Subscribe to our Newsletter

Big data: the end of the beginning?

I looked at traffic statistics of popular 'big data' websites: publishers, media companies. I noticed that most are tanking. Most of these websites were two stars (US traffic rank better than 100,000) a few months ago.

Web traffic trend observed in the last 30 days

Two of them (AllAnalytics, BigDataRepublic) lost one star and were downgraded to one star, but two other ones were upgraded from one to two stars (DataScienceCentral, AnalyticBridge). The data source is Alexa.com rankings, as of June 22, 2013. Numbers are subject to a 20-30% error rate due to Alexa inaccuracies, but the pattern identified is observed over many large web sites, and shows consistency over time.

Interestingly, the ones that are losing traffic fast are run by journalists, not data scientists. One of the biggest losers, AllAnalytics.com, seem to be a website where four or five people are talking to each other, mostly writing comments such "I agree with you Tom", "Great comment Tim" etc. It is so bad it is actually funny to look at these conversations, who knows, these five guys are maybe just one person. But even someone talking to himself would do a better job and post more diversified or antagonistic comments.

BigDataRepublic has been using very poor statistical methodology when they wrote their landmark article in April: the list of top big data influencers. They were heavily criticized. You can not make this type of mistake when your audience is analytic minds. Even today, their #1 big data guy is an obscure market research guy, and you are wondering if he paid to be listed at the top. Needless to say, their traffic went down after this fiasco.

Yet, two of these websites are growing steadily (they are run by real data scientist and computational marketing experts), VC funding is strong, and a rich ecosystem is developing fast. So what is going on?

I call this the end of the beginning: we have reached a point where you can no longer claim that you are a big data guru or a data scientist, if you don't know what you are talking about or know nothing about data science, or sell fake data science. This era is gone. We are entering the second stage where many new companies are still being created, mergers are likely to increase, but the pure opportunists get weeded out while hype evaporates like ice in Greenland. For many, the question right now is show me the ROI.

Related articles

Views: 3417

Comment by Vincent Granville on June 23, 2013 at 10:08pm

You can only go so long posting over and over the same stuff. At some point people are becoming immune to your hype.

Comment by Jean Michel LeTennier on June 23, 2013 at 10:12pm

It's only HYPE , if it's not true..  and to be fair, I have reached out to you and you have wrongly assumed that I am exaggerating..   What I find sad, is that a Scientist of your caliber, would write anything off without seeing it.. trying it or even discussing it.. So, we will let the Market decide.. and I can assure you they are interested.. in no longer being held hostage by IT..

Comment by Jean Michel LeTennier on June 23, 2013 at 10:15pm

Associative Database Technology does exist, and will over take anyone and anything not using it as  base

(10) Patent No.: US 8,051,102B2
(45) Date of Patent: Nov. 1, 2011

Comment by Vincent Granville on June 23, 2013 at 10:33pm

@Jean-Michel: I have made no assumptions. 

Comment by Jean Michel LeTennier on June 23, 2013 at 10:43pm

Sorry but the use of the term "HYPE"  indicates as much..  for the first time in 45 years, there is the opportunity to dramatically expand the abilities of all researchers , all businesses.. all manner of innovative minds.. no longer constrained by a two dimensional digital solution..  .. This is a paradigm shift, we are simply bringing it to people's attention..  Having read a few of your papers, I can easily see where this technology vs the technology you were forced to use would have made a significant impact..,. but please let's be clear , this is not HYPE, as you can see by my profile, I am a no nonsense person, and too old for crap.. and I am in the data word 30+ years myself.. so I do have a clue. ;-)   my best to you

Comment by Jean Michel LeTennier on June 23, 2013 at 11:00pm

@Vincent, as a last aside, our new PHP/JAVASCRIPT version is available now and because we are no longer constrained by .NET libraries.. we are seeing performance increase of 300x SQL server. Our engineer e - basically removed the bottlenecks form the java data objects, so we run natively in JAVA as well.. 

 I think you would have a blast Dr.. I really do.. ;-)

Comment by jonathan seller on June 24, 2013 at 7:20am

This is great news! For this to become 'real' it has to pass through the various phases of the hype cycle. Next up, the 'Trough of Disillusionment"; and once that's over we can get real work done without the buzz getting in the way. http://en.wikipedia.org/wiki/Hype_cycle

Comment by Jean Michel LeTennier on June 24, 2013 at 7:31am

with all due respect..  Too take Wiki as the ultimate source of correct information? surprising.. but I will refer you the Patent and that may help,  additionally to several PDFs on the subject if you like.. by world renowned scientists..

(10) Patent No.: US 8,051,102B2
(45) Date of Patent: Nov. 1, 2011

 

To the end of becoming real, so to speak, We have engaged Dr Victor Tang of MIT to review for one of client projects as well as several other professors ( in the field)  in the UK, Brazil, Australia, Belgium, Canada,  and India.. and Dr Munoz of UCAL (hopefully) and of course Dr Granville (hopefully eventually) ;-)

 

Comment by jonathan seller on June 24, 2013 at 8:19am

lol. Starting a comment with that line says you don't at all. Get off your high-horse Jean; it was a comment to provide some input. The study of 'hype' is hardly deterministic, but but thanks for the more 'correct' information. 

Comment

You need to be a member of BigDataNews to add comments!

Join BigDataNews

On Data Science Central

© 2019   BigDataNews.com is a subsidiary of DataScienceCentral LLC and not affiliated with Systap   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service