5. Beyond the Social Listener No Actionable Data Marketing Centric No Link to Contact Center Indexing and Sentiment Analysis Social & Community Monitoring 3
12. Decisioning Based On: Sentiment Social Activity Influence Real-time Decisioning Service History Lifetime Value Geography Purchases 25,000 Social Keyword Results Should Not Equal 25,000 Service Requests Business Rules Cases and SRs Why a Social Decisioning Engine? 5
13. Applying Our Contact Center Operational Best Practices 1,000s x per day Queuing & Routing Decisioning Matching Natural Language Processing @JetBlue just lost my golf clubs in Phoenix. Awesome! Missing Southwest about now. Natural Language vs. Keywords Signal vs. Noise Intelligent Engagements Optimizing Social for the Contact Center 6
23. The Front Lines of Social Engagement Actionable Social Data Based on Rules and Policies Community Advocate Controlled Environment Social Advocate Contributing to Social Conversation 8
24. Assisted Network Outage Use Case Actionable Listening Realtime Clustering Natural Language Processing KPI Alerts Recommended Action Proactive Self-Service Intelligent Routing to Tier 2 Engineer Prioritized Queue 9
Let’s talk a little about what’s happening today-> There are tons of listeners on the marketYou could spend nothing, or into the millionsBut all of them are relatively the same, with no real ACTION-> so, what do I do next? While dashboards are nice, there needs to be a link to operationalizing the data to improve the overall customer experience
Take any of the brands each of you represent…Even with rudimentary key word analysis it’s too much on the case creation frontBut more importantly, how do you Service the brand detractors, but really amplify the brand enthusiasts?One of the ways is to consider a larger picture of the person’s profile and value (right portion of the slide)
As team sizes grow that address social interaction, the same problems we’ve been addressing for years on the (voice) routing and queuing front, ariseMore importantly, the lack of efficiency of having dedicated personnel manage only Twitter for example may become problematic, or at minimum, expensive- a much more efficient approach is melding the media types and having multi-skilled agents handling multiple types of interactions (email, chat, text and social… not just social)Finally NLP provides the filter and intelligent dialog to cluster and identify that signal that must be addressed, shutting out or deferring the large amounts of noise (and now spam) that is becoming more and more prevalent
The technology offering is really enhanced by two different and distinct roles, powered with actionable dataCommunity Advocates work exclusively within a defined environment, typically a discussion group on your web site:Responsible for Identifying the Question type or General Discussion also if in the right category, and propose responseChecking for Abusive ContentPosting the reply on the site within agreed upon SLASocial Advocates work in a much more public domain such as Twitter or Facebook. Similarly, they handle responses, but are much more aware and trained in adherence to company/legal/support policiesThey will also likely see large spikes in interaction traffic and be responsible for facilitating appropriate response
1. Network outageBlogs, Twitter, FB goes crazyMeanwhile, ABC Service provider corporate not even aware. Business as usual. CxP is being damaged.2. Convergys comes in…Actively listening and doing realtime clustering of posts, consumer generated media, applying NLP and KPI alertsSocial Decisioning Engine takes recommended actions which are: - prioritized routing to Tier 2 network support - prioritized queue for customers calling in from Northeast (affected area) based on account+zip lookup in IVR3. Within minutes:Network engineers work on issue Customer care post status on ABC_Corp.com support page, Facebook Fan page, Twitter page with ETA for service restoration, IVR menus updated with network outage message and ETA