I have never been a fan of  Forrester Research – but  they produce a lot of good research. By the way, have you noticed a lot of the research firms have been acquired  or are merging of late.   So, it wasn’t entirely surprising to me Forrester Research would tell it’s own analysts they can’t have blogs of their own any longer.  As reported in  SageCircle

Forrester CEO George Colony is well aware of that savvy analysts can build their personal brands via their positions as Forrester analysts amplified by social media (see the post on “Altimeter Envy”). As a consequence, a Forrester policy that tries to restrict analysts’ personally-branded research blogs works to reduce the possibility that the analysts will build a valuable personal brand leading to their departure. In addition, forcing analysts to only blog on Forrester-branded blogs concentrates intellectual property onto Forrester properties increasing the value of the Forrester brand.

My understanding is the average analyst gets burnt out with all the travel, writing (writing several research papers each quarter) and consulting ,within 3 years.

Forrester Analysts have a tough life and maybe, one of the reasons why it’s worth putting up with less than optimal pay and frequent travel is to build your own brand – in fact, that’s probably the main reason why people sign up to become a Forrester Analyst in the first place.   Well, no more of that.

All things being equal, if your a Forrester analyst and you can’t build your own brand – what’s the point of being an analyst for them?

I don’t care about Forrester analysts that much – but what if it’s not just about Forrester – what if more and more employers decide they want to own all the intellectual IP while you work for them, including your personal blog?   Is that in our future?

Reading more closely, it turns out Forrester Analysts can continue to blog outside of Forrester if it is on a subject not related to their role at Forrester – in other words, you can blog about music, travel, art, cooking, even politics, but you can’t say anything on a subject Forrester is paying the analyst for.

“…..Make no mistake: Forrester is committed to social media, and the number of our analyst bloggers is increasing, not decreasing. Analysts will still have the ability to blog outside of Forrester on topics not related to their coverage areas.“

What, if in 5 years, most employers decide to feel this way ?

Anyway, decided to try to make sense of this action by exploring Social Media and using Sysomos MAP (here’s the query - Forrester AND Research AND blog AND analyst).

Here’s the popular phrases around the decision of Forrester to ban blogging in most cases

Here’s the Key Conversations that happened in the last 3 days around this subject in blogs:

Forrester is investing in building social tools and associated best-practice training for our analysts so that more of them get involved.

In addition, forcing analysts to only blog on Forrester-branded blogs concentrates intellectual property onto Forrester properties increasing the value of the Forrester brand.

Sagecircle shared rumors that a change to Forrester blogging policies would prevent analysts from having personal blogs and would aggregate analysts’ posts into Forrester-branded role-based blogs..

The research we write for clients has always depended on a rich two-way conversation with experts and practitioners in the marketplace.

Blogging is an extension of the other work we do — doing research, writing reports, working with clients, and giving speeches, for example..

Blogging is an extension of the other work we do — doing research, writing reports, working with clients, and giving speeches, for example.

There is only one augie, and the thoughts i share on my blog are now based Upon the research i do, the people i meet, and the information i am given access to thanks to my role at Forrester.

So it was interesting to read of Forrester research’s directive to analysts using their own personally branded research blogs: take them down or re-direct them to the Forrester site.

If Forrester is worried Analysts will build their own brands and then leave – instead of punishing Analysts, I suggest they make it more enticing for them to remain.  Here’s some ideas – bet I’m not the first to think of them, either.

1. For any paper that an Analyst writes for Forrester, let the analyst get 10% of the sales money added to their salary every time they’re paid.
2. Provide free education to expensive college and graduate courses if you remain an employee.
3. Cut down on billable travel by 25% and replace it with virtual meetings at a similar billing rate.
4. Allow Analysts to use 10% of their own time for a project of their choosing (and let them get 20% of any sales coming from that work).

That is some of the ideas I had today -let’s think of rewarding employees, not punishing them, for taking initiative.

Social Media Week is over as of Friday night but I didn’t get a chance to write a post about Influencer Identification based on the tools I have access to.  As in the other posts in this series, I will place a datafile you can download at the end of this post so you can examine the information for yourself if you want to.

And, I did post about Finding Influencers – based on the idea that being influential is function of

…. being connected to others who have short chains to many other people with high betweenness. Or, looked at differently, betweenness is a measure of how many social circles, or social scenes, a person is connected to.

What kind of influencer list can we get from the tools I have access to?

Radian6 has an influencer widget that provides a list of sources that are considered influential for the keywords provided in a Topic Profile (or a collection of Topic Profiles).

However the Influencer ranking is somewhat dependent on how sliders are set up in the Influencer EQ dashboard in the Topic Profile configuration (see below).

Rather than explain what each of these slider bars means – but on the face of it the influencer list provided by Radian6 in this case, looks good.    You can also drill down to the social profile of the blog/website owner (if Radian6 has information) (see below) and get various stats like website traffic on the far right (if you purchase the Compete.com site data).

The Social Profile information is a best guess, I’ve found, it’s often sketchy and inaccurate  – but it’s better than nothing.  On the other hand, you’ll almost always have to go to the website and get the contact information anyway – so having some information from Radian6 doesn’t hurt.  Unfortunately, the Social Profile isn’t exportable, even though the Influencer list, itself, is – a feature I wish Radian6 would add.

I’ve tried to get a blogger list out of the Influencer widget, but found of the 250 sites provided – I had to throw out about 95% of them – upon close inspection.   The list has often been way too noisy to be useful in building lists and I concluded the Influencer Widget, despite the name, was never intended to be a blogger relations list – and had Radian6 wanted to implement such a feature, it could have – the Influencer Widget is more like a Prizm that reflects all the influential sources (or light) that it found – and ranks them by Influence (which you define, partly, by the Influencer EQ settings).

According to this Influencer list SocialMedia.biz, Mashable and SocialMediaWeek.org were the most influential sites and while there are many good measures here, defining betweenness (Influence) is a measure of how many social circles, or social scenes, a person is connected to is  missing – because

1. The Influencer Widget measures the influence of websites, not the individuals who write them – which is the information most people actually wanted.   Also, you need to drill down to the actual posts that are influential to find out why that site is considered influential, and the author of the the posts in question would be considered the “influencer”.
2. The information about “who” wrote the post that is considered influential is buried deep in the Influencer Widget, making pulling out any kind of useful list of Influencers very time consuming – and the vast majority of sites usually have to be thrown out – in my experience of using this widget.

If you have a good clean topic profile, expect to spend 4-10 hours  of hard work to get about 10 – 20 names of influencers out of it.

Alterian/Techrigy/SM2 does have an Influencer Report of sorts and it does identify individuals rather than websites, and is immediately more useful, in this sense, than Radian6’s Influencer Widget – however, Radian6 was never designed to create a influencer list – so i see that just as an oversight that could easily be corrected in a future release.

According to Alterian/Techrigy, I’m considered the 6th most Influential “author” – and I intend to use the Techrigy Top Authors report more often.   However, a closer look at this report shows it’s flaw, Influence is a function of how many posts an author has done, which is a misleading metric, in my opinion.   Never the less, you can get the top authors my media type and by sentiment.

For example, if you wanted to get the top authors by negative sentiment you can do that with Techrigy (see below)

But when you drill down (as I did with Mike Moran, who wrote 1 tweet that Techrigy said was “negative” – it turned out to be, not negative at all).

So …. Techrigy’s problem, to me,  it’s good features are riddled with the same problems all the vendors are having -  the technology is immature and there is a lack of standards that are not yet in place for vendors around influence and sentiment, etc.  But Techrigy’s Top Authors report could be useful as an Influencer list for a subject if the noise is filtered out of a profile and your willing to accept volume (posts or tweets) as the measure of influence.

Sysomos Map Top Influencers (based on authority + recent posts) looks right on to me – and the first blog on the list is the same as for Radian6, SocialMedia.biz – but the rest are mostly different.   Being familiar with many of the blogs and individuals who write them I think Sysomos’s Influencer lists are right on.

Authority (which I equate with Influence, according to Sysomos) is determined by the following features, depending on the type of website being looked at:

Blogs

• The number of unique inlinks to the blog over the last year
• The number of bookmarks at social sites such as Delicious
• Readership information if publicly available
• Posting frequency
• Number of followup comments

Twitter

• Followers and Following data
• Number of tweets
• Number of re-tweets

Forums

• Reach
• Inlink count
• Posting frequency

Traditional Media

• Inlink count
• Reach (if available)

Clear enough – and not bad, since Sysomos does a very good job with noise supression – but it’s also not a measure of Influence that I defined at the beginning of this post as a measure of  a measure of how many social circles, or social scenes, a person is connected to.

First of all, like Radian6, sites are the unit of measurement, not individuals.  Also, there is no real attempt at Social Mapping, much as Google is doing with Social Search (see the Social Profile, below).

My sense it that “influence” could develop the kind of “mapping” that Google has just released the first version of, but hasn’t yet figured out if and why it should.

Perhaps Google’s entry into the Social Monitoring Space(read plus read my long article about Google and Social Media Monitoring (Read more: http://www.webmetricsguru.com/#ixzz0evICP23S ) will galvanize vendors in this space to try harder to reinvent themselves -  Sysomos should follow Google’s lead and and start building social profiles with it’s vast collection of data.

Brandwatch has a Top Sites report that is the closest thing I can find to an Influencer List – esp if the sites are filtered by high credibility.

Honestly, this list isn’t bad and most of the sites make sense – without too much trouble you can get the names of the authors and their contact information.   The problem here is this report is just a proxy for Influence, and it’s really designed to be a list you can use to find influentials connected to a profile – also, there is no social mapping, as I described above.

I would not use BrandWatch to find Influencers as it clearly wasn’t designed for it – neither was Techrigy – but… at least Techrigy does have a top authors report, which does do influence based on number of posts where as Brandwatch doesn’t appear to do influential lists, at all.

Biz360 does have a Top Authors report/widget as well, but not a influencer report – but the Top Author report is decent, as far as it goes (see below).

… and I can’t argue with many of the names on the list above, except 2 – John Q. Public and Admin.

In fact, by clicking on any author you get a drill down of all their posts for a specific time period:

Sentiment is also broken down by author

But it’s not clear if content is being rated by the number of postings or some other factor – and mapping individuals is missing from this package.

But one other approach to finding influencers is using PostRank Analytics – a free report can be generated and if the subject is defined (on their lists) you can get good stuff – but if most of the subjects I’m interested in knowing about – there is no category for.  Still, the list for Web Analytics was / is impressive (see below).

Social Media Week NYC did not exist as a list of blogs I could get, so I tried Web Analytics, and found my blog up near the top.   But for those subjects that do exist, PostRank provides the “wisdom of the crowds” with a change on engagement with a post by the audience.

Still, Postrank and the other platforms examined lacked the ability to map interconnections – and therefore, are immature.

Klout.com also has an influencer list – or Topic List which is very useful.    However, many words aren’t in any list – for example, Klout knew nothing about Social Media Week but does know a lot about Web Analytics.

I am not sure how Klout calculates the Influencers.

Also wrote about Traackr last month and it does a good job at identifying  influencers.

But Stowe Boyd is right – none of these tools maps Influencers as well as it could.

By the way, here’s the Influencers for Social Media Week NYC

Since it’s SuperBowl XLIV today I worked that into my “editorial schedule” (chuckle) and decided to do a post focusing on what chatter I could pick up from other NFL Players who might be blogging and tweeting and offering their own predictions (or just expressing their sentiment).

However, Radian6 teamed up with Mullen to do some of it’s own predicting – and real time following

Mullen and Radian6 today announced BrandBowl2010 (http://brandbowl2010.com/), a Twitter/Super Bowl experience that combines tweeting, ad reviews and a host of metrics that will let viewers see close to real time ratings of the TV commercials running on the big game.

“This is a perfect opportunity for us to showcase our capabilities and functionality when it comes to measuring online conversation and brand sentiment.”

Users can log on to BrandBowl2010 using their Twitter ID, tweet directly from the page, view tweets using the hashtag #brandbowl and see how the Twittisphere is responding and rating Super Bowl commercials and the brands running them.

The site will provide an overall ranking of the brands advertising on the game based on a composite score that takes into consideration both volume of tweets and sentiment (positive or negative). The Top 10 most popular brands will be featured on the site, with second tier brands displayed in a sub-section called the Locker Room. In addition, a simple click will display the volume of tweets over the duration of the game, the sentiment breakdown along with a word cloud that highlights what people are saying about a commercial.

So, along with my analysis – you can check out what Radian6 is doing, directly.

Note: by the time I got to the end of this post – I was unable to find any players that were chiming in on predictions – I’m sure some have – but I didn’t have a good list to work from.  However, the data file is contained in a link at the bottom of this post – please share your insights here if you have come up with something with this data.

Since I don’t follow football and am not really a sports fan – I don’t have subject area knowledge of the details that most fans have, so I’ll just use the tools and let them speak to/for me.   What I share here is me – my ability to dream up and use platforms and tools – I don’t care if I share the recipes, and someone else does something even better with it, as  I can always dream up more.  So if you see something in the data I uncover, run with it (all the way to the finish line, if you want).

Having said that – used three tools – Radian6 to pull the last 3 days of data ( a lot of data – attached at the end of this post for your own analysis, if you wish – I limited myself to the latest 10,000 records – there are volumes of data just for the last few days)  from the River of News, TweepSearch – to find NFL players that aren’t on the Colts or Saints team – and Sysomos Map, to get a better sense of the keywords to enter into Radian6 – using it’s “popular phrases” report (I wrote about this a few days ago).

Many of the profiles haven’t been updated recently – still, I’ll use the twitter accounts to search out if there have been comments about the game taking place today – and this post is written several hours before the kickoff.

The Popular Phrases uncovered by Sysomos (for Twitter, shown below) are a perfect example of why we need multiple platforms and tools – Sysomos MAP as a capability to uncover popular patterns of words that I have not seen anywhere else).  Using these platforms together is a superior setup and I believe agencies currently need to have access to multiple tools – this is one example of why.

The popular phases picked up by Sysomos were pumped into the Radian6 Topic profile  to improve it’s accuracy, while TweepSearch was used to identify NFL players who have an online presence and qualify them by influence (crude – in this case – just followers).

If you download the data I provided you’ll be able to filter on media type

In fact, using Radian6 and looking at the last 3 days (again, if I downloaded all the data – I’d have a massive file that no one would want to download and would take me way too long to analyze).

Half  of the activity in Social Media round Superbowl XLIV is on Twitter – not surprising – but simply, amazing!

And if we take the combined Twitter followers of those who are tweeting in the last 3 days, there is a potential to reach 206 million accounts (though there is currently no de-duplication in Radian6 – these are not “unique accounts” – consider this more like an “impression” or “pageview”).

Potentially, there’s 206 million messages that could be streamed just by Twitter along, in the last 3 days, around the Superbowl – but if even 1 or 2% actually saw the messaging – that would still equal a few million message – yes, Twitter is Mass Media, more than any other Social Media, right now.

So, let’s get down to the meat at the table, so to speak – who is saying what?

First, I could not find any player that tweeted about the game in the last 3 days based on the TweepSearch list – that tells me two things.

1. I need a better list – tweepsearch isn’t updated often enough for this kind of event (need a list of players who have updated in the last day or two on Twitter).
2. Most NFL players might have been warned not to tweet about their own predictions before the game.

So, I did the next best thing and took the tweets of anyone who has a million followers or more – about the game – here it is.

And here’s the tweets of Twitter accounts that have between 100,00 and 1 million followers.

what I’m suspecting it the “interesting stuff” is not coming from the accounts with the largest number of followers.

How about predictions on Twitter (see above) plus one for Comscore…hmmm

Well … Comscore’s survey does predict the winner – most thought the Colts would win.

When asked which team they expected to win the Super Bowl, the Colts were selected by 54 percent of respondents compared to 34 percent who thought the Saints would win (with 11 percent undecided). However, when asked which of the teams they would prefer to see win, the results were reversed with 52 percent selecting the Saints and 33 percent selecting the Colts (with 13 percent undecided).

Respondents were also asked to choose the most likely candidate for Super Bowl MVP. Not surprisingly, Colts quarterback and league MVP Peyton Manning was the overall top choice, garnering 50 percent of the vote. Saints quarterback Drew Brees ranked a distant second with 15 percent of the vote, followed by Saints running back Reggie Bushat 8 percent.

The upshot is – I think Social Media tools could be very helpful in handicapping sports events – but I need better information than TweepSearch can provide for a sports event – if we had that – I think I could have found someone who is NFL player now – who tweeted who they thought would win.

And I just want to say – I got this idea for the approach while at dinner last night with a group who attended Eventcamp here in NYC.

Here’s the data – SuperBowl last 3 days River of News Radian6 -  you can analyze it if you really want to – but bear in mind the file is almost 7MB in size, so don’t bother unless you think you can add value.   Radian6 provides the most complete data feed of any vendor – if I had unlimited bandwidth and resources – plus the time, I could have done something very nice with this data.

Some of my readers have asked me to weigh in on Stowe Boyd’s post on It’s Betweenness That Matters, Not Your Eigenvalue: The Dark Matter Of Influence as it impacts how influence in a social network is determined; Stowe Boyd’s post is also a useful framework to talk about how good Social Media Monitoring platforms do with determining Influencers.

… influence doesn’t seem directly linked to how many people you are connected to. It’s a function of being connected to others who have short chains to many other people with high betweenness. Or, looked at differently, betweenness is a measure of how many social circles, or social scenes, a person is connected to.

….do you have short paths into other social scenes, both incoming and outgoing? That is the deep structure of being truly connected: bridging over different social scenes, acting as a conduit, a vector, a filter and amplifier for ideas good and bad, the best insights, and deadly viruses.

Stowe Boyd is right, there is no Social Monitoring tool is capable of this level of analysis today, but  it’s fair to say that Google is working towards it, and , probably, no one else is  as close to it as Google is or will be (see John Battelle’s post on  Google Rolling Out Social Search: But Does It Leverage Facebook? last week.

Hmmm….. looking suspiciously a lot like what Stowe Boyd was talking about …  he thinks Google – so sure – this is probably something Google is working on … who is more influential ….. and here’s the paths – now all they have to do is score them and figure out what “communities”, an how many.     Anyone else doing this?  Raise your hands.

I planned to compare several of the monitoring platforms on how they locate influentials – but that will be done later this weekend, and I’ll use this post as a jumping off point.

By the way, keep in mind that Google working with the NSA on the China Security Breach (and maybe other things – see video above) plus read my long article about Google and Social Media Monitoring and then think about what Stowe Boyd is saying and Google Social Search.     Hmmm ….