Half of Analytics Investments By Companies and Banks Will Be a Waste–What Do We Analyze with Big Data and Does It Have Val
Posted Feb 25 2013 6:38pm
This is a great article and some folks who have some foresight here. I write the Medical Quack to bring some understanding to the layman level as best I can and already I’m seeing this in some of the stuff in healthcare. You can’t turn linear data into non linear and expect to find relationships that have value. CEOs will squash their Quants to the max though to come up with something they can sell. Boy have we seen that and it just doesn’t work. I keep touting the videos on the left hand side of this blog and if you want some answers and explanations, watch them. These are people smarter than me and in my time of developing software it’s the same mechanics with queries and developing analytics software, just bigger today and more headaches to find that one in a million model that “may” be correct and “may” work.
When you see all these stats thrown at you day in and day out, it’s the old bait and switch as anyone who has any kind of a sales or marketing background knows that’s the oldest trick in the book to start firing back with statistics whether they are correct or not. Money drives this of course and if you have bunk stats and the person at the other end sucks in, then you have made a sale and put money in your pocket. I like accurate data and how it makes me smarter and how I can use my human brain based on good research with good clean data to make decisions. We are not getting that anymore out there all the time and it’s getting difficult to tell the difference as everyone has some really nice marketing campaigns that are done so well, it’s almost hard not to believe. Well let’s see that math formula looks good, it even has a square root in it and I’ve seen it all week, so it must be good, right…not always.
The article says around 50% and I think that’s a fair estimation and one man said it was 40% being good data. This is one of the better posts I think I have done to explain how this works as you have big companies, NASA and others in a forum discussion on what in the heck are we doing with our big data. The gal from T-Mobile is my favorite as she comes right out and says what we are doing is “silly”. Hats off and kudos there. First part of fixing anything is recognizing there’s a problem:) Watch the video at the link below for more insight as to how this works and the discussion as it’s a good one.
We had in the news the big TIME article about the expenses of healthcare and there was no mention of Health IT, so how can you leave that out? It was a good article but left out the math and formulas and it’s the latest link in what I call my Attack of the Killer Algorithms series.
I probably read too much but occasionally I take on the garbage stats that appear when they get to the point to where they are claiming savings that “nobody” can predict. When I start seeing “trillions” in savings claimed, red flag:) I realize that I see things a little different than how how the average consumer does but keep in mind, perception is everything and I get right down to the reality of how this works as I’ve done some of this and someone who has never written a stick of code has no clue on the mechanics and thus so their perceptions will be all over the place as you just don’t know.
There’s plenty of folks out there much more versed in areas than I am because that’s their area of work or focus, been that way for years so I don’t know everything but the mechanics of how this works with selling data is black and white as I intend to present it, no algorithm fairies just mechanics on how it is written and how it works. Politicians and others can talk all they want but until changes are made at this level, nothing changes.
If all of what you see out there, and let’s use the markets as an example, were so good, we could fire every analyst off the Street, right? Point made here is there’s still levels of errors and in linear calculations, machines do it better and they the human with machine assistance can come to some “educated” conclusions, hopefully, if they are not Algo Duped if the answers don’t look quite right:)
We see a lot of “strange” perceptions too from those in Congress and a big part of this is a low or no level of participation with consumer technologies and that old paradigm of “its for those guys over there” just won’t go away. They say they participate but come on if they did, we would not be hearing some of the idiotic stuff we hear today..non participants stick out like a sore thumb, especially when they are really reaching out to gain some control over some issue and that’s why we have this deal on woman’s health that comes up, digital illiteracy. I should have started a series on the “those guys over there” as there’s certainly enough of it (grin). Those folks making laws certainly need help understanding technology for sure and the Sunshine Foundation is right here.
So coming back around on topic here, the analytics that you see out there do not all have value but developers, quants and others who have this big pressure on them to make money are getting very creative today and you about likely to see any and all kinds of analytics out there. The one big farce that I see is the FICO medication compliance analytics, this is being sold to make money only. They are doing one heck of an effort to market though but they are using both credible and non credible data together and sure you crunch numbers to trend as that is helpful but not when you take it to the limit they have with individual scoring and these bogus results, kind of like the old grading on the curve if you are old enough to remember…yeah…you don’t' get your real score (grin), that’s the name of that tune.
Remember that company Accretive and their analytics?
How do you like these business analytics at the link below? Which 50% area of Business Analytics should this fall in? I guess that depends if you are a consumer or a big corporation trying to make some millions selling data.
I agree with this article, some folks with some smarts who are not afraid to speak out, even if it is in Australia, it all works the same world wide. So you can read what I suggest to help this out because folks are just running wild with analytics and not all of it is good and manufacturing can use a boost and this is the way to do it with weaning folks of the algorithm for lunch trail and at least make some take a look at “junk” if that is what they are anticipating selling for profit when they know “real” value is not there or if they have had their quants fictionalize the risk. That happens.
Organizations have a 50-50 chance of yielding any benefits from their business analytics investments, according to Commonwealth Bank of Australia (CBA) manager of information systems and frontline analytics David Tanis.
During a panel discussion at the Gartner Business Intelligence & Information Management Summit 2013 in Sydney, SAP innovation evangelist Timo Elliott said all businesses will gain something from every business analytics investment they make, since there is a lesson to be learned even in a failed project. But Tanis, who was also on the panel, was not so optimistic.
The third panellist, QlikView vice president of global marketing Henry Seddon, was even more pessimistic, tipping the failure rate of investment into analytics to be about 60 percent.
Big data has commonly been defined by the three Vs: Variety, velocity, and volume. Many organizations are attempting to harness the potential of big data for a number of uses, from gaining meaningful business intelligence to targeting marketing strategies. Each face their own challenges in analysing the vast amounts of unstructured data that are available, which can come from many different sources, including social media and video content.