Aug 30, 2012

How to Fail Your Way to BI Success

“Insanity”, according to Albert Einstein, is “doing the same thing over and over again and expecting different results”


With Gartner now reporting that more than 70% of global BI projects will fail in the 2012-14 period, is it time to get really concerned and review the path we are walking on. It is time to check whether you are rushing on a path of insanity, or willing to stop for a while and get your bearings straight.

Decisions taken in a state of fear are reported to consistently choose higher risks, cost and lead to an almost certain failure. BI is no different.

The biggest difference between you and Picasso, or Einstein, Beethoven, Bill Gates or whoever your heroes are, is that they spent lots of time in whatever they did. They spent more time in front of a canvas, or guitar, or computer, working away at applying their minds passionately to specific subject areas and goals. According to Malcolm Gladwell’s book ‘Outliers’, patterns only form when someone has done a similar task for 10,000 hours, which in his score equals to 1. According to his research anyone that has spent less than 10,000 hours in any field should not be trusted as any sort of ‘advisor’.
I have spent over 30,000 hours in BI projects ranging from Oracle, Teradata, Informix and SAP BI (95%) so this paper comes with an outlier score of 3.
Looking back it is the failed BI projects, which I was assigned to, that taught me more than the ones I led and were successful. In one project we scored 102% user satisfaction score. Looking at the last six years close to 70% of the projects I was assigned to were projects either heading, or gone, due south. In flying terminology it is called a nose-dive.

Here is the standard question I asked the CIO, BI managers and technical leads from the client, in such projects and when I undertook a 'Strategic BI Healthcheck'

Me: So, how was/ is your BI project?

..and the answers I have received over the last few years.

Client:
[1] Our BI project was an IT success, and a total business failure;
[2] Our System Integrator is working on a fixed bid project. They were supposed to finish in December, now its June so it’s their problem not ours;
[3] We are only 6 weeks away from go live, and right now if we get 40% of the reports we expect then I will be more than happy;
[4] We are in a Severity 1 status two to three times a day so I don’t have time to answer your question;
[5] we have been clearly told by our SI, that according to best practices we must keep business stakeholders out of the doors of the BI project

In one recent BI Project strategic BI assessment, I found basic BI naming conventions missing. When we asked the customer if they had a BI –Center of Excellence they responded a strong ‘Yes’. Rather surprised, and after taking a very shallow dive we found that their SI was calling their offshore support staff COE – and even had the customer executives convinced that they indeed had an operational COE. So all their BI-COE did was daily load checks and low cost developments in BI from across the globe. Zero business participation. Zero standards and processes.

One thing all these respondents were missing is that today it is possible to not fail.

Most BI projects fail because the client does not even try. Both clients and SI’s seem to get entangled initially in an bright vision of hope, smoke and mirrors, followed by an endless spiral downwards where the fear of failure overturns all conventional logic- as both continue to proceed along a very predictable path of failure.

Currently with a sound global methodology and documented scientific principles we can keep delivering world class BI projects consistently ‘Right every time’. But somehow the choice remains towards a consistent path of failure. For the skeptics, were these statements wrong then 70% of BI projects would not be speeding towards imminent failure even as you read this paper. Unless you think Gartner writes all junk and their analysts smoke strange things in stranger bars. That thought would be self-cannibalistic in its very structure.

So your first step is to undertake a few ‘Acid-Tests’ depending on what stage your BI project is at. However the ‘Acid-Test of BI Success’ is pretty easy and can be conducted by the customer alone. So if your BI is sick then please stop all activities, which I know is almost impossible to imagine, but if your car is heading into an obvious accident situation the advice always is todo a very hard stop. Now that you have stopped, it is time to conduct a serious Strategic Health-check by a qualified BI Business Value Architect.
Should you choose not to do this , then you have personally, or collectively, once again consciously chosen to continue on your predictable path of failure. The solutions, though rarified, are written on the wall, but one has to take time to slow down and at least read it.

Important to remember' It will cost 4 to 60 times more to try and fix a BI fault after go live than during the planning phase' BI Valuenomics 2010

More often than not, the fear of failure is so haunting that most of us continue to do what we have been doing, hoping that somehow the future will change. That, by the way, is walking right into the Einstein’s lunacy definition at the begining of this paper.

The first step is to try and never get your BI initiative on any path to failure.

However, once you think that your BI is on a path to failure – the second step is to accept it. Once you think your project might be on a failure trajectory proceed to write down on a piece of paper ‘I will now face my fears and take the following steps to achieve success ( fill in these blanks). In each step there must be a quantifiable goal that mitigates attribute of the fear of failure)

• Step one:
• Step two:
• Step three:

Now sit back and imagine your life three months from now when you have managed to achieve your goals and actually managed finding a remedial path. When suddenly there is a light at the end of the tunnel. (Feels good, right?) So what's stopping you?

The two greatest reasons for BI failures are [1] Assumption and [2] Fear of Failure

Assumption that your partner is going to provide you all the solutions, that your advisor has over 10,000 hours ensuring their BI project is a success, that if you involve business in your BI project they will cause unnecessary delays and increase costs, that the BI project can crawl speed full-speed-ahead without any formal documented methodology- i.e. a ‘Global BI Cookbook’, etc.. This is like Ready-Fire-Aim kind of a BI project.

Fear of failure is the second biggest obstacles to success and we all have to battle it. What separates success from failure is the ability to accept the fear, find out if you are qualified to find the solution and then applying a scientific method out of the mess with consensus of all key stakeholders. The failure-types prefer to manage in the state of crisis, I’ve met a few that seem to thrive on the daily adrenalin-rush of multiple Sev 1 issues on a daily basis sometime for months on an end. Sev 1 issues should not happen more than a few times a year. While others continue on their current path of no-processes, no-standards, etc. hoping things will somehow change in the future- which they rarely do.

It is critical to realize that the failure of a BI project is the failure of the enterprise, of its assets, time, confidence but most important of all its capability to make decisions (till it is fixed) and least of all that of the SI.

Often I have sat in ‘Lessons learned’ meetings and have found that people who are afraid of failure treat them as mistakes. Mistakes are not failures. Mistakes are actually your path to success, but only if you learn from them and then avoid them the next time. Mistakes look like failure but are actually solutions in disguise. True failure is not recognizing failure in a timely manner, continuing on a path of failure without changing any of the variables, or blaming someone else for the failure. True failure is quitting. Everything else is a path of learning.

If humans never made mistakes evolution and development would stop. In pharmaceutical and R&D mistakes is the path to success. As Bernard Shaw had stated “Show me a man who has never made a mistake, and I’ll show you a fool”.

Anyone who tries will make mistakes and I have made my honest share. However, we must also learn from mistakes. It is ill advised to make the same mistake again and continuously over time and space. In almost every sport it is recommended to firstly go beyond the ‘pain point’ and then go to the edge of failure. This is done to establish a breaking point. The best of best then take this breaking point further and further and succeed where no person has before. Gold medals are neither won by athletes who work within constraints nor by ones who continue to make the same mistakes.

Great players will focus on the times they have failed to remind themselves on how not to add to that number. Lesser athletes will focus on the times they have won and try very hard to shout and promote only the few successes they have had. In my life I have learned a hundred times more in failed projects, and my BI Valuenomics book is all the lessons learned from failure- as the scientific principle of all management established in 1897- that by eliminating all failure points our only option is success. An ‘Acid Test’ for a SI is to ask them how many BI projects they have completed in the last 12 months and then try and speak to each one of them independently. A lot of SI’s have many BI customers but almost none are referenceable.

The Silicon Valley is full of the ‘10th time heroes’, i.e. 1 out of 10 ideas will succeed. The ones who succeed either have uncanny luck or sheer persistence, i.e. the ones who never make the same mistake twice and work as a team. The valley is also full of inventors who have held on to their ideas, not sharing, not teaming and they are left standing where they were 5 to 10 years ago and will probably be right there 10 years ahead.

Successful people look at points of failures, learn from it, establish some form of rule to avoid it the next time and succeed. Failures for them are opportunities to learn. In life one can either lead or follow, and fearful leaders try to do both at the same time and at the most critical moments it is the fear of failure that totally paralyzes their thinking. Henry Ford quoted: "Failure is the opportunity to begin again more intelligently."

So with an established BI failure rate of 70% start with accepting that there is a very high probability in your BI initiative/s will fail unless you take serious review of your situation. Recognize that 70% of the BI projects in the known universe will continue to fail all around . That you now need to eliminate all the failure points – for then failure gets eliminated as an option.
Again one way is read ‘BI Valuenomics’ . Up until mid August, that’s when I read the 2012 Gartner report, I had though this book was a nice read and a sound concept. After reading that report I today firmly believe that companies that do not read this book will have BI projects like dinosaurs and face a probable failure of around 70%. Right now this is the only book one can use as an actionable roadmap out of the Gartner 2012 BI nightmare- bar none.

Your ‘advisors’ are going to bombard you with what to do and how they have done it with other customers. However you need to start reading Gartner reports more carefully, speaking to fellow customers on what they did wrong. Follow this through and hire yourself a reliable ‘BI Business Value Architect’ and learn what not to do from them. Train your business stakeholders and arm them with checklists of failure and success criteria’s, i.e. what to avoid and what to ensure you have.

To be ultra successful you have to be able to find your failure points in your mind, find them often and predictably, and then avoid these pitfalls. If unable to do so accurately find yourself a reliable advisor who can undertake all these tasks on your behalf. This must under no condition replace your business stakeholders, be part of your current BI team, or be someone without solid business background and BI skills to match.

As a takeaway your scientific path to success should be:

1. Do your BI right the first time, every time.

2. Accept Failure is a very high possibility in BI (current at 70%)

3. Develop and get hold of BI Checklists by each phase

4. Anticipate and avoid failures. When failure happens welcome it and learn from it

5. Apply only scientific principle in BI, not because they are handed to you as such but because they are.

6. If you don’t know how find someone who does and has a proven string of successes that are referenceable

1 comment:

  1. I enjoyed reading your article. Please make more interesting topics like this on.
    I'll come back for more :)

    From Japs a researcher from Always Open Commerce

    ReplyDelete