Wish List for Social Media Analysis

| 11 Comments

What do you want that you can't have today? Specifically, what would you like to see in a social media analysis tool or service that you can't get (as far as you know)? Most of the vendors are probably reading this blog, so here's a chance to air your wish list while I get out of the way. I'll start with one that Lee Odden mentioned on Twitter yesterday.

SMA for SEM
Lee wants an automated keyword analysis tool for social media sites (process described by Marty Weintraub):

Really need to the social media monitoring industry to step up and come up with a good keyword analysis tool specific to social media sites
@leeodden
Lee heard from three companies in response (ok, the rest of you need to monitor social media keywords on Twitter). Is anyone else working on that application or want to talk about it (publicly or privately)?

What else do you need in 2009?
Open mic time. What's on your wish list? Mention it in the comments, and I can guarantee that companies in a position to offer it will see your suggestions. You might even hear from someone who already does it.


11 Comments

Nathan, by keyword analysis tool, are you referring to what Lexicon does with Facebook? I may be misundestanding Lee's request, but we have been providing real-time trending anaylsis tools and capabilities on keywords/keyphrases for quite some time, against variables like source type, geography, sentiment, and posting history (mo/yr) just to name a few.

Joseph, yes that is the sort of thing I am asking about, but across numerous social media sites - not just Facebook.

Thanks!

Lee, there are a number of ways we accomplish what you are requesting. The graphing/analysis tools we provide are in real-time and aren't limited to social media sites alone (Twitter, Facebook, MySpace), but across all source types (blogs, forums, consumer/review sites, video, images) and these can be cross-referenced against specific metrics like tonality, geography, post history, just to name a few.

I'd be happy to demonstrate what I'm describing here - feel free to tap me @RepuMetrix when you've get a moment.

How about getting a consistent definition for the term "social media"?

I'd like a spam filter that traps off-topic posts, not just casinos & viagra (like Akismet).

Niche sites like sphinn.com would be able to easily reject well-written but off-topic posts about holidays to India, or electric cars. It could also help with ensuring that posts are submitted to the right category of larger sites, like digg.com.

Anti-spammers need to raise their game to deal with social media, you know that spammers are already working on it.

Having seen most of the tools (about fifteen or so, including Radian6, TruView, Trackur, etc)

I would like several things. I could write on any of these points for a page or two, but here is the bullet list.

1- duplicate post origin. Too many professionals post duplicates to create buzz effect and distributed readership.

2- Better tie-in to keyword tools, perhaps Google API to bring in more clusters of like-minded words that may get left out.

3- User-mapping across networks: including Facebook Connect, Google Connect, OpenID, and possibly Linked-in / Zoominfo API.

4- ROI metric / conversion support. Conversation interaction numbers correlated to marketing and PR spends, establishing short URL items for tracking purposes.

5- Audience and conversation locating based on demographics, perhaps tying in to metrics from a service like Quantcast to find like-minded individuals.

6- More friendly GUI. For the most part, all of the current GUI's stink. You have to be a social media guru just to understand everything on the control panel.

Trendrr may do the trick or at least structure a good bit of it.

http://trendrr.com

Good API as well for those who want to build off that data

Trendrr

Great set of questions and feedback. At Visible Technologies, we use TruCast to work many of the issues above directly using multiple methods - keyword and keyphrase categorization, learning algorithms, and conceptual categorization which matches like subject regardless of keywords by looking at the relationships between words and phrases across multiple posts/comments/twitters, etc. These aren't all applied in the basic use of the software that is sometimes demo'd, but we'd be happy to show you more and discuss. The process as Marty describes it is similar to the work we do with our Benchmark reporting for many companies like Microsoft.
Barry makes some great points as well and all are areas we're continuing to improve. On the duplicates for instance, our influence measurements are continually tuned to work through that exact issue. Otherwise, we'd be assigning levels of influence and sentiment to authors and sites beyond their scope. Keep the feedback coming. We take it seriously.

Keep in mind that, even if Google won't let you scrape and bake behind walled gardens with the keyword tool, you can scrape to create your OWN HTML pages and run the tool on that.

Cool thread, thanks... :)

Barry - HiveSight does user mapping across networks, as well as demographics. We've also been told that our GUI is nice, but that's for you go judge.
Our approach is different from most tools - we focus on people rather than conversation, and try to support strategic marketing decisions rather than just PR.

Thanks for the comments, everyone. I don't want to give the impression that I've left the room; I'm just staying out of the way.

Keep in mind, if you have an immediate need for a function that you haven't found, I may be able to help directly.

Comments are now closed for this entry.


About Nathan Gilliatt

  • ng.jpg
  • Voracious learner and explorer. Analyst tracking technologies and markets in intelligence, analytics and social media. Advisor to buyers, sellers and investors. Writing my next book.
  • Principal, Social Target
  • Profile
  • Highlights from the archive

Subscribe

Monthly Archives