In my last blog I gave you five reasons to use Social Media Analysis. This blog provides some technical depth that can be used for analyzing Social Media Analysis solutions.
SMA technology offerings range from small companies that monitor individual, named social media accounts using keyword search to offerings with extensive natural language processing (NLP) that process millions of social media sites including the various "firehose" APIs (e.g., Yahoo and Twitter streaming APIs) and provide historical data for trending. Choosing an offering with the right capabilities depends on an organization's objective for mining social media dat ... see my previous blog for some example reasons. There are several SMA decision criteria that have varying levels of innovation. My top five areas of innovations to include in SMA technology decision criteria are: Aggregate Monitoring, Global Technology, Text Analytics, Social Model, and BI Integration.
1.Aggregate Monitoring
Most of the initial offerings in Social Media Monitoring (SMM) technology were targeted monitoring tools. The user entered in the accounts they wanted to target, and the tool monitored those accounts. Simple enough! In recent years, an innovative trend for vendors is to provide aggregate SMM tools -- that is, they monitor millions of social media sites and provide an aggregate of all data for all users. Most of these tools provide filters that can also target individual accounts. Have your cake, and eat it too. The aggregate SMM tools are normally more expensive, so the objective for SMA is imperative when making this choice. A small company may be fine with using a targeted SMM tool that can target 100 customers for $40 a month (or even use a free open source tool for this), while a large organization may need a more expensive aggregate monitor to collect what customers are saying about a brand.
2.Global Technology
Many SMA tools only support English across Twitter, Facebook, and a few other sites. Global organizations need a tool that can process multiple languages and support leading regional sites. There are a handful of innovative tools that have incorporated multilingual processing tools into their SMA offering, allowing an organization to pick one tool that can monitor multiple languages. Si, señora. For global organizations, it is also important to choose a tool that can mine data from social media sites that are regionally preferred. For example, brand popularity in growing BRIC nation economies may be important to a global organization. In China, Qzone is the largest social media site, so monitoring Facebook for analysis of public opinion in China may prove to be an unproductive endeavor. Facebook recently (last year) surpassed Orkut in social media prominence in Brazil (and India in 2010), but Orkut remains a site to monitor for public opinion there. There are similar examples in other regions of the world. Understanding the region and choosing a tool that has innovated to cover specific regional sites is imperative.
3.Text Analytics
Whereas the initial SMA offerings provided keyword and keyword list search, recent innovative products are integrating natural language processing (NLP) and/or other text analytics that provide a much deeper analysis of the social media content. The text analytics market segment is another growing market, but it has a richer history than SMA. Some of the vendors in this segment date to the early 1970s -- explains the Beatles' lyrics used as test cases. NLP has been used in corporations for several years to mine unstructured data and make it accessible to existing analysis tools. Some SMA vendors have recognized the value of NLP and other text analytics and have innovated their products to use NLP-related capabilities like syntactic and semantic parsing, named-entity recognition (NER), relationship analysis, entity disambiguation, summarization, and sentiment analysis to name a few. These capabilities provide an increased confidence over traditional keyword and keyword list search in both matching relevant social media posts as well as the accuracy for classifying the information in those posts.
4.Social Model
Most organizations have some form of a model to represent their concept of a brand -- often depicted as Venn diagrams, concentric circles with pie slices for the brand facets, relationship graphs, or some other chart that has taken the marketing staff entirely too much time to produce. Many companies embracing social media have added social model concepts to their brand model, such as the social view of a brand by customers, which might be different than the social view of a brand by investors. Some of the key views included in social models are customers, investors, competitors, partners, analysts, media, employees, and general public (not included elsewhere). Some SMA tools are recognizing that social model is becoming key for the overall brand model, and are leveraging social models as an anchor for posting events being monitored. The model then acts as a way to view the various events filtered by the SMA tool in an organization's model of their brand. This innovative area will continue with more tools providing standards-based representations of social models, such as the standards prevalent in the Web 3.0 community.
5.BI Integration
Business Intelligence (BI) is a high-growth, multi-billion dollar market segment that has been deeply rooted in many organizations for years. SMA is actually one form of BI. However, only a handful of SMA tools (typically the providers of BI tools) integrate with existing corporate BI tools. Many SMA tools are creating silos of information containing the immediately-available social aspects of a brand, while BI tools are working with corporate data that represents traditionally mined aspects of a brand. Blending these two worlds together might create a perfect Brady Bunch effect, possibly allowing companies to correlate changes in the social model of a brand with changes in production, logistics, or customer support. SMA tools that innovate to provide integration into existing BI tools will be the leaders of the SMA market in the future. The SMA market is already an acquisition target, and I anticipate BI vendors will continue to acquire SMA tools this year and next, specifically the ones that are innovative.
The Social Media Analysis (SMA) market is undergoing several innovations, including my top five described above. Organizations looking to leverage SMA tools to better understand their social model should consider including these innovation areas in their decision criteria. As always, many of these decision criteria are based on the primary one -- the organization's purpose for analyzing the social media data.