OK, so I wrote about at Using Radian6 to predict the Presidential Election - Part 2 at The Analytics Guru and I started seeing how you can use tools like Radian6 if your a Search Marketer or SEO firm, though I don't think the tool was really built with that goal in mind and I created an outcome "spreadsheet" that I embedded into a PowerPoint slide that you can download - see my outcome spreadsheet
I based the spreadsheet on the Radian6 New Topic Widget with the idea I'd get the main phase or keywords for each of the topics above, especially those that were "hot" (red, tan or yellow and larger). But as I went on to other keywords, I found the a sentence or two necessary, and I increasingly found myself wanting to use Radian6 as a news summarizer to News Intelligence tool, more than a Keyword tool.
And yet, based on the configuration of the New Topic Widget (below)
there should be plenty of keyword phrases that you could pull - if not so much that people will search on those phrases, they'll still be looking at the content - which means that having the phases and ideas distilled into a phases, are valuable for contextual search - especially when you contrast my spreadsheet with what typical SEO/SEM Keyword Research tools give you like Google AdWords Keyword Planning tool:
Typical Keyword tools don't really deliver conversations - they look more at descriptive term - but that's not really telling you much about sentiment or influentials - or even, where to advertise - hence, their weakness or deficiency in this area.
If you take Column E of my spreadsheet (see my outcome spreadsheet) which is embedded in the PPT slide you can download from the link, above, you'll see what Google makes of it:
It's awful - all the good stuff - the entire conversation - was left out - showing the immaturity of standard keywords tools for Social Media analysis - or to hone in to conversations and find influentials.
So what would I like Radian6 to do?
Answer: Use the Radian6 New Topic Widget to extract a list of keyword phases out of the contextual text it picks up, sorting by it's engagement algorithm and making that a down loadable spreadsheet both a marketer and a copywriter can use.
Why a copywriter?
Because, on the face of it, much of what I picked up, especially as I went on with this exercise, are the questions both candidates need to answer -as I outlined what I found over at The Analytics Guru in Using Radian6 to predict the Presidential Election - Part 2 with these issues that are bound to come up and needed to be addressed, and that, to some extent, are being addressed by both campaigns.
And I have argued these tools, things like Radian6, ought to be in the Web Analytics groups, not just Marketing or Communications - as the involve sentiment analysis - which affects web measurement and the weight one puts on traffic, in other words, it's meaning - which is the domain of Web Analytics.
For example, I see a need for semantic analysis of the content around the keyword phases - with an eye towards "who" is saying what - then crawl back and build an opinion database that will show the likelihood, based on who is saying, what, the sentiment of the the text.
Rank the text by Radian6 Influence algorithm and give us an overall sentiment for each set of terms in a profile - and also give me the spreadsheet of terms, along with who said them - which can be used both for Keyword Planning AND, Intelligence.
In fact, you'll find that as you move up the Web 2.0, Web 3.0, whatever you want to call it, continuum, many of the tools being created have multiple purposes - perhaps they were created for one thing, but are adapted to others, like Twitter - look at what people are doing with it … far more than the founders ever intended or imagined.
And the same for tools like Radian6, Comscore AdMetrix and ComScore Segment Metrix - all three tools are being promoted as marketing tools - or PR Tools, but they are also useful for other things - perhaps even more useful, give the right hands, than what they were created for, initially.
I'll have more to say about this in a later post as I'm not done here yet.



