The Musings Of An Opinionated Sod [Help Me Grow!]

Is The Only Thing Big About Big Data Is The Hype?
May 31, 2016, 6:20 am
Filed under: Attitude & Aptitude, Comment, Data, Insight

I am a big fan of data and information.

I am a firm believer it has incredible power inherently within it.

I am absolutely convinced – mainly because I’ve seen it and been a part of it – that when used correctly, it can liberate opportunity.

But despite this, I am also of the opinion it’s current position on the pedestal, is misguided.

To be honest, the reason for this is more because of how people are using it – or choosing what data they think is important – than the actual data itself, but like most things in life, when people look for convenient answers and rely on individuals who have either limited or myopic experience in their field for understanding what’s happening, the results can end up killing potential rather than releasing it.

But that’s just my opinion and I’d be interested in hearing yours.

Especially George, Pete, Baz and Lee … who I know interact with it at the highest levels and with the most qualified of all people.

Over to you …

20 Comments so far
Leave a comment

It’s memorial day. No one needs to think about things like this on memorial day.

Comment by DH

I’m proud of America giving us a holiday you haven’t had.

Comment by DH

Bank Holiday in UK – so Rob is the only one here not working when he should be, He wins again!

Comment by John

So I’m missing out on 2 holidays? I will have to sort that out immediately.

Comment by Rob

youre slipping campbell. better up your fucking game.

Comment by andy@cynic

Big tits not big data.

Comment by Billy Whizz

“Especially George, Pete, Baz and Lee” seems like biased data selection to me. What grades did they get on their statistics tests?

Comment by John

I’m being facetious of course, but my big problem with all the talk about data is that it all too often comes from marketing departments and ad agencies populated by people who studied english or history. Nothing wrong with that but the dodginess of effectiveness papers speaks to the pervading lack of numeracy.

Comment by John

Of course I’m being biased based on the simple fact they work in companies and are in positions where they have access to the sort of data most people who bang on about’ big data’ can only dream about.

As for their qualifications to make an educated comment, well you’re right, it’s based more on their position and the duration of that position than anything more substantial but in George’s case, I know he studied super-human mathematics at some ponce university so he probably has a better view than most. Though as you pointed out, it’s a national holiday everywhere in the World [except bloody China] so this post is totally redundant as everyone enjoys a day out doing the opposite of deck writing – which is what my day has planned for me.


Comment by Rob

wouldnt trust any of them to make me a grilled cheese sandwich, forget their fucking maths ability. i like commenting when all you fuckers have gone, it lets me spread the insulting. give you all a fair go. you all fucking deserve it.

lee. throw in a couple of biz class tix and ill trust you to make me a grilled cheese.

Comment by andy@cynic

no fucker reads campbells bollocks so theres no point linking to yours.

Comment by andy@cynic

Data as you say is a good thing. However I see three main problems.

-Accuracy of data. If it’s not explicitly requested, error rates are high (50%). This leads to the garbage in garbage out problem.
-Reputational risks of holding data. With hackers everywhere, the more you hold, the more you’re exposed.
-Chilling effects of data. Survey after survey points to people modifying their behaviour as a result of increased surveillance.

It seems overall that there is a view from big data booster that given enough data, you’ll have perfectly targeted advertising for products that everyone wants to buy. The problem with this is people aren’t econometric models and constructing the data model required to approach this involves corporations constructing surveillance states.

More data, yes. But the right data and think critically about the costs of collection – including the externalities.

Comment by Neil Major (@Neilmajor)

Thank you Neil … and thank you Ballantine70, both for your comments and saying something that is on-subject, that is very rare for this blog.

Comment by Rob

you fucking tried to slip this serious shit in when i wasnt looking didnt you campbell. well it didnt work because i caught you red fucking handed. dont blame you mind, its a shit subject discussed by middle management to look like they have power when all they are doing is writing powerpoint decks to the boss full of buzzword shit they dont understand. and george, baz and mob would definitely know more about it than most because theyre senior middle management and get to choose the font of the deck. and its never fucking impact. take note campbell and let me never see you try and slip this sort of shit in your blog again.

Comment by andy@cynic

You were missed.

Comment by DH

Andy’s comment has made me too frightened to make a comment.

Comment by Pete

first sensible shit youve ever done.

Comment by andy@cynic

Very missed.

Comment by DH

At my job we have access to a lot of data on the activities of people – not their usage of various websites or their searches, but on their human activities of dealing with their families, their job and their friends. And time and time again, because the data isn’t consistent behind how it is generated, the same activity by 10 different people will lead to data that would suggest at least 6 different activities actually took place. So, to Neil’s point, the accuracy of data is tremendously bad, and when you just have more data it doesn’t mean that the answers will be “more” right, in fact, it can lead to very bad results. But people don’t want to hear that, they want to believe that with more data comes more specific answers, and they’ll twist the data to make it fit the answer they want time and time again. And, because the data is so inaccurate, they can…

Comment by Tim

Leave a Reply