SAP

Twitter is the well-known platform based on micro blogging that allows continuous communication among its users.
As it is defined on its site: “people use twitter to create, discover and share ideas with others”.
The communication is friendly, fast generally public (but private conversations are also possible) and free.

People share information and their point of view, provide or receive feedback, build relations and stay in touch with people or just stay connected to monitor and to listen.
Twitter’s number of users is huge and constantly growing. According to the report “New Data on Twitter’s Users and Engagement” from RJMetrics Inc (1)  (Jan 2010), at the end of 2009 Twitter registered users have reached the number of 75 million, with a growth of 6.2 million new users in Dec 2009.
The number and the growth rate are considerable and relevant, but according to the same report, with a potential limit: the percentage of active users is on average at only 20%.

How can we leverage twitter for customer services?
I identify 4 ways that are (at least 3 of them) dynamic in the sense that interaction with users can continuously move from one scenario to the next depending on the specific situation and the stage of the interaction. 

* Use twitter as an official and alternative channel for customer service.
In addition to the phone contact center, email, or also chat through the official web site, you manage twitter as an additional channel for the service contact. 
The company service representatives are officially registered in twitter and provide direct support to the customers. 
The customer goes on twitter and starts the conversation, addressing the support request to the sales representative.
The communication can be public and can be seen by other users, or private, or start as public and become private and vice versa depending on the situation.

* Use Twitter to discover complaints and to address unresolved issues.
You constantly monitor Twitter in order to identify unhappy customers that are just sharing complaints and problems on the net. You contact them for addressing the issues and facilitate the resolution.

* Use Twitter as a public forum. You periodically publish information on twitter (news on products and services, resolution of issues, ways to improve usage, etc).

* Use Twitter as a source of information for intelligence of products and services. You scan in order to find out the opinion of your company and competitor products and services and build up indicators of products and services quality.

However, a couple of aspects need to be considered with attention. Firstly the information via twitter is very concise and the format very rigid, but in that format you are free to write everything in an unstructured way. Secondly the information can flow continuously from users located anywhere in the world, facilitating high volumes. 

These are the key characteristics of twitter, but on one side they limit the quality of the information and whereas on the other side they require a structured organization and information system in order to address the continuous flow.
In other words you need to have a personnel infrastructure, such that each member of the team is dedicated to a specialized area or topic. Furthermore, an IT solution should be available to integrate the process with traditional customer service systems.

SAP has presented (last quarter of 2009) a first solution that supports some of the uses aforementioned: twitter is integrated within the CRM suite and Business Object is leveraged for text analysis.

Basically the solution allows a service representative to navigate from the CRM service module directly in twitter. He can then read and analyze the messages sent directly to his twitter account or search for postings related to his company product and services. Also, he can have a conversation in either a public or private way with the twitter customers and finally, if needed, create a service ticket in the CRM and notify it to the customer via twitter.

In addition, the solution provides functionality for sentiment analysis. An engine analyses the postings in twitter related to the company products and services, classifies them (positive, neutral, negative) and creates a dashboard, which compares the trend of the “sentiment on twitter” over time with the number of service tickets created in the same period of time. Finally, it allows selecting the messages based on the level of sentiment, addressing them in a personalized way.
 
Please don’t hesitate to provide feedback on this topic; I'd love to get your thoughts.

(1) Courtesy from: http://themetricsystem.rjmetrics.com/2010/01/26/new-data-on-twitters-users-and-engagement/ .