|
Blessed
are the Geeks, for they shall internet the earth |
The Next Big Thing: Just-In-Time Data Warehouses
Greg Bromage


It’s rare when the IT
industry lags behind the other, more traditional business models. But one such
case has occurred, and it provides an opportunity for business to save
millions of dollars of needed expense.
In the past, supply chain
managers have used warehouses to maintain their stock. Many items would sit on
shelves, lifeless “potentials” until a need arose. It represented a cost
to the business, however when the need for a particular item arose, the lead
time to supply it was much shorter. The client gave the specifications, the
warehouse staff found the item (or the closest approximation) and provided it.
A simple system; and one
which the IT industry has learnt from. We created the concept of a “data”
warehouse, where huge amounts of data could sit, waiting for Management, Data
Miners or Business Analysts to ask.
With the advances in IT and
communications, however, the supply chain people realized the problem.
Warehouses cost a lot to maintain, and stock (which may end up being never
being used) piles up, gets lost, or is in the wrong place. With the help of
the IT professionals, they developed the concept of Just-In-Time supply
chains, where they can eliminate the warehouse and just take the customer’s
order and manufacture just the required quantity and specifications. It
reduces inventory and staff time, and still keeps the supply chain short.
This is what the IT
departments have not yet realized. The concept of a data warehouse was
originally intended for business intelligence. The data does nothing by
itself; it sits there until some asks a question or has an idea to improve the
company. They can then interrogate the databases and extract the data they
need to justify their business case.
Since the client is coming in
with a preconceived idea of what data they want to see, it is far easier to
dispense with the data warehouse entirely. Eliminating the costs of the
hardware, software, staff, backups and communications systems associated with
a data warehouse generates significant cost savings to a business.
Staff can simply make
requests to “virtual” warehouse, possibly operated by a suitably qualified
call centre in
India
, and the required data can be fabricated to exactly meet the expectations and
requirements of the client.
I have decided to call this
concept the “Just-In-Time Data Warehouse) (J.I.T.D.W.), and realization of
this concept has the potential to save millions of dollars.
There is, of course, a
possibility that the data provided may not always reflect reality. However, it
is a well documented statistic that 80% of business projects are considered
failures, for various reasons. Therefore, if 80% will fail anyway, the
accuracy of the data is of no consequence. With the remaining 20% of projects
that could potentially succeed, if we assume that the data manufactured has a
50-50 chance of being right, the end result as a worst case would be a 10%
attrition rate of business cases which could fail because of faulty data.
Given the huge costs of
setting up and maintaining the data warehouse, the cost savings realized will
usually offset the costs of the failed project, which translated to a high
return on investment for opting for the JITDW.
|