At the IBM Connect conference, we heard about some of IBM’s approach to Social Analytics. In a nutshell, you can break the strategy into two main areas:
- External Customer Analytics
- Employee Analytics
In both areas, the overall approach involves social listening, data collection, segmentation and deep analytics. What we heard from IBM was a focus on Campaign Management and how social analytics can be an integral part of your campaigns. The idea is to be able to target individual people or very granular segments with your marketing.
From a customer analytics viewpoint, here are the basic steps to managing a campaign with social analytics involvement:
- Derive ideas insights and actions from social media. This can be called sentiment analysis and you want to understand what consumers are saying about your company and products. Just as important, you want to understand how people feel about you.
- Pull consumers into the conversation on the web. Once you analyze the customer social data, you need to target communications to those areas of the social network where you can take advantage of one-to-one marketing. In an example, if you find that many people see you favorably on a particular forum, you could suggest on that forum that people sign-up with you (opt-in) for more information.
- Execute the campaign using individual data for consumers who opt-in. This is where you target those people with your individual ads, special offers, community involvement, etc.
One very good question is how to mine all the terabytes of social data. By using IBM’s products you can take advantage of many data sources that IBM already has access to and then you can add your company’s data if needed. The IBM products used on the consumer side include the following:
- IBM Social Media Analytics
- SPSS Modeler
- Cognos 10
- Connections 4
- Core metrics
- Unica
What if you want to analyze employee behavior instead of consumer? In this case, you want to get the social pulse for enterprise. Similar to the consumer side, this analysis includes real time data collection, sentiment analysis, and topic extraction of unstructured social media data of the enterprise.
Here is the lifecycle for the internal side:
- Derive measurable aggregate insights for employee sentiment from social media. This would include both the internal social network and external social media. To ensure that employees privacy is maintained, it is important to make sure that data is collected in a way that employees agree with. So allow employees to opt-in or out and allow them to designate which external accounts can be accessed.
- Model and micro segment employee groups to refine policies, focus communications and drive culture. The analysis will tell you where employees are having issues with the company, with other employees, or with customers. This information should then be used to improve policies, processes, communications, etc.
Here are some of the techniques used in the employee analytics space
- dynamic recommendations
- community metrics
- social influence
- social network building
- sentiment analysis
On the Employee side, IBM is still in the lab-development process of testing approaches and products on the to perform sentiment analysis through IBM Connections. However these products will likely be involved in the process are involved in this process:
- SPSS Modeler
- Cognos 10
- Connections 4
