SAS's Coming Out Party - Analyze This!

SAS's Coming Out Party - Analyze This!

Summary: The DisclaimerSAS's legal department, which seems to be the only SAS department retaining the outlook of the early 20th century, asked all the bloggers who are posting who had their travel expenses covered to put in a disclaimer that they had all their travel expenses covered.  My disclaimer is as follows:  See Denis Pombriant's disclaimer.

SHARE:
TOPICS: Software
2

The Disclaimer

SAS's legal department, which seems to be the only SAS department retaining the outlook of the early 20th century, asked all the bloggers who are posting who had their travel expenses covered to put in a disclaimer that they had all their travel expenses covered.  My disclaimer is as follows:  See Denis Pombriant's disclaimer. I am officially in complete agreement with that as mine too.  Plus I think that SAS legal is off base and misinterpreting the December 2009 ruling of the FTC that covers paid blogger endorsements being revealed.  First, I'm not a "blogger" per se. I'm more of a consultant and an analyst who is hired as a consultant by almost every company in the CRM space for some reason or another. Second, I NEVER EVER will endorse anything for money. Ask any vendor I've consulted with about that .  Third, I really think they misinterpreted the ruling.  But, hey, they're the legal department.  I don't mind the disclaimer. I mind that they gave me suggested language.  But then again, the early 20th century was like that.  CYA .

But that is a mere poorly done blip by comparison with SAS's triumph in April 2010 - the 21st century.  They shined this week.  Let me fill you in.

A Bit of History

About 4 years ago, I was on the SAS campus to do a presentation to one of their internal groups.  I had the opportunity to both work with them and to get a tour of the campus, which among other things included a private school, a staffed medical facility, a picnic area, and a world class, museum grade, gems and minerals collection.  I was awestruck. It was no wonder to be that SAS was considered one of the top five or so places in the world to work at that time, before that time and up to and including, now. In fact, at the opening of the SAS Global User Forum and Executive Summit on Sunday night, the Great Place to Work Institute gave them an award for being their choice to as the #1 place in the United States to work.  It was no wonder they had an attrition rate of south of 2% routinely when it came to staff churn.

But at the time, there were several things that puzzled me and even disturbed me. The world was starting to boil as a very clear cut new communications revolution began to emerge and people in their customer clothes were beginning to think quite differently how they wanted to both think about the purchasing decisions they made and second, how they trusted - and who they trusted.  It wasn't that hard to recognize. The presentation I gave to the internal group of about 100 database marketers, thanks to the really smart and very good person, Marketing Director, Deb Orton who brought me there, was on the death of the 4Ps.  Marketing was clearly changing and what had to be done to accommodate those changes was emerging.

Yet, in the public domain, SAS was peculiarly immune to the changes. They didn't really respond to them in their advertising campaigns. They didn't participate in the new communications that their competitors were jumping in on.  They had such a loyal customer base and such committed employees that they were pretty much being assured growth of some kind regardless of the world roiling around them.  In fact, I remember I thought of them as a wonderful but insular organization living in a real life version of an Oz of their own making.

But this company of 250 products if nothing else, was serious about its products and about using them to see what was going on in the world.

I'm not sure what the actual motivation was but I have to tell you right now that not only has SAS released a truly excellent social media analytics product - a unique one in fact, but they have attacked the market and their own culture in a way that has transformed the company to something new and perhaps equally as good as their "place to work" accolades indicate.

The Product - An Analogy

In France, there are what many think might be outdated, but nonetheless influential classifications of Bordeaux wines. The classifications go like this - in English - First Growth, Second Growth, through Fifth Growth. Most of these were classified in 1855 with a few exceptions, but all in all they are the ne plus ultra of wines. If you make the grade at one of these levels, you are golden.  For example, Chateau Lafite Rothschild is a First Growth from the district of Pauillac.

But there is one wine that stands out so greatly from the pack  - a Sauternes - a dessert wine - Chateau D'Yquem.  It has its own classification above First Growth as a unique product.

While I like Chateau D'Yquem better than I like SAS Social Media Analytics - especially after a great dinner - the product that SAS put on the table has a unique character that stands out in the market as, at this point, one of a kind and thus doesn't have anything to compare against as of yet.

The Product - for Real

Don't get me wrong. It will have competition. It may even be bested by someone if SAS gets complaisant (though I'm not feeling that right now, know what I'm sayin'?) But, at least for now, you can't compare this to any product on the market.

First, let me tell you what it isn't.  This is not a social media monitoring tool on the order of Radian6, ScoutLabs, Attensity, or any of the 170 roughly other SMM tools out there (see Nathan Gilliatt for the best aggregate look at the SMM market).

Am I say that the Social Media Monitoring tools do NO analysis? No, I'm not. There are products like Attensity that do some - and in fact, in the case of Attensity classify themselves as Voice of the Customer products. But what SAS SMA does better than any of them is social media analysis with enterprise class depth and a surprising ease of use. At this stage, none seem to compare.  I say "seem" to compare because I have to provide a caveat since this is a brand new product with not a lot of customers yet so it doesn't have the seasoning.  But its something like Jason Heyward, the Atlanta Brave's red hot rookie, who is called the J-Hey Kid because Willie Mays, one of the greatest baseball players in history was called the Say Hey Kid. He has that potential and shows flashes of it all the time.  Just like SAS SMA.

What It Is

What makes this a great out of the gate start for SAS is that they thought through all the components in a way that makes this, as the always astute, the redoubtable Esteban Kolsky calls it, not just a solution (which is what SAS tends to think about it as) but a platform that is eminently configurable and customizable.

In developing this product (I had the pleasure of doing a day of paid consulting at SAS HQ in February so I got an early view of this thing), they considered all the requirements of companies in a contemporary environment. As key product development leader Marc Chaves put it "We worked from the end back."  They identified what the customers would likely want from the solution and did it by, gasp, talking to some of their customers to find out. Initially, they concentrated on the hospitality industry and the financial services industry, both high touch, both with clearly measurable objectives on what they wanted to monitor and both with specific and identifiable taxonomies, which were especially important in how sentiment analysis algorithms were developed. Once they were able to figure out that (I may not be getting the exact order right, much of this was done in parallel) out, they took some real care in trying to figure out the communications channels for the customers in those industries. Obviously, they drilled into the usual - Twitter and Facebook - but they also smartly allied with third party data sources who monitor things like threaded forums and user groups.  So they didn't do what everyone else did, they actually figured out how to handle the less traveled and less sexy , but still conversation heavy traffic in alternate channels.  That way they nailed the traditional, structured data sources like Reuters, the now popular unstructured data sources like Twitter and blogs, and the less known but still really important data sources like those threaded forums.

But they didn't stop with that.  They understood this wasn't strictly a matter of a singular quantitative approach to how influence works. They knew that what defined an influencer on Twitter was not the same as what defined it on Facebook.  So their approach to each, for example, is different, in how an influencer works.  More on that later.

They also understood two other things of real import. First, they are a company that has historically dealt with the enterprise so they had to scale to the enterprise. Second, the enterprise was increasingly looking at delivering solutions in the cloud or at least on demand - while, of course, the product development environment at SAS was more on premise focused, though not strictly. They had other on demand products like Customer Intelligence on Demand, for example.  But they understood, not just for the delivery requirements of the customer's IT department, but for their own flexibility in developing the initial rules, roles, taxonomies, etc. the model for on demand was better suited than on premise.

They didn't stop with just those bits of engineering either.  To put it politely, SAS was never known for the beauty of their products.  No catwalk or runway for them.  But they spent serious production time in developing both a pretty attractive interface (its still not beautiful, but it is nice) and an eminently navigable, easy to use interface that was appropriate for actual human use.

Here's a screenshot of a dashboard that can give you the idea.

First, note one important thing. As stated above, this is a readable, attractive dashboard. And it comes from SAS, who, to put it somewhat mildly have been, shall we say, graphically-challenged over the years. Not this puppy.  This dashboard can be understood by anyone from the beginning. The deepest parts of the analysis are not that visible but what you need to know at a glance - the purpose of a dashboard - is right here.  Note something else, On the top you see pull down menus - one called source, another, business unit. Underneath it you see Days.  So the way you can change what you're looking at? Pull down a source or business unit name, move the slider on days or go to another tab and something new is there for your perusal. Easy.

Okay, I'm not waxing rhapsodic.  This isn't art. Its software. But it is pretty nice and step forward.

The other three things that make this an excellent product out of the gate?

  1. Predictive analytics
  2. Social network analysis
  3. Speed

Predictive Analytics

Of all the products out there that do some form of text analysis including those above, Clarabridge, IBM and multiple others SAS does what no others do - that would be predictive analytics. They have scenario building capabilities - the what if stuff - that allow the users to predict the volumes (with of course, no certainty of correctness - just a better guess) of social media conversation. But what adds to that capacity is that the tools can also then take those forecasted volumes and anticipate their impact on a brand or a campaign or an event.   No one else, at least to my current knowledge does that.

Social Network Analysis

SAS and a director of Los Angeles County, did a fascinating though slightly scary (implications) demo of the use of social network analysis when it came to public aid fraud detection. They showed the analysis with some beautiful geographic visuals - exposing patterns of activities among social networks to show where fraud rings resided and who were likely candidates for the title of "Prisoner 2010" due to their possibly fraudulent activity. There are some scary privacy and ethical concerns from my end on the whole approach - but the quality of the analytics themselves was clearly superior.  Social network analysis (SNA) has enormous business implications due to its ability in either a B2B or B2C environment to detect influence, which as easy as it sounds when thinking about Twitter, is far more complex than "Retweets = level of engagement" (which, BTW, I contend is not as significant a signpost as its made out to be).  Influence varies from time to time and moment to moment (a big future post coming on influence). However, what SNA can do regardless of that is identify continuing patterns of influential activity - which has value when trying to find out who influences what decision makers in a company in a given timeframe or who is a brand influencer in the B2C world.

SAS's articulation of SNA for the product is deep, visually attractive and important. Its a rare instance where a product that does other things offers that. The value of SNA in the B2B environment is that it can use the social graph, and internal data to find out who influences the decisionmakers, though I have to say, that with this product, it would be an awkward, overly expensive use of the product, if doable at all.

Speed

But keep in mind, doing all of this takes time and processing power. Hence the on demand nature of the product. SAS provides the processing power.  But what's even more fascinating is the constant improvements SAS is making to these products speeds. Without saying too much, on the opening night of the event, SAS execs showed a capability to deal with up to trillions of records really fast. What once took 18 hours took 2.26 minutes when it came to reading and analyzing records.  No more needs to be said.

My  Concerns

Pricing is my first concern. The set up fee is $50,000 - not an insignificant sum. The continuing subscription price is $5000-$15,000 per month - not an insignificant sum either.  Since there really are no similar products on the market, hard to say how customers are going to view that, but there is a popular perception that any tools with the words "social" in it should be free or cheap.

Second concern, their claimed 90% accuracy rate for sentiment analysis. I don't doubt the veracity of their claim. SAS, if nothing else is scrupulous about numbers and won't release anything they can't stand by. But at the moment their volume 90% success is their testing in a couple of clients. Are they going to be able to continue to reach those accuracy levels when they are scanning trillions and quadrillions of lines of text.  Not any easy task - and dangerous to some extent - if you're staking your claim in part to these high accuracy rates.

My third concern, and perhaps the most volatile, is that SAS pretty much has to do the work when it comes to setting up the rules, making this concerning when it comes to the time to completion of a rules addition or change. The conversations are occurring in real time. Something that needs - at this point - the vendor to make changes - while the conversations continue on - is a potentially real problem. In fairness to SAS, though, Jim Davis, a rather dynamic CMO, told me that SAS is aware of this and working on it.  This, BTW, is not the tools that do the analysis that I'm talking about. They are easy to use.

One Final Thing: The Culture

I'm going to start with a story.

We were sitting around at the launch with Mark Chaves talking about the product with several leading industry journalists at the table. Among them were Marshall Lager, President of The Third Idea and industry up and coming thought leader; David Myron, Editor in Chief of CRM Magazine (among others); me and Alex Williams a senior editor at ReadWriteWeb.  Mark had identified the SAS SMA as a game changing product for SAS and gave some business reasons why that was the case. I piped in and said "I think that its more than that.  It pushed a change in culture when you had to think of how to create, launch and deliver the product. It was a node in a major change in culture for SAS." Alex Williams asked "What do you mean? Can you elaborate?" I said, "Three years ago, Alex, you wouldn't have been here.

In short, that reflects exactly how this great place to work but insular company, is now doing its outreach via social channels. They've adopted contemporary technologies and new strategies to begin to get to people who wouldn't have ever been on their radar screens at all if it weren't for how the world shifted over the past 5 or so years.  They have a kick-butt Social Media Manager, Dave Thomas who not only knows the social space, but also how to organize the kind of social campaigns using the tools that scale them - and do it well. Additionally, Alison Bolen, a former editor of SAS magazine, is "former" because they've transitioned her to teach the internal staff (chosen ones that is) how to blog.

But much more than that, they've recognized the value of the very space that they're analyzing with their product.. They've understood that influence in the outside world matters, and is not a static thing. So there were several independent analysts - me, Esteban Kolsky, Denis Pombriant, Brent Leary, Josh Greenbaum who were invited. There were social folks like Alex Williams of ReadWriteWeb and local Seattle blogger influential Shannon Paul.

Not only that, but they had an army of retweeters who RTed all the stuff that people like me and Esteban Kolsky put out via Twitter as the launch of SAS SMA was announced. At the keynote on Sunday night, there was a live and rather gorgeously rendered Twitter stream on the stage that was readable by all. Meaning they did it right as a launch with a social component.

My only concerns were two fold. At the launch event, there were NO institutional (e.g. Gartner) analysts with butts in the seats.  That's because they are treated differently than people like me or Denis who function (at least in moment) as independent analysts and consultants - the distinction is an artificial construct when it comes to dealing with that.  They still distinguish the function.  In a way that helps, because our homies love working with SAS PR star, Angela Lipscomb and that's a good thing.  Even if the function remains separate though, the institutional analysts like IDC's superstah Enterprise Apps ruler Mike Fauscette should have been sitting next to me at that launch as should, say Michael Maoz, no matter who they deal with.  Second, there was no representation of long time enterprise analysts except for, thankfully, Josh Greenbaum.  Thing is they carry huge clout in the world of practitioners and there is no trade off between social and enterprise - since we're trying to build social enterprises. SAS can't get blinded by the social part of this to the exclusion of the 20 year influencers who are called by the very prospect companies that SAS is courting.

But all in all, those are not major concerns, just things to be corrected for next time. SAS's first venture into the social marketplace was a product, social, and cultural step forward for the company that positions them incredibly well for the future.  Now they have to deliver.

Topic: Software

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

2 comments
Log in or register to join the discussion
  • Keep the feedback coming!

    Thanks for your thoughtful post Paul. It has generated a lot of conversations internally. We really welcome the feedback.

    Angela
    SAS Media Relations
    angela.lipscomb@...
  • Glad to have SAS coming to the party

    SAS is absolutely one of the best software companies of all time and it's extremely exciting to see them joining the social media analytics game. No, Radian6, Attensity or ScoutLabs are not competitors to this product.

    That being said, SAS's social media analytics platform has a VERY, VERY LONG way to go before it's an enterprise solution. Right now, it's in poor form compared to Social Radar, Sysomos, or Crimson Hexigon.

    That massive amount of "Mixed" sentiment is because their natural language processing (NLP) algorithm stinks. SAS either isn't taking this project seriously yet or underestimated the level of effort in pulling knowledge from the written word vs. counting dollars and cents, inventory or customer service calls - numbers are easier to count an analyze than is a line from Emerson.

    User interface and graphics? Good god, this solution is ugly. The only thing more outdated than the look and feel of this application was the presenter's microphone and the odd, old school news broadcaster format.

    $50K topology set-up fee? First, how long is any topology going to stay current - the net changes every day as does the questions you'll have for the platform. The web is not a relational database - it is fluid. The topology fee is a complete waste of money - either due to a misunderstanding of how the solution needs to work or because SAS needs to support on-premise type margins.

    SAS is a great company, with great people. But, they have a very long way to go before they can compete with real solutions in this marketplace. Some of their enterprise customers may be clueless enough to invest in this solution, but most of them are much more savvy than SAS realizes. If they actually knew the market, SAS would have waited to release this solution or would partner with one of the leaders to leverage their data and algorithms to get their own offering up to par. Many companies are already using powerful enterprise social media analysis platforms - SAS's won't be one of them anytime soon.
    @...