Implementing business intelligence-whitepaper

Implementing Business Intelligence:
Why You Should Build BI Before You Clean Your Data
DATA QUALITY
DATA
MANAGEMENT
BUSINESS
PROCESS
MANAGEMENT
RISK
MANAGEMENT

Jet Reports | 10450 SW Nimbus Ave | Suite B | Portland, OR | 97223 | www.jetreports.com Pg. 2
Implementing Business Intelligence: Why You Should Build BI Before You Clean Your Data | Jet Reports
In this White Paper
Overview Pg. 3
Wait, wait, what is Data Governance again? Pg. 4
So, how do I become the Governor? Pg. 5
4 Steps to implementing Data Governance Pg. 6
Data Governance – not what it appears to be? Pg. 10
Back to Business Intelligence Pg. 11
About the Authors Pg. 11
DATA QUALITY
DATA
MANAGEMENT
BUSINESS
PROCESS
MANAGEMENT
RISK
MANAGEMENT

Implementing a Business Intelligence environment against
data you don’t trust? Sounds counter-intuitive, some
would say: ‘just plain wrong.’ How could you possibly build
a Data Warehouse and cubes on data that might not be
accurate? You shouldn’t of course, and that is the point.
Accurate data is the non-negotiable cornerstone of any
Business Intelligence effort.
But accurate data does not happen by itself. Data needs to
be enthusiastically governed to a state of truth and reli-
ability, and this is a big job. So big in fact, that attempting
a global data governance initiative can cost so much time
and energy as to obliterate the initiatives it was designed
to support in the first place.
If your company has an urgent need to manage three or
four KPI’s around inventory, the six months that it might
take to understand your inventory data and make sure it is
correct can seem like a very long time.
But there is an even bigger risk. Think about how you find
out that you have bad data. It is almost never by putting
data into a system. You only find out about bad data when
you try to get data out. So this begs the question – if you
are going to embark on a data governance initiative before
you roll out BI, what exactly are you going to govern? How
do you know what to fix, if you can’t see that it is broken?
Nothing highlights inconsistencies in your data like trying
to build a dashboard against it. Nothing will focus data
governance efforts like clearly understanding the num-
bers you need to drive your business. Not just the num-
bers themselves, but what are the components of those
numbers? What were the transactions that placed those
numbers into the system? What are the processes around
those transactions? Are they contributing to accuracy, or is
there something buried in there that is making the num-
bers faulty?
There is magic in seeing those faulty numbers appear in
an analysis because now you have a clue about what data
is broken. BI will give you an effective data governance
roadmap, with the starting point in the form of numbers
that your subject matter experts and other stakeholders
will recognize as wrong. It is the BI environment itself that
makes your numbers accessible. And with that accessibili-
ty, data quality issues become glaring, and the root causes
behind them become easier to trace.
This whitepaper is going to discuss
some specifics around that concept,
but first it is going to give some
practical advice around
data governance itself.
Accurate data is the non-negotiable cornerstone
of any Business Intelligence effort.

Wait, wait, what is Data Governance again?
Data governance is a discipline that embodies a convergence of data quality,
data management, data policies, business process management, and risk man-
agement, all in the context of a company’s data.
The stripped down quantitative definition is that data governance is simply a set
of processes that ensures important data assets are managed properly through-
out the enterprise. Qualitatively, it ensures that data can be trusted and that peo-
ple are accountable for anything bad that happens because of low data quality.
Note the important words: ‘people’ and ‘process’ – it’s about putting individuals
in charge of creating processes that prevent data issues, and fixing them when
they inevitably occur.
Through data governance, organizations are really looking to answer the deeper
systemic questions about people, process and methodology. With those an-
swers, they put themselves in a position to exercise positive control over the
processes and methods they use to handle their data.
Altering the company’s way of thinking and setting up new processes to handle
information is not something that happens quickly. The current methods and
habits that you use in your business took years to develop, so the changes you
are asking for will demand an evolutionary process. Data governance is most ef-
fective when it is treated as an ongoing program with a continuous improvement
cycle. It rarely works when attempted as an ‘all at once / one-time initiative.’

So, how do I become the Governor?
One of the biggest mistakes we see is jumping in too soon without clearly defin-
ing what the company wants to accomplish. It is hard to measure effectiveness
without first outlining the business case and determining where and why to apply
data quality efforts.
The most direct way to identify data quality problems is to tackle them from the
perspective of specific business issues rather than general purpose Data Gover-
nance. Specific issues commonly surface in the context of business processes
– a business process being defined as any activity that spins off metrics. Man-
aging Accounts Receivable, for example, will give you past due amounts, days
outstanding, client payment history, etc. Working backwards from a suspicious
result surfaced by your AR dashboard will make the business case jump out
(we’re acting like a bank for our clients!) and will lead you directly to the underly-
ing issues in your AR process.
The advice below is best read in a ‘think globally, act locally’ context. Rather
than applying these steps to the entire data set, look for one defined problem,
understand the specific areas that need to be addressed and then put the most
pressure on those steps where a clear result can be visualized.
Working backward
from a suspicious
result surfaced by
your AR dashboard
will make the business
case jump out

4 Steps to implementing Data Governance
#1 Identify and Establish the Vision
There is no point in trying to align your people, process and policies unless you
know what you are trying to accomplish. Any vision statement should tie to a
foundation of ‘what we do here’ and ‘how we do it.’ Visions should be clear and
specific. Applying the vision of ‘Being the easiest office supply company to buy
from by offering one step ordering with 100% on-time delivery and prices within
5% of the competition,’ produces some targeted goals for customer facing pro-
cesses:
Increase customer satisfaction and encourage repeat business by
streamlining the order process
► Electronic order form that is sent at client order cycle times
► Collect enough information, but not more than necessary
Increase loyalty by instilling trust
► Collect, record and maintain preferred delivery times and method from
the client
► Ensure delivery information is delivered to the responsible person with
fail safe back up process
Increase quality of corporate revenue by customer retention
► Collect and monitor competitor pricing
► Create a process for adjusting prices based on competitor data
The advice below is best read in a ‘think globally, act locally’ context. Rather than applying these steps to
the entire data set, look for one defined problem, understand the specific areas that need to be addressed
and then put the most pressure on those steps where a clear result can be visualized.

In these examples, the data is not really doing the work. It is what is happening
around the data that is making the difference – the data is a means to an end.
But now that the vision is put in terms of specific goals, the data is working for
you. If you have goals, you can derive metrics to measure them. The data that
you need for these metrics will become evident, as will the processes needed to
collect this data.
Vision drives data governance. For example, our office supply company will want
to monitor their on-time delivery rate. The quickest, most accurate way of doing
this is using a BI solution to aggregate the relevant data from where it currently
resides, attempt to display it in the way you need to see it, and identify what
might be missing. Any gaps or inconstancies in the foundation are quickly recog-
nized. For example, customers requesting custom delivery dates outside of the
typical shipping time frame are currently getting calculated as ‘late’ – skewing
the on-time delivery rate. That can be fixed by a process tweak and by adding a
field in the order form.
You are actually seeing data issues in the form of the end result of your data
collection efforts. This ensures you aren’t left wondering whether the governance
changes you make to support the vision will create accurate results in the end.
Increase loyalty by instilling trust Increase quality of corporate revenue
by customer retention
Increase customer satisfaction and encourage repeat
business by streamlining the order process

#2 Appoint Your House of Representatives
Essentially, Data Governance is all about getting your house in order. Making
sure that the way you handle, store and process data is legal, efficient, and right
for your business. Quite often, individuals with no background or specialty in
data are tasked with the job of managing or processing it. Data governance is a
way of making sure that, despite this, you get the most out of your data assets.
Here are the major roles needed to get data quality embedded in
the organization:
Executive Sponsor – best case is to get one that is as senior as possible, with
as much authority as possible. That needs to be balanced with enthusiasm and
the capacity to do the work. It won’t help if your executive is passionate about
the subject, but too busy to do anything about it or vice versa. Most important
is that the Executive Sponsor has a solid vision of how trusted, secure data will
improve the business.
Data Stewards – Your stewards are your business and IT subject matter
experts who can most effectively translate how your data influences the busi-
ness processes, decisions and interactions most relevant to the organization.
Your business stewards must be IT-savvy and your IT stewards must be busi-
ness-savvy. Both must be strong communicators and facilitators across the part
of the organization they represent. Think of them as part subject matter expert
and part project manager.
Data Governance Driver – Proper data governance is a business process in
itself and someone needs to be there to keep it moving and keep it on track. The
Data Governance Driver coordinates tasks for data stewards, communicates all
decisions made by stewards to relevant stakeholders, and drives ongoing data
auditing and metrics to assess program success.
Solid vision...
Subject matter expert...
Drives program...

The easiest way to identify these roles, and avoid arbitrarily guessing who
should be appointed, is by tracing the data backward. You have defined metrics
to drive your business and you know how the underlying data should behave.
Your candidate pool for the roles above should come from the people who are
most invested in two somewhat disparate things – what the metrics are saying
about the health of the business, and the accuracy of the metrics themselves.
In our office supply company, the VP of Sales is ultimately accountable for reve-
nue from existing clients. She is probably the person who looks at the retention
dashboard the most, and she might be your Executive Sponsor. The Data Stew-
ard might be the person who is actually responsible for the customer retention
process. He will know what activities his team is doing and what data points
result from those activities.
The Data Governance Driver might be the person who is responsible for putting
the data points together to form the metrics. In order to do this effectively, she
will have to understand both the processes that create the data, and the me-
chanics behind how that data comes together.

#3 Build the Process
Funny thing about processes; when they are bad, they usually start being bad
at the very beginning but only get noticed at the very end. That is why bad data
that finally surfaces in a Data Warehouse or a reporting tool is so helpful. You
start tracing the problem from a number that doesn’t make sense on the dash-
board, to the cube, to the Data Warehouse, to the transactional database and
then finally, there it is – bad data as a direct result of the method that placed it in
the database itself.
This is the classic ‘garbage in/garbage out’ syndrome, and it is precisely what a
data governance program is designed to eliminate.
Rob Karel of Informatica summarized three major data lifecycle process stages
in a classic blog post1
–
here is a summary:
Upstream Processes – These are the business processes that capture, create,
import, purchase, transform, or update data and introduce it into your organiza-
tion’s information ecosystem. One of the most common, and toughest to solve,
data governance challenges centers around the reality that those in the organi-
zation responsible for these upstream processes rarely have visibility, or incen-
tive to care, about who is consuming this data downstream, and why. This is
where an influential and senior executive sponsor comes in. Your sponsor must
enforce and evangelize amongst her peers the recommended data capture and
maintenance policies generated by the data governance organization.
Stewardship Processes – Stewardship refers to the actual human powered
activities around defining the data policies, business rules, standards and defini-
tions created and mutually agreed upon by your data governance program. The
activities in question include identification, notification, escalation, and mitigation
of any exceptions to your data rules that require manual intervention to resolve.
It is a living, breathing, ongoing thing, and this might shed some light on the skill
set needed among your Data Stewards.
1
Link to Rob Karel blog: http://blogs.informatica.com/perspectives/2012/08/30/data-governance-sustains-your-data-lifecycle/
Bad data as a
direct result of
bad methods...

Downstream Processes – These are the operational and analytical processes
that consume, protect, archive, purge and otherwise draw insight and value from
data. This is where the users interact with the data – it’s the dashboards, distrib-
uted operational reports and ad hoc reporting capability that your company makes
decisions on. Your executive sponsors will only agree to support changes to your
upstream processes, systems and organizational behaviors if you can deliver sig-
nificant business value and ROI against these downstream processes.
To better understand these abstract processes, let’s apply them to our office sup-
ply company who offers and sells three different colors of laser printer paper. The
controller analyzes the Downstream Process of inventory cost, by analyzing in-
ventory turns through a BI dashboard. This information relies on the Upstream
Processes of a receiving clerk who records when laser printer paper arrives, and
shipping clerks marking when it leaves. When the controller identifies that printer
paper is being ordered and received into the warehouse daily, despite the inven-
tory turn being low, he realizes the Upstream Process isn’t recording the variation
on paper color. While white paper is turning quickly, yellow and blue are sitting in
the warehouse. This discrepancy is communicated by the BI solutions and then
addressed through the Stewardship Process.

#4 Policing the state
Your company already has Data Governance policies in place; you likely just
call them something else. If you have a form in your CRM system that sales-
people use to enter new client data – then that counts. It governs the data input
process. The tricky part is that while this form might be sufficient for the up-
stream process of capturing client data, it might be ‘garbage’ for the downstream
process of profiling, segmenting and marketing to your client base later in the
lifecycle.
There is one big, strategic goal when thinking about building policy – get your
organization to manage data as a true asset. Even if you get one department to
do this, or one business process, you are helping achieve corporate goals.
Here are the common policy areas that will ensure you get started
on the right track:
Data accountability and ownership – You would be surprised at how often
corporate data does not have an owner. ‘We are all responsible for data quality,’
or ‘Data integrity is everybody’s job,’ is not the right answer. If everybody owns
it, then nobody does. It is easy to go wrong here because ownership can quickly
take on a political tone. Ownership should be in place for one reason; to ensure
that the data is being captured to answer business questions.
Organizational roles and responsibilities – Remember the Data Stewards
and Drivers? These policies document and make clear the responsibilities of
them and other dependent stakeholders. I like to think of the RACI matrix when
devising these: who is responsible for the day to day work, and who is ultimately
accountable, etc. If you are not familiar with the RACI matrix, it’s a fascinating
management tool – see the link below.2
2
RACI matrix
Strategic goal;
Manage data as
a the true asset
that it is.

Data capture & validation standards – Stop the “garbage in”! These policies
define minimum required data capture standards, data validation rules, refer-
ence data rules, etc. The goal is to ensure the people, processes and systems
that capture, import, update, transform or purchase critical data do so in a con-
sistent, standardized manner with a focus on quality.
Data access and usage – Usage policies ensure appropriate use of data by
appropriate stakeholders. Limiting access to sensitive or confidential information
ensures regulatory compliance of course, but also ensures optimal use of data
assets. It’s helpful to think of roles when considering data access: salespeople
have certain responsibilities and they need certain data (usually limited) to fulfill
those. Finance folks have different responsibilities and so need different data…
Establish a track record of this type of behavior, and all of the sudden the quality
of your revenue rises because you can predict it:
GOOD CLIENT RECORDS =
INCREASED ABILITY TO KNOW CLIENTS =
INCREASED ABILITY TO SERVICE CLIENTS =
SURPRISED, DELIGHTED AND HAPPY CLIENTS =
INCREASED REVENUE

Data Governance – not what it appears to be?
If you start to think about data governance this way, it quickly changes from
being about data to being about risk. It’s the risk of losing clients because your
accounting people are using bad information to do the billing. It’s about the risk
of missing opportunities because your marketing people are using bad data
to try to understand the client. It’s about general enterprise risk because your
processes and technology are not putting a tall enough hedge around sensitive
client data.
You can quickly appreciate that data governance is not really about the data at
all. The data is actually just records or events that have been recorded – call
them facts stored in a database. Sometimes they are calculations based on a
combination of these facts, but even those are just math, subject to a static set
of rules. They really have no personality or will of their own.
The term ‘bad data’ is actually misleading. The data itself didn’t do anything
wrong. It’s the people, processes and policies around the data that are doing
something wrong. These are the things that have free will, that get broken or that
get changed with unintended consequences based on somebody’s whim.
So instead of asking: How can we fix our bad data? Ask: What capabilities are
we trying to enable for our organization? What decisions are we trying to sup-
port? How can changing the way we initially record the facts about our new
clients make them love us more?
Now we are talking about things that can make a difference! When we ask: What
capabilities do we need? we are really asking a question about our people; what
do our people need to do? How do they need to do it? How can we make our
data work to support them in that mission?
2
RACI matrix

Back to Business Intelligence
A successful BI implementation is often an exercise in setting priorities: ‘we need
to gain visibility and control over our AR balances first.’ ‘Next we need to tackle
inventory, then sales.’ At Jet Reports we believe the best way to deliver results is
to take an iterative approach. ‘Deliver fast and deliver often’ is our motto.
This means it is important to approach each business area as an almost indepen-
dent BI project, tackling each report, KPI, or metric separately. The implications
for data governance are that once you see the data flowing out of the system in
report form, you know what to tackle, piece by piece, in terms of data quality.
This is very hard to do otherwise, where the tendency is to look at data gover-
nance as a whole and try to fix the data in the same holistic way. That becomes
an overwhelming project that can stop other initiatives in their tracks.
It is a regular occurrence that when companies install a BI solution, all the incon-
sistencies in their data are highlighted. A common one is sales orders that are
posted without being attributed to either a customer or a salesperson – revenue
is unaccounted for.
‘How can this be?!!” is the usual refrain. ‘We have business rules around posting
invoices that are supposed to prevent this from happening.’
Whether it was an innocent work-around or a full-on policy breach, something,
somewhere caused those business rules to be violated. What once was a
hard-to-unravel mystery is now easy to spot, and easy to solve.
At Jet Reports, we treat Business Intelligence as a data management project,
and we have helped hundreds of companies make sense of their data by building
an environment that facilitates data integrity and complete visibility at the same
time. We invite you to talk to one of our BI and data fanatics about how can help
your company today.
Remind me why I care?
Let’s get this out of the way. Data Gov-
ernance sounds boring. The words
themselves imply general unpleasant-
ness like rules, regulations, compliance
– a governed state. It’s not as exciting
as a flashy dashboard full of insightful
KPI’s. Why is it important?
Because simply put, nothing, absolutely
nothing works without it. Let’s start with
the most important thing – revenue. No
company survives without it, and you
are probably monitoring it very, very
closely. But, if data around this income
is not right, things start to fail.
Take the direct sales company that re-
lies on repeat business from a loyal set
of clients. If your office supply vendor
kept sending you duplicate bills, and
the pizza place kept delivering to your
angry neighbor, you would do business
elsewhere. Here’s the formula:
Bad client records = inability to
service clients = lost clients =
lost revenue

Your Authors:
Jonathan Oesch
Loves ERP, Loves Data. Loves what happens when data from a transaction processing
system gets turned into usable information that people can rely on; “It’s like that chemistry
lab in college where you walk in with your raw supplies and leave with a perfectly round
rubber ball.”
VP of Business Intelligence Sales at Jet Reports. Speaks, writes, blogs and tweets about BI
strategy – sometimes in his sleep.
jono@jetreports.com
Tara Grant
Obsessed with taking the company wish list and turning it into hard-hitting strategy. Dedicat-
ed her entire career to ridiculously useful reporting and BI solutions. “If you aren’t con-
cerned with being in the business of collecting information for a competitive advantage, you
aren’t really interested in business at all.”
Director of Worldwide Sales (aka Queen of ‘Strategery’) at Jet Reports.
tarag@jetreports.com

Jet Reports provides robust, easy-to-use reporting and Business Intelligence
solutions that empower business users across the globe to make informed
decisions.
► Proven solutions – used by over 100,000 users worldwide.
► Support you can rely on – renowned customer service.
► Empowers users – Training to match every learning style.

Implementing business intelligence-whitepaper

More Related Content

What's hot (20)

Viewers also liked (16)

Similar to Implementing business intelligence-whitepaper (20)

More from Kaizenlogcom (20)

Recently uploaded (20)

Implementing business intelligence-whitepaper