of users to be targeted for retainment by the Web containment strategy, and

wherein the hot lead rate is based at least in part on a hierarchical level of each URL, customer behavioral attributes, and customer journey attributes.

14. The apparatus of claim 9 , further comprising:

a processor configured for determining an acceptance rate from a multivariate model and using the acceptance rate to determine when to present the user with an invitation to the chat channel.

15. The apparatus of claim 9 , further comprising:

a processor configured for identifying and servicing optimal segments of URLs based at least in part on exit rates and user ratings of the URLs.

16. The apparatus of claim 9 , further comprising:

a processor configured for text mining transcripts of chat sessions between users and live agents of the call center to identify the top call driver;

the processor configured for identifying Web pages associated with the top call driver to target for deployment of the Web containment strategy.

17. A computer-implemented method for effecting containment of users to a Web-based customer support environment where issues can be resolved, the method comprising:

gathering voice call data for a voice call placed to a call center,

wherein the voice call data comprises customer relationship management (CRM) data, and

wherein the CRM data comprises transaction data related to the voice call, including any of call time, date, total call handle time, agent who took the call, call issue type handled, attributes of a caller who placed the call, attributes of the agent who took the call, and attributes of prior history of interactions with the caller;

identifying a top call driver from the voice call data,

wherein the top call driver represents an issue responsible for driving a largest volume of calls to the call center;

identifying a Web journey corresponding to the top call driver,

wherein the Web journey includes one or more electronic footprints left by visitors to a website seeking to resolve the issue, and wherein each electronic footprint specifies a referral Web page, a landing Web page, an exit Web page, a clickstream of visited Web pages, a wait time corresponding to each visited Web page, a count of visited Web pages, or any combination thereof;

after the Web journey corresponding to the top call driver is identified, computing traffic on Web pages related to the top call driver and on Web pages corresponding to the Web journey;

based upon gathered Web analytic data and the traffic, analyzing behavioral attributes of the visitors to the Web pages related to the top call driver,

wherein the behavioral attributes include time spent by the visitors on the Web pages related to the top call driver, hover activity, and clicks performed; and

based upon the behavioral attributes, performing an operation that includes at least one of:

inviting a visitor to a chat channel having an automated feature;

presenting the visitor with a self-service opportunity on a self-service channel; and

presenting the visitor with an assisted self-service opportunity on an assisted self-service channel,

wherein performance of the operation causes the visitor to remain within a Web-based customer support environment to seek a resolution to the issue rather than initiate an interaction with a live agent associated with the call center.

18. The method of claim 17 , further comprising:

determining a percentage of visitors who end up chatting for a given number of visitors for a particular vertical, based upon a percentage of visitors who are invited to chat and a percentage of the visitors who are served.

19. The method of claim 17 , further comprising:

estimating chat volume; and

once the estimation of chat volume is determined, determining a number of agents associated with the call center that is required to effect containment of at least some visitors to the Web-based customer support environment.

20. The method of claim 17 , wherein the traffic comprises a complete distribution of visitors to each Web page and a distribution of how long the visitors spend on each Web page.