Had a dream about the perfect Influencer platform but when I woke up I couldn’t find it! Ha!
Anyway, it’s been a few days since I posted last and found that, lately I have been learning and taking in so much new information I didn’t feel I had as much to say about it, because I had not yet internalized it (and decided what was of value to me). Broadcasting information might be OK for me to have done in 2006-2010, but lately I need to step back, internalize the information, and then decide what it all means.
When something hits me in the gut, sorta like a instant reaction, yes, I think it’s great to write about, but lately, as information accelerates, simply repeating the message is adding to the noise, not generating signal.
I was lucky to ride out the storm in Rhode Island though many of my friends weren’t as fortunate and I don’t think we ever saw anything this bad in the Tri-State region in human memory. The only time NYC was flooded like it was this week was 200 years ago, and at that time, the waters went to Canal Street. All classes for my Rutgers students were cancelled till Monday, November 5th (when I’m returning to NYC) as well as all New York schools. I think its’ the same story along most of the East Coast.
Decided to look at local Impact of #sandy using Marketwire Sysomos over the last week (Oct 28th – Nov 4th) – almost 80% of the online volume is Twitter mentions, retweets, etc.
Obviously, getting any sort of intelligent readout would mean knowing a lot more about precisely what we want to explore or know about. One thing that would be interesting is to measure the volume of mentions about the general election at the same times and locations, to see where the storm, as expected, became more important than who’s going to elected next week (hopefully – let’s how there’s not big delay over contested votes).
From what I can tell the main impact of the Sandy on the elections (from the standpoint of Twitter was on October 31st and November 1st), else, little change in online volume, though the volume of mentions about the storm far surpassed anything being said about the election by a factor of 50:1 during the first couple of days of the storm.
I know that I’d like to add value to the post by talking about what people expect me to focus on (analytics) but my feeling is the data needs a lot more work and definition before we can do much more than provide a buzz map. And just looking at NY, it’s pretty obvious, once you have the context, what the buzz map means (see below). Most of the conversations were about the storm hitting landfall and evacuating the Subways due to flooding.
I suppose we didn’t need this buzz map for that insight!
I guess the problem is that you need to know precisely what you want to explore on a very granular level, else your just going to pick up little signals here and there like “Hurricane Muhammad” that don’t really amount to much of anything.
Query = (Sandy AND (Hurricane OR “Tropical Storm”)
Popular Phases on Twitter
Delving into “Hurricane Muhammad”
Overall Share of Voice of 7 hardest hit States (all channels)
I noticed Recorded Future’s new look and I like what I see, though matching up output about Hurricane Sandy between platforms is probably not going to yield much of value. I think we must look at each platform for it’s own unique insights and leave it at that, for the time being. I embedded a view from Recorded Future about “Sandy” but I don’t know how well it will display, below.
The Recorded Future doesn’t seem to work the same way it did before, I can’t tell you what the outcome will be based on this map. Again, I think the results will be better with more work and segmentation of the sources, as that always seems to help.
What I thought was more indicative of the future on Tuesday, November 6th, was all the fervor over FiveThirtyEight blog and Nate Silver’s predictions based on sound polling theory that Obama had an 83% chance of winning. It seemed as if people want to live in Bizarro world and just believe the data they want, regardless of how it’s collected.
As long as “quant” and “nerdy” Nate Silver was predicting Baseball, no one had any problems with his methodology – but as soon as he started talking politics, esp this year, people on the Right (who didn’t like what they heard) jumped all over him and it has become very ugly.
Of course we can expect and almost certainly predict “election fraud” will be in the news the next couple of weeks, esp in Ohio and Florida, where there is reason to believe there is fraud. In Ohio, Tagg Romney, one of Mitt Romney’s sons is invested in Diebold and many of the voting machines are running “new software” and in Florida, well, it’s already a mess.
One doesn’t need to be a soothsayer to expect the election to be contested save a decisive victory, which looks unlikely – hope I’m wrong about this.
Here’s another presentation that predicts an Obama victory based on Web Search Volume – of course, who knows what will actually happen at the voting booth, while voting and the vote itself. But the indicators all seem to point to an Obama victory – for the most part, but we’ll have to see what it all looks like Wednesday morning.
Influence and towards a New Influencer Dashboard
Everyone who read my book pretty much knows how I feel about Influencers and online Influence – there’s not much new to say – but as I had that dream last night about the perfect dashboard (and I’m not saying I came up with one), I feel it’s necessary to say something about what’s wrong and doesn’t work, along with what would work here.
First of all, I think we’re at a point now where we should stop building platforms to satisfy a particular audience or potential buyer and step back and think about what we really need to build, and then decide to build it. With that in mind, here are some of my thoughts about an Influencer Platform I’d find useful – not the one’s that exist (except to say there are some features that might exist in various platforms, but the whole thing exists nowhere).
- Features of a new influencer dashboard (sort on any four dimensions)
- channels I am most active in over the last year and over the last month
- best times and channels to reach me (historical data)
- the topic cloud over the last year and over the last month
- any contact information such as a phone number and email address
- any particular event tweeter or posted about over the last 2 weeks (good for an opening pitch)
Additional (harvested data based on public data)
- offline influence markers
- events I spoke at
- events I attended but did not speak at
- meetup groups I belong to and the meetups I attended
- charities I done to and causes I talk about outside social media
- when applicable, any particular marker such as a disease or common life experience (say, the holocost or 9/11)
- member of professional organizations
We can also score on the features – I would tend to favor offline signals equally with online, perhaps even more than online as they take more effort and effort demonstrates interest
I would leave out financials, credit and other kinds of public info as it is too contentious to add them, even if it were possible to so. This is not about technology as much as it is about what actually makes sense to do a human level.
This is already a long post, but I leave you with a video – digest this (but don’t go with Adobe, they’re as much hype as anyone else)