Business Intelligence - Failure is not an option

18 January 2012

Alan Martin

Alan Martin

Business Systems Technical Manager

There is, I am told, an estimated zettabyte (that's 1021 bytes if you are remotely interested) of data swirling around in cyberspace. A lot of this is a result of the rapid uptake in social media ably supported by advances in personal hi-tech devices: smart phones, tablets, clay tablets, carrier pigeons….well okay, maybe not clay tablets. 

Corporate entities also have an ever-growing morass of data in the bowels of their organisations forcing them to find a way of acquiring, sifting, analysing, processing and generally play alchemist to try to turn data into jolly useful information and that, as we all have been told, is power.

So it is just as well, that we have a set of methods, tools and processes called Business Intelligence (BI) and a host of expensive consultants to advise us all how to develop strategies to make all this happen.

The only problem is that over 70% of BI strategies - or any strategy come to that -either fail completely or fail to deliver against expectations

Why that should be so is a topic in its own right, or even a book - ah, already I see a Pulitzer Prize with my name on it.

Any strategy is roughly 5% cerebral and 95% implementation. Get the first part wrong and you are either wasting money or finding yourself alone in the CEO's office with the clear understanding as to which drawer the revolver is sitting in.

So given such a high rate of failure, are there ways of avoiding being in that 70%?

There are certainly things you can do to increase your chances of success and, for what it is worth, here is a whirlwind tour of the less cluttered part of my mind and opining on this particular subject.

Data Governance

Data really is the blood of your company. Without quality data there is can be no quality information and hence decision making is likely to be based on questionable facts.

Governance of data i.e. actually having and enforcing an information policy will ensure your business keeps data accurately, secure, preserves meaning and has robust health. Fact based decision making seems preferable, and rather useful, than having to make assumptions, unless of course you are an economist, politician or like trying driving around cities using a map drawn on the back of a napkin.

Ownership

You need a god - or at the very least, a demigod with a lot of thunder bolts - to sponsor your BI strategy.

BI must be led by the business for the benefit of business. Simply leaving your BI strategy solely to IT is akin to asking them to push big rocks up a hill; you just know there is a broken toe or foot just around the corner.

It might be funny to watch but just think of the rise in your insurance premiums.

Align objectives

Goes without saying really, if BI isn't aligned to your overarching business objectives then failure is pretty much guaranteed.

This is another valid reason why BI must be business led.

Avoid conflict

Business groups have their own aspirations and generally try to steer things the way they want them. There are many reasons why they do this: ambitions of middle managers, job protection (well spreadsheets rule don't they?), a Freudian need to prove their naughty bits are bigger than. …..well you get the drift...

...the only solution to all this, is strong leadership

Powerful leaders are needed to cut across the schoolyard politics and keep the strategy aligned. Therefore, think of getting a leader who is an empowering despot, only without the desire to physically go out and shoot people.

Goal setting

Whilst business units take on an identity of their own, individuals can easily stray from the path of righteousness.

SMART goal setting has been debunked repeatedly (seriously, there is almost zero evidence this works with people who actually possess a brain) and yet we persist - love you really HR …honest. The pursuit of individual goals can run contrary to your plans.

Make individuals goals interdependent to develop a common purpose around your BI strategy. You will quickly identify those not playing nicely and correctional procedures using rats and glass tubes initiated….only kidding.

Skills

Hire a data modeller. Get someone who dreams about data, buys Data Analytics monthly and has pictures of Ted Codd, Chris Date and Ralph Kimball on their bedroom wall.

Don't skimp here because a well designed model is going to pay you back in spades. Good BI specialists will help your business to learn to self-serve and get the most out of their data.

The power of BI has never lain in producing nice dashboards and well-formatted reports, it has been the ability to ask the 'What' and 'Why' questions. Those questions are best asked directly against the data - or you can persist with more spreadsheets and reports than Millwall football club have supporters, your choice.

Technology

If you did the grey cerebral thingy right, then you know that you will need technology that will grow with your company. At the end of the day how quickly you can process data and serve it up to the business, will directly affect how useful your shiny new BI systems will be.

Ideally, you want to try to get all pertinent data in one place and yes that kind of means keeping it close to the business and not in an offshore account in the Cayman Islands. If it takes 28 hours to process information required by the board at the start of each day then quite simple, Houston we have a problem.

Technology is always advancing, do not be too distracted by shiny things, you can always add them on later.

Flexibility

Look at how other data sources can help improve the value of your data. A single tweet can, quite literally, change customer preferences in the blink of an electronic eye - think footballers woodenly telling the tribal amongst us how great the latest X-Box football game is -  gee better go buy that if ultra-rich footballers play it.

So how flexible is your strategy?

Can you include mining the murky depths of Facebook, LinkedIn, MySpace or Google +? Are you able to incorporate weather patterns, footfall and other consumer related data?

Bringing this type of data into your super new data mart/warehouse/filing cabinet,  will help you gain better insights into your customers preferences, how emerging social trends might affect your product sales and how trashed your sixteen year old daughter got at that party you didn't know about.

You need to be able to react to consumer trends; you snooze you lose - guess I lose every between 10pm and 6am every day then…bummer.

Whatever strategy you have, it needs to be flexible enough to bring in data that will enrich the information available to the business and that, as I have stated, means better decision making.

Remember...

...a BI strategy is not just for Christmas, implemented successfully it can last at least a year. Get it right and you can sit back, take the plaudits and put your feet up.

Of course, all your efforts will ultimately be rendered useless: on the 23rd December 2012, the Mesoamerican long count calendar ends and with it the world - apparently.

I will of course be holding a post apocalyptic party on the 24th; I'll send invites and see how you got on.

--------------------------------------------------------------------------------------------------------------------

Amateur psychologist, rational thinker, one time Enterprise Architect/Data Architect, Development/Support Manager, Database Manager and lastly BI manager. A Social Economics graduate from a backwater education facility, Executive mentor come coach, a Humanist, anti-IT IT professional, all round sceptic (skeptic for our American cousins) and a scouser with a world renowned collection of hub caps - @MARTI_AD

Disclaimer: The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of Bluefin Solutions Ltd or Camelot UK Lotteries Limited.

 

 

View comments

Comments

Blog post currently doesn't have any comments.

About the author

Alan Martin

Business Systems Technical Manager

Bluefin and SAP S/4HANA - welcome to the one horse race

We use cookies to provide you with the best browsing experience. By continuing to use this site you agree to our use of cookies.