Blended Community and Social for Customer Care

By Edwin Margulies Peer-to-Peer Community software has proven to be a great way to leverage the power and knowledge of your customers, advocates, and experts in a collaborative way. Whether you're sharing threaded discussion forums, ideas, blogs and wikis - peer-to-peer software is great way to embrace and support your customers. But until now, the big, blue social ocean of Twitter, Facebook, Google+ and others are in their own wild domain with separate listening and engagement platforms. The good news: You can now blend peer-to-peer community and social channels together with a unified agent and supervisor interface. How is Peer-to-Peer Community different from Social? The whole idea with building a community is for your company to "host" a collaborative meeting place where ideas, advice, and self-service can be shared by your support team, community experts, and of course your customers themselves. For example, companies like Dell, Apple, and Cox Communications have successful implementations of community software. The prevailing model is for the hosting or sponsoring company to pay for the cost to install and maintain the software and to also offer experts to moderate forums or post answers to questions. Community sites share the same "crowdsourcing" model as big blue ocean social, because the denizens of community sites are primarily the customers themselves - and they like to share their experiences and tips with everyone else. Most communities that are sponsored by enterprises also have moderators or experts who work for the sponsoring company. These experts should clearly double as social care agents too. The difference between Community and pure Social is that peer-to-peer community sites are much more structured and have built-in mechanisms for offering knowledge bases and other self-service outlets. Discussion forums have full threading capability and topics can have long tails that are easier to follow than on native social channels. On the other hand, Tweets, Blogs and non-fan-page Facebook mentions are much less structured and kind of like flying fish that sometimes fall into your boat, but not all the time. Social engagement for customer care platforms attempt to put structure around big, blue ocean social posts, but it's still not as well-behaved as a peer-to-peer community. So What's the Big Deal about Blending these Models? It is a boon to productivity to have a unified, blended agent and supervisor interface to handle the myriad conversations, posts, threads, ideas and responses for all of these channels. I predict that in the future this will be commonplace, but now it's best-in-class to have this blended capability. For many enterprises, it is a nightmare to expect your service experts to "swivel chair" betwixt multiple interfaces. That is to say: a separate interface for community responses, a separate interface for each Facebook fan page, a separate interface for each twitter handle, etc. Counting the Benefits of Blended Social and Community Let's take a quick look at four of the more obvious benefits you can take advantage of with a blended approach to peer-to-peer community and big, blue ocean social engagement: 1. Unified Interfaces mean More Productive Agents 2. Leveraging Centralized Natural Language Processing 3. Common Rules Engine means Comprehensive Actions 4. Unified Role-Based Analytics means Better Supervision Unified Interfaces mean More Productive Agents Having a unified agent and supervisor interface means higher productivity for your staff. Instead of jumping around from one interface to another, imagine being able to filter easily on a common media stream that includes community chatter, blogs, forums, tweets, and Facebook mentions all in one, single interface. A not-so obvious benefit of this unified interface is the ability to share common agent assistance and knowledge base articles and to be able to push those articles to users - no matter what the source. For example, you can use the knowledge base of the community and push a KB article URL to a Twitter user. That has the effect of "reeling-in" that person from the big, blue social ocean and getting them to swim in your community pool. The same idea applies to shared agent assistance scripts and the use of URL reduction technology. Being able to use these tools across the board not only saves time for your agents, but it helps to ensure a comprehensive and standardized outbound messaging schema. Leveraging Centralized Natural Language Processing Most peer-to-peer community software does not have built-in Natural Language Processing (NLP/NLU) capability. That is the ability to automatically categorize content, automatically de-spam posts, automatically understand sentiment, or establish trending topics. Because big, blue ocean social posts are more unstructured - NLP is pretty essential - otherwise you're just swimming in non-actionable spam. Now consider merging community and big, blue ocean social posts together. Imagine being able to apply spam control, sentiment and automatic clustering to each and every piece of content. You can now establish trending topics across the board and better anticipate issues that are becoming important to your constituency. Ditto keeping a better eye on the sentiment of posts and having a unified time line of how sentiment changes over time with each customer. Common Rules Engine means Comprehensive Actions Let's face it. Your enterprise has its own set of policies, rules and best practices for each given situation. You have these rules even if you don't commit them to an automate rules engine. But now consider being able to trigger automatically on the rules you have set up and do this across the board on community posts and social posts simultaneously. Take, for example, this rule: "Any unhappy customer who complains about a newly-launched product, should be automatically routed to a special brand protection team and offered an instant replacement." Wow, wouldn't it be great if all of the posts in your system - community or social - could be scrutinized by that rule and automatically filtered and routed based on that rule? Wouldn't it also be great if after the filters have done their job, the agent got a next best action script to tell them what to say or do? The good news: This is all a reality in the most modern, blended community and social engagement software offerings. Unified Role-Based Analytics means Better Supervision To round-out the benefit parade, let's consider how much easier it is to supervise and coach your customer care professionals with united reporting. For example, simple agent KPI (Key Performance Indicator) metrics can be gathered across the board. This includes post handle time, resolution time, and outcome reporting. Another benefit with unified reporting includes advanced outreach data by agent. Here, supervisors are able to view the sentiment breakdown of customer outreach on an agent-by-agent basis. Now imagine being able to see which agents are doing the best job of reaching out to angry customers - regardless of source. The idea here is that role-based analytics with a blended community/social approach provide a much more comprehensive view of what's going on with your agents and workgroups. Once you have all of this unified, you'll never look back. Conclusion If you are investing in peer-to-peer community software and also in social engagement for customer care, you should seriously consider a blended model. Blending community and social together offers economy of scale in staffing, continuity in messaging, and the ability to leverage state-of-the-art rules engines and natural language processing. If you're on the market for these capabilities, look for a truly unified agent and supervisor view, not disparate packages. You'll end up with a much more productive customer care team and as a result - happier customers. Be aware that one company offering community and also social engagement does not mean these two offerings are truly blended. If the benefits articulated here are not truly manifest, then it's not really a blended offering.