I interviewed Connie Bensen of Techrigy about her company’s service SM2 and played around with the tool. When thinking about Techrigy’s SM2 I thought about the SaaS social media monitoring service in four ways.
1) What information was discovered by SM2 and how do you set up a campaign? To use SM2 you search by specific keyword phrases. SM2’s online demo conducted a search on three keyword phrases, and displayed a number indicating the total volume of results for those three keyword phrases. This means any data displayed by SM2 is dependent upon the initial keyword phrases you enter into SM2 to be found. It's important to have specific goals in mind when conducting a search, especially when SM2’s business model is higher fees for more searches.
SM2 gathers information from a number of sources, including: blogs, online bulletin boards and microblogging platforms, for an update list of sources check out Techrigy’s blog post regarding the company’s list of sources.
2) How the data is analyzed? The SM2 reports dashboard enables you to see reports that cut into the total volume numbers, detailing the number of people by gender, or age; where it was possible for SM2 to determine those demographics from all of the posts found by SM2 within the period of time collected.
In addition, SM2 then allows you to categorize the content results by predefined categories. Many of those additional predefined categories related to the type of social media technology where the results were discovered, twitter, blogspot, or livejournal to name a few. SM2 also categorizes by technology, specifically and generically, so you have a category for microblogs and also a category for twitter. SM2 predefined categories also enabled you to search the content found results by sentiment and tone, on sentiment you can find content that is negative, positive or neutral, you could read all of the content posts related to each type of sentiment and also see the aggregate total number. SM2’s sentiment engine analyzes all the results against a lexicon dictionary. The dictionaries are available in 4 languages & can now be adjusted by the user.
SM2 categories included another type of category called related, these included political, religious, stock etc, but also product problem related, these related categories were determined by setting up phrase triggers, a series of related keyword phrases were added to the category folder, and any content that includes those keyword phrases would be put into the folder. For most of the related categories, you are able to see and edit the predefined categories, adding or deleting keyword phrases.
Several categories including tone, did not allow you to see or edit the keyword phrases.
The dictionaries that are used to determine tone, sentiment & emotions can be edited depending on industry specific needs.
You can create new categories by setting up new rules. You can take the data SM2 discovers and conduct deeper analysis by separating any found overall results into different buckets, categorizing the data by sentiment, tone, or related content, such as keywords that describe a competitor. You use keyword phrases to help build a new category folder.
Connie Benson explained the use of category rules this way:
“Keyword searches are used to gather the search results from the SM2 Social Media Warehouse that consists of over 1 billion indexed pieces of information that we're collecting on an ongoing basis & date back to late 2007.
The searches create a databse in the profile. These results can be subdivided into subsets of information using category rules. When setting up a category rule you are providing SM2 with terms or phrases that determine the contents of those buckets of information. A category rule & the category that it maps into are separate objects which allows boolean capabilities and the creation of complex search hierarchies. SM2 also can be trained using Bayesian analysis so that it learns as search results are added to a category. More information here:”
When starting a new project on SM2 in the best of all worlds with unlimited funds you’d want to find all related content that you can find using keywords related to your brand and competitors. & industry related. By having all the data available you can then segment the data with further analysis, comparing results for all of your products against the marketplace and other competitors. However, because SM2’s business model is based on the number of keywords used and volume of searches, you have to think carefully about what information you really need about your brand, marketplace and competitors, otherwise costs escalate.
3) How you can access the data, or how is the data stored? SM2 provides you with a number of existing reports or analysis reports; these include daily volume, share of voice, and compare dates, demographics and domains. SM2 also provides sentiment reports; including brand references, content tone and emotions.
Lastly, 4) how SM2 integrates with CRM systems or gives you the ability to manage a response to social media opportunities found within SM2. When reviewing the actual results of reports you have the ability to assign a single content record into a workflow. Unfortunately I ran out of time in exploring this area, so I will write more about this last aspect in a later post.
Tips for using SM2
Focus in on the share of voice and compare analysis reports, you will probably wish to compare how your product is doing with other products in the market place, these analysis reports with a little bit of manipulation will give you the ability to track how you compare with other competitors and products.
Connie gave five tips for getting the most out of SM2:
1. Decide on business goals - This will guide the scope of the search & timeframe. How broad or narrow is it? Decide on regular reporting & plan to respond to actionable results.
2. Use a spreadsheet to plan the search hierarchy.
3. After setting up the searches use the Author tag cloud & Domains to identify spam.
4. Utilize the category capabilities to organize the search results into subsets.
5. Superusers realize the value of taking subsets of conversations & creating a category. Then that information can be analyzed across the various reports. For example: You'd like to see the popularity for all conversations originating on the West Coast. To do this create a category 'West Coast'. Use the Map Overlay & assign the conversations from the appropriate states in 'West Coast' category. (Search results can be in multiple categories.) Once completed, go to the Demographics report & filter by the West Coast category. Now you'll have the popularity, age & gender displayed for those search results.