A tip from Charlie Osmand over at Fresh Networks (who I believe spoke in London at #msm10 last week, where I was at and also spoke) caught my attention this morning and I took a look at their (his) review of Social Media Influencers.   The link to download the PDF  is contained in Charlie’s DM  below.

curious to hear what you’ll make of this one: 9 Social media monitoring tools compared for ability to find influencers: http://t.co/pANb2h

Direct message sent by charlie osmond (@cosmond) to you (@webmetricsguru) on Dec 04, 10:10 AM.

The report is excellent in that it goes into some depth of influencer identification because it confines itself to actually focusing on a particular aspect/application of each of the 9 platforms (I was not familiar with PeerIndex, but have worked on all the other platforms listed and have active accounts on most of them).

But the closer you get to the truth, the more you realize there is no “one truth” or infact, one tool good for everything.  According the FreshNetworks paper  BrandWatch is best for finding Forum and Message Board Influeners while Radian6 was better finding Twitter followers (I’d put my money any day on finding Forum followers over Twitter followers as there are so many free tools like Klout and PeerIndex that handle Twitter followers – but identifying  Forum Influencers by topic – that’s rare and potentially much more valuable).  Scout Labs/Lithium was best at finding YouTube and Image influence while Synthesio was better than the rest at finding Facebook Influencers (again, much more valuable because it’s harder to get into Facebook and really find a consolidated influence score there).

The rest of the platforms really stood out for other features but not  for influence identification such as Alterian/Techrigy/SM2 for segmentation, Sysomos for best overall tool (USP) and Social Radar for the best data visualization platform.    The subject of the influence study was “organic baby food” – something every platform should be able to do a good report on.

There is so much “fragmentation” between vendors and what outputs they produce and buyers (and what inputs they need) that each platform tends to be good at certain things but not as good at others – you pick the platform you need for the relationships and capabilities you need on one project, then find it’s unsuitable for others, and you get into the datasoup mess most agencies are in, with a hybrid set of tools that are not interoperable and do not support the workflow or analyst, never mind the stakeholders.

But still, this is the best overall influencer study of a comparative set of social listening platforms I’ve seen so far and does a good job at defining influence in a workable way that is useful for most people reading this blog.

FreshNetworks quotes Dave Sifry, founder of Technorati as …

targeting the Magic Middle of online influencers, sandwiched between Chris Anderson‟s Long Tail and well-known A-list influencers.

Makes sense to me as it’s unlikely your going to get Lady Gaga to advocate much of anything (if you can even get through to her handlers) but you may have some luck getting through to people just ahead of the Long Tail (who have groups of niche followers and aren’t so inundated with requests that they are putting up barriers to being contacted).

It’s also interesting the FreshNetworks study talks about 4 types of Influencers yet only shows the three (above), where is the forth type hiding?

Another great thing about the FreshNetworks paper is how of it defines 5 levels Engagement – I don’t know if this is the best definition, but it’s certainly workable and even possible to manually rank influence based on it.

Inactive – aware of the relevant sector or topic but not actively engaging.

Brand Conscious – aware of your brand. Generates some buzz but not fully engaged and mostly fleeting references online.

Word-of-Mouth Amplifiers– spreads key messages and updates that relate directly to your brand, products or services.

Brand Advocates/Evangelists – support and champion your brand online.

Brand Ambassadors – key advocates who have a strong involvement with the your brand online. Could be a commercial relationship.

I doubt you could automate the classification of influence (but it’s worth a try).

There’s also a section on “what’s in it for me” in approaching influencers,  how to do it and how to select the right tool/platform.   I honestly could not have done a better job if I had written the report myself, though I might have tried to pick clear winners and losers instead of giving every platform kudos for something they do better than anyone else, as FreshNetworks did.

On the hand, to be fair and honest, as I am known to be – the real issue here isn’t influencer identification capabilities, it’s rather, the right tool/platform you need for your particular situation – that’s a custom consulting case scenario, not something that a paper like this can fully address.

My book will attempt to set the case and rationale for more strategic investment in tools and processes than any of these types of studies has produced (the basis for the FreshNetworks study is to build more clientele – nothing wrong with that ) but doesn’t really address that fact, nor can it, that every business and situation is different and the precision needed in influencer identification continues to evolve past where any of these tools, profiled, goes.

Also notable, is that this study covers 9 platforms and leaves out others that could have also been interesting such as Traakr (not to mention that just about every PR, Advertising and Marketing firm now is either creating influencer lists of their own or buying them and white labeling it from firms like Traackr or other similar type services).

Prices vary for such lists and are widely dissimilar, often for the similar lists to tell the truth, but having done some of this myself, in various capacities, I am of the belief we should not invent the wheel  unless we can show a much better cut of influence than what the existing tools and platforms produce now, nuff said.

This is going to be a long post and the ideas have been peculating in my mind and everything I’m saying now came me as a result of drafting my Monitoring Bootcamp presentations in  London late this month and taking a long walk (like J.J. Rousseau, many of my ideas come to be while I’m walking – sometimes, I have all I can do to jot the ideas down before I lose them).

Measuring Influence Online

Influence is a very deep subject and I’m sure I am missing parts of the puzzle – I’m hoping  readers weigh in.

Influence has a number of components which can be calculated independently and often with entirely free tools;  scores can be weighted and an overall influencer rating can be assigned based on what is found.

First, I hold individuals are  influential on select subjects, mainly, and determining online influence requires the ability to categorize websites and backlinks by subjects.   By extension, we can categorize individuals through the content they author and who links to that content – therefore,  websites can be categorized by the following factors, at least:

1. Web Directories (Yahoo! Directory, Google Directory, to name 2 of the main directories)
2. Analytics Platforms (Such as Compete, Quantcast, Microsoft AdLabs, Comscore, Nielsen, Hitwise)
3. Blog Directories (such as Technorati)
4. Twitter Analytics (Tweetlevel, TweepSearch, FollowerWonk, Klout,  to name a few)
5. Frequency of Keyword Usage
6. Categorization of Keyword phases into categories/segments
7. Twitter Lists

First Deduction:

Registering for a web directory is a good thing to do now.

Here’s my reasoning:  While Yahoo Web Directory sucked and does not drive traffic it is still used by search engines to categorize your site – and can also be used to categorize conversations and backlinks – therefore….. it’s a good thing to be signed up to Google Directory and Yahoo!Directory again, but more for categorization purposes than anything else.

As web directories are often categorized by human beings – it’s going to be more accurate then automated analysis to segment sites into categories.  I suspect, however, that if your perceived to be an expert in to many categories it degrades your overall influence.

Here’s what I’m considered to be an expert in (and this can probably be confirmed by the amount of times I mention those topics in my online content and the backlinks coming to my site)

But what if I had 10 areas of specialist instead of two or three – would the extra levels water down the overall importance of any category?

I reason that it would - just like the case of having 300 links on a web page – they all pass pagerank but it’s diluted by the number of links on the page (think back of “pagerank sculpting” that some have talked about in SEO circles – though I think Matt Cutts came out against doing Pagerank Sculpting).

Also, those who design algorithms to detect influence make them in their own image – the human mind may not be normally (with a few instances) be able to hold into short term memory more than, say, 8 numbers – it’s not that people can’t have longer phone numbers, for example, but most people could not remember them if they were, say, 15 digits instead of, say, 7 digits (again, with some notable exceptions such as idiot servants).

I believe we as human beings, try to make search engines, influence mapping based on how our own minds work – even though we, as individuals, are far richer – we may not, at our present state of evolution, be able to “contain” more than a few things at one time.

So – as we function as people – I think online reputation mapping tools are based on the same “limitations” we as people function on – even though the algorithms technically don’t need to be limited that way.

It seems to me – unless there is some other way to do it that I’m not aware of – identifying Influencers must, by definition, include Categorization that is done on a vast level.

And,  if we have something like this weighting schema – we might be able to test out and see if we can just get our own influencer list – even out of Google, or Trackur, or any number of free tools that’s every bit as good as anything we can get somewhere else (providing we take the time to vet the list).

There’s more factors to be sure – but I don’t want to make this first post too long and this at the idea level right now – I’ll show you examples soon enough.

Do want to say something before closing – I picked a Six Month rolling average rather than 1 year or 1 month or 2 weeks, etc.   A lot of Twitter Influence calculations (ie: Klout.net) is too much based, I feel, on what your doing now – in the last day, week, two weeks ,etc – because that’s what Twitter is all about – the moment – and that’s fine.

But …. real influence is built over time – and Klout’s approach is too much based in the moment.   Take one example – I go to museums alot  – including Modern Museums like MoMA – there’s all kinds of stuff in MoMA that is interesting – but it hasn’t been tested much over time – and the whole point of Art in MoMA and other modern museums is to interact with it ….. today what’s in Museums might be considered Influential by someone who curates a show …..

On the other hand, I walk into the Metropolitan, see Art where there is a general consensus over time from many curators, alive and dead, and many communities, that value what’s hanging on the walls – that to me, is more valuable than a bunch of curators who think something is valuable or influential today (because it may not be, tomarrow, or next year, or 10 years ago).

Someone may do a lot of posting on a subject for a few days or weeks – then be silent – I think a 6 month rolling period is good enough – anything less might not do it – anything  more (a year, might diffuse activity too much).

Just want to point out I’m still going to review Social Radar today and other platforms as I’ve been doing – but the stream of “ideas” and insights comes at it’s own time – I’m simply trying to write it down and share it – before I lose grasp of it (as often happens in my life).

Yes, this is all very general and hazy – but you need a framework first – before you build concrete examples.

Any thoughts?

Feedback from the Influencer Scorecard Summit – two weeks ago appeared from K. D. Paine’s Notes from the Influence Scorecard Summit and Philip Sheldrake posted at Marcom Professional on How the Influence Scorecard radically transforms marketing and PR, both which I’ll comment about here.

The Influencer Scorecard transmutes Social Media and Public Relations by identifying, perhaps, for the first time, where every function of strategy, planning and execution belongs, who owns it, and what the right sequence ought to be (see the Influence Scorecard 2009,  below, along with a little movie I made on day one of the summit – to provide a flavor on our meeting ).

Influencer Scorecard Summit – short video — day 1.

Philip Sheldrake suggested major Social Web Analytics vendors adapt a methodology that can be built into a  scorecard – which might not be easy to do – but here’s what it would look like, in concept, if they did.

That’s contingent of course upon the leading social Web analytics vendors quickly picking up this approach and developing their products and services accordingly.

Philip envisions a “Chief Influencer Officer” rather than the CMO that currently exists, who ..

…. knows precisely the state of all six influence flows at any point in time. She is sensitised to her organisation’s environment in a way that makes most CMOs today look like they work in little bubbles where they had no choice but to “make stuff up“.

But, what’s interesting to me is the Chief Influencer Officer could actually be filled by a “Web Analyst” in much the way  Eric T. Peterson said (last year), someday, a Web Analyst would be on the cover of BusinessWell and Fortune Magazine -  of course, he or she would not be called a “data analyst”, but may end up being one, none the less.

Philip also saw the transformation of online commerce into “Buyer Marketing”

the balance of power shifted somewhat from the massive dominance of large organisations at the end of the 20th Century back towards the individual. This rebalancing will continue during the next decade. It will give rise to something I call buyer marketing; similar to what Doc Searl’s VRM initiative refers to as “personal RFP” and most recently what Scott Adams has labelled “broadcast shopping“.

Philip Sheldrake builds off my own prediction that Google will enter the Social Media Monitoring space in the next year or two and blow away half of the vendors currently in the market (by merging what Google already knows about us into Google Analytics profiles) – I think this is quite plausible and likely (see my deck, slide 15, below).

And PR will be transformed because advertising and marketing are moving into Public Relations, more and more, as CMO’s become Chief Influencer Officers (as Philip puts it, and some of them end up running PR agencies).