I compared the “Data Management” with a story of Elephant
and Seven Blind Men in my last blog post http://manageyourdata.blogspot.in/2012/09/data-management-elephant-seven-blind-men.html.
I promised to start writing my views about the issues that I heard so far from
my customers, consultants and product companies that I worked with.
The first issue is confusion with the whole terminology.
This seems very basic thing but if all stakeholders in this value chain are not
talking same language, same terms with same meaning the initiative is doomed to
fail.
5 years back in 2007, when I was working on evaluating a “data
mapper” product for a company in California – I realized how complex a
terminolgy issue can be. I was at their office for 2 days, going through
product screens, meeting high value real architects. The product was really
good, nice UI, good functionality – but less appeal in the market. I was
talking to marketing head, and he said – Prashant ask me questions about the
product. Ask me when you seek a functionality in a product, what would you ask
for. And I did that. You know what – the company worked hard based on such
feedback from industry users and within a year launch new version which talks
in terms of users and then bought over by a giant. Value unlocked. That was one
off example.
What were my questions to them ?
1.
How does ETL works in this ?
2.
What is the data profiling mechanism?
3.
Does this tool normalize my data ?
4.
Does this tool classify my data ?
5.
Does this deduplicate ?
6.
Can this gives
me exceptions to rules data sets ?
7.
Can I override tool with my own rules ?
8.
Does this toll enrich my data ?
9.
Can I specify attributes for enrichment ?
10. Can
I have multilevel enrichment ?
11. Can
I map different datasets to one common datasets?
12. Can
this tool produces validation reports with data ?
Few of these seems typical
techno-functional and not so business based questions. But when you Go to market –
who is your customer? For data – mostly its IT people – who are technical. The
business will surely have a different set of questions but they will map to
these invariably. How ?
What a Supply Chain or Procurement
business person will ask typically for ?
1.
How do I cleanse my multilevel redundant BOM
(Bill of Material ) Data ?
2.
How do I generate required spare parts listing
with good data ?
3.
Can I have cleansed, good quality suppliers list
?
4.
Where do I aggregate my suppliers by parents –
to leverage spending patterns ?
5.
How do I search my customer list and aggregate
that ?
6.
Can I have a classified product list – to get
necessary dashboard, inventory or spend amounts ?
Have a close look and you will see
most of it maps to core questions above.
So in next post – I will try to
provide glossary of terminology that invariably and interchangeably used while
discussing Data Management .
As usual – Please keep sending your
feedback & questions.
Prashant Mendki
Pmendki at gmail dot com
Twitter - @pmendki
Comments
Post a Comment