Thursday, May 21, 2009

The 10 principles of BI Success

Introduction:
Business Value Attainment 2:

2009 is a landmark year for all DW and BI Professionals, agnostic of technology platforms. On one side of the equation we have a deep recession on finances, corporate expenses and overall project expenses. One the other side we have 1,500 CEO’s worldwide that have said BI is their # 1 priority for the 2009 -12 period. Gartner Feb 2009
(source: http://www.gartner.com/it/page.jsp?id=636310)

BI was #2 in Jan 2007, it shifted to #5 by August 2008 and has resurfaced as #1 in Feb 2009.

The reason it was high until 2007 is that management thought BI imperative to the overall decision process.The reason it declined in 2008 is that numerous project did not provide anticipated BVA (Business Value Attainment)

It is now our ownership to ensure that we redefine what project success means and then extrapolate that to BI Projects.

One thing is certain we have moved away from the traditional ‘Build mode of ‘Fire, Ready, Did we Aim’ methodology towards a sustainable planning mode.

My hope is that 2009-10 will delve a lot into introspection.
Question like ROI and TCO will become paramount. BI Business Value audits, Checklists and foundation layers will become imperative. Though this sounds like common sense but we find it still uncommon by rating the customer satisfaction 4-6 months after go live.

We may have to revert to the traditional method of success – plan. ‘Plan your work and only then work your plan’ as our motto for 2009.

This note highlights the strategic, and success, components of a Fundamental Principles of BI Projects.

What are the lessons we have learned either from your implementation or from that of others? Failures should catalyze evolutionary knowledge, and successes create proven structures and processes as wisdom. The question is: Are we learning from our failures, and sharing our wisdom, professionally and not politically, to assure future successes?

Start now! Because it’s all about ‘ethics’ and ‘business value’, ‘standards’ and ‘processes’, ‘people’, ‘communications’, ‘quality’ and ‘governance’ we need to consider the solution holistically with a sound business strategic goal in planning.

All BI Projects end on a binary outcome – either they work or they do not.
Either it provides business value, the measure of which is business user satisfaction, or it do not.

Everything else, as far as these Principles is herein defined as ‘playing the game’
Ethics remain the foundation of this Principle.
Recognize short term 3rd Party providers for they might simply work on a principle of ‘Play the Game’.
Long term partners work with standards and processes and then simply govern them. Experience must be empowered to trust their own intuition more than the recommendations of profiteering 3rd Party.

Principle 1: BI must only be about Business‘’
Two thing lacking in some BI implementations are business and Intelligence’’ Alex Paleologoes, BI DataBridge Europe.
Great BI can only be accomplished by teaming with business to meet business needs. Business Value cannot be attained by keeping business in isolation.
Action: Let business take ownership, or participative ownership, of all BI projects. SOX mandates business ownership and accountability of all BI. Don’t let your vendor recommendation or beliefs steal this fundamental right from business. Someone in the company has to be accountable for business value assurance of BI projects – who better than business itself.

Principle 2: Successful business is all about People & process
Employees are at the root of all successful companies and thus BI projects. A single hour spent with business will assures more success, than the same many hours spent with concepts, apologies, ideas or recommendations. Great BI projects keep business and data as the center of their universe; average BI keeps business in the loop, and bad BI keep business and people out of the BI decision process.
Action: Trust the people that have made your company great. Trust your advisors only if your business owners fully trust and sponsor new ideas without constraints or boundaries. If business is the final ‘customer’ then let them take ownership and accountability of their BI. Build accountability if new ideas fail to add strategic business value.

Principle 3: Data Governance [1] is not possible without Data Quality
Structure defines the path to success, standards the boundaries, and processes the methodology, Data Quality the foundation of Data Governance. Only after these four are established can any form of governance take place.Great BI’s have complex layered architecture consisting of technical and business layers. There are tens of millions of data elements in any data warehouse. 1% data cleansing cannot provide global data quality. Data governance driven by MDM[2] is like monitoring only state leaders, in the fight of global terrorism, and letting all other passengers pass through unchecked.
Action: Get a Global Data Quality profiler and score every data element entering your data warehouse. This is the only way to lay the foundation of Data Governance in BI. Measure every sucker that enters your data warehouse, not just a handful of them. [1] Gartner, Forrester, TDWI and all advisors are recommending Data Governance, however currently this is limited to Master data elements only. Researching all providers we find all of them center around MDM elements only. [2]Total DQM, i.e. scoring and profiling every data element, remains the holy grail in SAP and more so in SAP BI. Data profiling is the foundation of Data qualifying, and data scoring is the key to assuring data quality, thus data quality is the foundation of Data Governance. What you can’t measure you cannot improve’ , stands true in data quality. If you do not have rules there is nothing to govern, stands true to data governance. According to SAP in January 2006 there are less than 300 MDM elements in all of SAP. The average cost of cleansing one to two of these, i.e. customer and vendor, normally runs some pretty high budgets.

Principle 4: Plan your work and only then work your plan
Very common saying, very uncommonly usage. Makes one believe that sometimes common sense can be most uncommon. All plans must be based on structures, methodologies, processes, standards and best practices. Processes must be measured, Principles governed. Data must be profiled and scored prior to letting it enter the data warehouse. Without structures and processes chaos becomes the only available option
Action: Immediately install structure, standards and processes and then follow the plan. Baseline on proven success, improve it as you go along

Principle 5: You’re not dumb. They’re not smart
3rd Parties can sometime fool companies with their jargons and complex presentations. The fact is that your business owners know, best of all, what is good for your business should be leveraged in every BI process or deliverable. Vendor recommendations must be aligned and filtered with business needs. Business must take the final fine-tuning of their analytics Final ownership must lie with your business owners. Never let your business become isolated from what is being done on their behalf. Keeping business out of any BI process is a predictable recipe to disaster - some enterprises learn early others as they mature with the technology.
Action: Business must be a willing sponsor. They must own and be accountable for all BI deliverables. When in doubt let business make the final call.

Principle 6: Control is but a momentary illusion
We have seen numerous BI implementations where the vendor controls all BI deliverables, in some worst case scenarios even prohibits business from participating in what they will finally receive. There must be no single person in any Company that believes in thrusting BI on business users as the right way to implement BI. Let 3rd Party build relationships of trust with the business, rather than a protocol of control and instruct. Control destroys trust and eliminates ownership.
Action: Business based BI relies on business leadership and trust rather than on Vendor recommendations and control

Principle 7: Data is complex, requirements confusing, still must deliver to business needs
Most companies that do not involve business early in the BI ownership normally land up in a situation ‘..this is now so complex, I don’t know where to start..’. Business does not march along nice straight lines. It meanders. Organizations are organic and grow organically. Your project must follow structure and standards in order to assure success. Stop fighting structure and eliminate ‘Quick-Fix’ options.
Action: Train business on BW checks and balances if you are new to BW. Get a mentor for your Operational management so they can become intelligent sponsorsInvolve business very early in BI implementations, and keep them integrated into the BI deployment process to assure success. Most projects that keep business isolated from BI end up with data and numbers but very little business relevant information. Build Data Warehouses and not report marts by planned Architecture and modeling checks along the way

Principle 8: There can be no secrets
If you think BI can be implemented in isolation and in a secretive way then you are only committing self cannibalism. This is akin to a cancerous methodology of BI implementation. Is most such cases the only predictable victim is the BI project itself. A project that keeps secrets loose respect of others and soon sponsor’s budgets and then success itself.
Action: View issues with a magnifying glass and successes with a telescope. Align all things near with goals of far away. Ensure that there is direct contact between business owners, users and the BW developers. Do not let any process or person insulate the key decision makers.

Principle 9: Always trust your ‘blink’ factor
‘Blink’[3] is the concept of first instinct feeling. This can professionally be eroded by sales experts and convincing arguments of big 3rd Party presentations. At the end of the day do not buy into any concept unless you personally feel the ‘Blink’ factor, i.e. personal warm-and-fuzzy feeling. Trust your instincts and those of your business owners as respect their initial concerns and negative feedback.
Action: Buy only into concepts that the company and business owners are personally confident with, and are willing to take ownership from its initial proposal. If unable to take ownership, either disassociate yourself from the process or find a mentor who will guide you through the maze or confusion and thereby enable you weed out the wool and direct you towards true Business Value Attainment (BVA). After each project add a BVA score to success measures that must be filled by business owners three to four months after go live. [3] Blink: Malcolm Gladwell (http://www.amazon.com/Blink-Power-Thinking-Without/dp/0316172324#reader) ‘The statue that did’nt look right and other cases.

Principle 10: The only constant is change
As Einstein confirmed and with everything being dynamic we must always question every Principle here. With proven expertise new BI concepts must constantly be implemented. Scorecards and metrics are pure example of change. With emerging BI technology complex decision methods are being enhanced periodically. No decision must isolate new conceptsat the cost of business familiarity. The other way, if proven, often just works fine.
Action: Special functionalities require special assistance. Do not let 3rd Parties convince you of their version of truth. Never change the terminologies, i.e. a COE can only be that and nothing else, as you change the name so does its essence. Small minds encourage re-titles as a form of originality, only to loose the very essence of the roiginal concept in a very short time. Hire the concept expert, where possible, to provide ‘Value Drivers’ to assure concept success every time. Do not waste time with any form of apologies.

Take decisions that assure high BVA.

No comments: