Association Insight International & European | Page 22
Association Insights | Expert Briefing
STAGE 2 – ASSESS CURRENT PROBLEMS
AND THEIR IMPACT
You now need to put some flesh on the bones by delving
into specific CRM data problems and their causes and
impacts. As you do so create and maintain a log of specific
problems.
An effective way of tackling this stage is to:
•P
rofile and analyse the data currently held within your
CRM system(s) to assess the scope and scale of specific
data problems. For example, how many addresses do
not contain valid postcodes? How many customers
have valid email addresses? And so on. Specialised data
profiling tools are available to help accelerate this work,
but eyeballing the data on spreadsheets can often be an
effective starting point.
•O
rganise interviews and workshops with stakeholders to
present the issues that you have found and use these to
explore questions such as:
-W
hat is the impact of these problems on customers,
suppliers, the business and IT?
-W
hat problems are most pressing? Which cause the
Association the greatest pain?
-W
hy might they be occurring? What are the root causes
of the problems? Process failures? Human errors? Poor
training? IT faults?
-Who is responsible for the errors and who should address
them?
•U
se these to home in on the problems that really matter.
You cannot solve all the data issues you will come across.
Remember the 80-20 rule. What’s important is that you
start to tackle the ones that really matter in the eyes of
your major stakeholders.
•O
nce you’ve got a good understanding of the core issues,
revisit and enhance the business case to start to tackle
them, and acquire the resources you need.
STAGE 3 – DEFINE DATA FITNESS FOR
PURPOSE
Here you now need to focus on the one or more key data
quality problems that have emerged from the previous
stage. For each of these problem areas:
•D
etermine the really important business data items &
records that need to be improved, for example email
address, customer contact address, company billing
address, product code etc. Again the 80-20 rule applies.
Focus on the 20% of data items that cause you 80% of
your data problems.
•W
ork with the business and IT stakeholders identified
earlier to define these items in clear business terms so
that all can understand and relate to them. Then set the
required standard for the data item or record and create
& enforce the business rules that control both their format
and content. For example, what does a customer record
need to contain to be mailable? What should a valid
telephone number look like and contain before use in a
telemarketing campaign?
22 | © Associations Network 2015
•S
et realistic and incremental improvement targets,
based on the current levels of adherence of the data
to the standards laid out. Remember that data quality
enhancement is a continuous process of improvement,
not a one off project, so aim for gradual change. For
instance, “Reduce returned mailings by 15% per annum
for the next three years”; “Ensure that 85% of customer
addresses have valid postcodes by April 2016 and 95% by
April 2017”and so on.
•F
inally it’s vital to put some measures in place to baseline
the current state of the data and to track changes. Ensure
these measures are updated on a regular basis and report
on them via a dashboard or similar tool to show that
improvement is happening. Wherever possible include a
business impact measure, for example each 1% reduction
in returned mailings saves your Association £1,000. This
helps to ensure that the benefits of the improvement work
are quantified and demonstrates the business value of
the work. Try to get the key measures reported to your
Associations Board or Executive Team to maintain their
interest and awareness. Use your senior champion to
facilitate this.
STAGE 4 – DESIGN THE IMPROVEMENTS:
PEOPLE, PROCESS & TECHNOLOGY
For each of the data problem areas identified, you then
need to undertake an improvement project. Note that as
data quality improvement is a business challenge, these
business change projects require business leadership, with
IT in a supporting role.
Important to note here is:
•A
s indicated earlier, data quality problems emanate from
both business and IT issues. So designing improvements
requires a holistic approach, usually encompassing
business process change, people education and training,
and new or amended IT. In some cases problems can be
improved without the need for new or enhanced IT. The
key point is that they are almost never an IT only problem,
so don’t consider them as such.
•A
s a general rule, ensure that the improvements you
design are sustainable. Do not settle for a one off data
cleansing exercise on its own as it never solves the
underlying problems. In a year’s time another cleanse will
be n eeded, then another and another. Find permanent
fixes that will continue to reap benefits after the project
has ended. Ensure these improvements become an
integral part of your business as usual processes.
• If you budget allows for it, look to acquire specialised data
tools to support the work. These tools can be used to:
- Profile & analyse data
- Create and manage data improvement workflows
-S
et up data glossaries and dictionaries to hold data
definitions and business rules
- Correct data defects
- Enrich existing data
www.associationsnetwork.org