Description
pon Microsoft v1BT Business intelligence
oday’s business environment demands
flexibility – whether it’s to respond to
unexpected events, customer demands,
or both. But to be in a position to make
timely, effective decisions, manufacturers need to be
armed with a panoply of information – and that’s also a
requirement for those grappling with initiatives such as
Six Sigma, Lean’s continuous improvement and
Balanced Scorecard, which not only demand more
information more quickly, but expect it to be available
to many more users throughout the organisation.
Why does this present such a big problem? After all,
the information they need is, these days, all in a com-
puter system somewhere isn’t it? The challenge comes
about because it’s not all in the same system, and in a
manufacturing environment, it’s rare that a meaningful
decision can be taken on data from one source alone.
How, for example, could a production scheduling deci-
sion be made purely on the basis of what’s happening
on the shop floor, without taking into account raw
material stock, product availability from suppliers, work
in progress, the current stock in the distribution chan-
nel, and current customer demand?
As Renaud Besnard, product marketing manager for
Microsoft SQL Server, explains, collecting data from dif-
ferent sources is more easily said than done. “A manu-
facturing operation generates a lot of data through
transactional systems such as ERP, CRM [customer rela-
tionship management] and so on. But the traditional
methodology to deliver value from that data requires a
lot of technology infrastructure, which carries with it a
danger of incompatibility between systems and differ-
ent versions of the truth.”
So the different sources of data need firstly to be
aggregated, then possibly cleaned – customer data
from a call centre may not be quite so neat, tidy or rel-
evant as measurements from a shop floor machine or
lead times and routing data in a production planning
system, for example – and finally transformed, to
ensure that everyone is talking about the same thing.
While that’s not easy, there has been an answer. The
best route to extracting information from a variety of
data in large volumes and from different sources has
been to use business intelligence (BI) software. Besnard
again: “But traditional tools do not make it easy to
explore the data: they are not easy to customise, and are
not easy to upgrade.”
And even where BI tools have been used, they have
often been processing out-of-date information. “The
need to extract data from ERP systems and others, then
process it, next move it into data marts and then finally
transfer it to client BI tools for analysis means that the
information becomes stale. As manufacturers move
more and more to using real-time data, or quasi real-
time data, this is no longer good enough.”
Operational advantages
Further, traditional BI tools have, in the main, been
solely the province of specialist users, rather than some-
thing a shop floor manager, for example, would be able
to use. “BI information is now being asked for by people
such as sales and marketing staff or operations man-
agers, and these people are not necessarily natural born
number crunchers,” says Besnard. “That means the tools
have to be simpler, and more accessible – perhaps
deployed through a portal.”
Rob Gray, product marketing manager for Microsoft
Business Scorecard Manager, adds: “BI tools have also
been expensive, and while that doesn’t stop you deploy-
ing them, you have to ask what ROI [return on invest-
ment] is this going to give me? And, there is the
question of availability. Traditionally, BI tools haven’t
Manufacturing Computer Solutions • March 2006 • www.mcsolutions.co.uk
t
36
Business intelligence
Manufacturers have long recognised the importance of analytics, but now the tools are ready to spread to
operational users throughout the business. Andrew Ward talks to Microsoft about the new strategy
had to be highly available. If you are making one deci-
sion a week it doesn’t matter if the system is down for a
while, but if it’s being used by operational users then if
it is down for 30 seconds that can be critical.”
And the data situation for manufacturing users is
about to get much worse. “A lot of manufacturers are
now starting to use RFID technologies, and this should
be another opportunity for BI,” asserts Besnard. “With
RFID you are getting much more granular information,
whether in the supply chain or on the production floor,
that might be something you can use. BI tools are ideal
to extract and analyse that kind of information to
improve decision making.”
Besnard cites an example of how manufacturers can
expect significant gains from processing this extra level
of information. “For example, a computer hardware
manufacturer with a fairly complex supply chain might
benefit from analysing how long certain items are stat-
ic at certain points in that supply chain. The ROI could
be a reduction of 35% in storage space.”
Microsoft believes that to achieve this kind of bene-
fit across organisations requires a different approach to
BI, getting away from the expensive and specialist tools
of the past. The company believes that in Microsoft SQL
Server 2005 – with BI now an integral part of the plat-
form, not an add-on distinct from the underlying infra-
structure – it has the answer. “BI is now integrated all
the way from the database itself, through BI tools, and
finally to end user tools and the Microsoft Business
Scorecard Manager,” explains Besnard.
“Microsoft SQL Server 2005 includes fairly critical BI
functionality itself – like SQL Server Integration Services
for the extraction, transformation and loading of data.
That’s the tool you would use to get good data from a
large number of different sources, whether from
Microsoft or others,” he continues. “By being integrated
with SQL Server, it is faster and more efficient, able to
handle very large volumes of data, while also being easy
and familiar to use.”
Why faster and easier? Because SQL Server also
includes Analysis Services and Reporting Services – an
unusual and very powerful addition to the database.
Typically, tools and services like this work on an entire-
ly separate copy of the data in, say, a data warehouse,
running analyses on information from batch data trans-
fers. By contrast, SQL Server BI services run concurrent-
ly with SQL Server itself. “This means that data mining,
predictive reports or other analyses are not loaded and
run as separate processes after the event, but can be
dynamically analysing live data,” explains Gray.
Of course, it’s not always practical to attempt to
aggregate and process data live, nor is it always neces-
sary. “For example,” says Besnard, “whether we look at
Microsoft’s monthly revenue reporting yesterday or
today may not make not much difference, but for a
manufacturer there could be a need to update the data
every few minutes. There will always be a trade-off
between real and non real-time data.”
Besnard cites an example from another sector, but
one that many manufacturers will identify with in
terms of the need to respond quickly in today’s business
environment. “In the retail world, sales volumes used
not to be reported in real time, but on a monthly or
quarterly basis, and shops may decide to change their
store layout on the basis of that information. But now
at one big London store they can analyse sales volumes
on each floor and department in real time and also use
predictive analysis. For example, if a restaurant is busy
they know that certain other departments may be busy
immediately afterwards, so they can deploy mobile tills
within minutes so they can improve customer service.”
Manufacturers, and particularly those working on
supply chain initiatives, recognise the growing need for
collaboration. That means a growing demand for data,
reports and analysis for a wider audience than just one
manager. “By integrating BI services with our portal and
collaboration technologies, information can be quickly
shared throughout an organisation,” advises Besnard. ?
Manufacturing Computer Solutions • March 2006 • www.mcsolutions.co.uk 37
finally comes
to the masses
Our Sponsor
Microsoft
Microsoft Campus, Thames Valley Park
Reading RG6 1QG
Tel: 0870 60 10 100
E-mail: [email protected]
web: www.microsoft.com/uk/business
? Enter 420 at www.mcsolutions.co.uk/enquiry
“BI tools have to be simpler, and more
accessible, deployed through a portal”
Renaud Besnard, Microsoft
operational intelligence sponsored by Microsoft
doc_111294955.pdf
pon Microsoft v1BT Business intelligence
oday’s business environment demands
flexibility – whether it’s to respond to
unexpected events, customer demands,
or both. But to be in a position to make
timely, effective decisions, manufacturers need to be
armed with a panoply of information – and that’s also a
requirement for those grappling with initiatives such as
Six Sigma, Lean’s continuous improvement and
Balanced Scorecard, which not only demand more
information more quickly, but expect it to be available
to many more users throughout the organisation.
Why does this present such a big problem? After all,
the information they need is, these days, all in a com-
puter system somewhere isn’t it? The challenge comes
about because it’s not all in the same system, and in a
manufacturing environment, it’s rare that a meaningful
decision can be taken on data from one source alone.
How, for example, could a production scheduling deci-
sion be made purely on the basis of what’s happening
on the shop floor, without taking into account raw
material stock, product availability from suppliers, work
in progress, the current stock in the distribution chan-
nel, and current customer demand?
As Renaud Besnard, product marketing manager for
Microsoft SQL Server, explains, collecting data from dif-
ferent sources is more easily said than done. “A manu-
facturing operation generates a lot of data through
transactional systems such as ERP, CRM [customer rela-
tionship management] and so on. But the traditional
methodology to deliver value from that data requires a
lot of technology infrastructure, which carries with it a
danger of incompatibility between systems and differ-
ent versions of the truth.”
So the different sources of data need firstly to be
aggregated, then possibly cleaned – customer data
from a call centre may not be quite so neat, tidy or rel-
evant as measurements from a shop floor machine or
lead times and routing data in a production planning
system, for example – and finally transformed, to
ensure that everyone is talking about the same thing.
While that’s not easy, there has been an answer. The
best route to extracting information from a variety of
data in large volumes and from different sources has
been to use business intelligence (BI) software. Besnard
again: “But traditional tools do not make it easy to
explore the data: they are not easy to customise, and are
not easy to upgrade.”
And even where BI tools have been used, they have
often been processing out-of-date information. “The
need to extract data from ERP systems and others, then
process it, next move it into data marts and then finally
transfer it to client BI tools for analysis means that the
information becomes stale. As manufacturers move
more and more to using real-time data, or quasi real-
time data, this is no longer good enough.”
Operational advantages
Further, traditional BI tools have, in the main, been
solely the province of specialist users, rather than some-
thing a shop floor manager, for example, would be able
to use. “BI information is now being asked for by people
such as sales and marketing staff or operations man-
agers, and these people are not necessarily natural born
number crunchers,” says Besnard. “That means the tools
have to be simpler, and more accessible – perhaps
deployed through a portal.”
Rob Gray, product marketing manager for Microsoft
Business Scorecard Manager, adds: “BI tools have also
been expensive, and while that doesn’t stop you deploy-
ing them, you have to ask what ROI [return on invest-
ment] is this going to give me? And, there is the
question of availability. Traditionally, BI tools haven’t
Manufacturing Computer Solutions • March 2006 • www.mcsolutions.co.uk
t
36
Business intelligence
Manufacturers have long recognised the importance of analytics, but now the tools are ready to spread to
operational users throughout the business. Andrew Ward talks to Microsoft about the new strategy
had to be highly available. If you are making one deci-
sion a week it doesn’t matter if the system is down for a
while, but if it’s being used by operational users then if
it is down for 30 seconds that can be critical.”
And the data situation for manufacturing users is
about to get much worse. “A lot of manufacturers are
now starting to use RFID technologies, and this should
be another opportunity for BI,” asserts Besnard. “With
RFID you are getting much more granular information,
whether in the supply chain or on the production floor,
that might be something you can use. BI tools are ideal
to extract and analyse that kind of information to
improve decision making.”
Besnard cites an example of how manufacturers can
expect significant gains from processing this extra level
of information. “For example, a computer hardware
manufacturer with a fairly complex supply chain might
benefit from analysing how long certain items are stat-
ic at certain points in that supply chain. The ROI could
be a reduction of 35% in storage space.”
Microsoft believes that to achieve this kind of bene-
fit across organisations requires a different approach to
BI, getting away from the expensive and specialist tools
of the past. The company believes that in Microsoft SQL
Server 2005 – with BI now an integral part of the plat-
form, not an add-on distinct from the underlying infra-
structure – it has the answer. “BI is now integrated all
the way from the database itself, through BI tools, and
finally to end user tools and the Microsoft Business
Scorecard Manager,” explains Besnard.
“Microsoft SQL Server 2005 includes fairly critical BI
functionality itself – like SQL Server Integration Services
for the extraction, transformation and loading of data.
That’s the tool you would use to get good data from a
large number of different sources, whether from
Microsoft or others,” he continues. “By being integrated
with SQL Server, it is faster and more efficient, able to
handle very large volumes of data, while also being easy
and familiar to use.”
Why faster and easier? Because SQL Server also
includes Analysis Services and Reporting Services – an
unusual and very powerful addition to the database.
Typically, tools and services like this work on an entire-
ly separate copy of the data in, say, a data warehouse,
running analyses on information from batch data trans-
fers. By contrast, SQL Server BI services run concurrent-
ly with SQL Server itself. “This means that data mining,
predictive reports or other analyses are not loaded and
run as separate processes after the event, but can be
dynamically analysing live data,” explains Gray.
Of course, it’s not always practical to attempt to
aggregate and process data live, nor is it always neces-
sary. “For example,” says Besnard, “whether we look at
Microsoft’s monthly revenue reporting yesterday or
today may not make not much difference, but for a
manufacturer there could be a need to update the data
every few minutes. There will always be a trade-off
between real and non real-time data.”
Besnard cites an example from another sector, but
one that many manufacturers will identify with in
terms of the need to respond quickly in today’s business
environment. “In the retail world, sales volumes used
not to be reported in real time, but on a monthly or
quarterly basis, and shops may decide to change their
store layout on the basis of that information. But now
at one big London store they can analyse sales volumes
on each floor and department in real time and also use
predictive analysis. For example, if a restaurant is busy
they know that certain other departments may be busy
immediately afterwards, so they can deploy mobile tills
within minutes so they can improve customer service.”
Manufacturers, and particularly those working on
supply chain initiatives, recognise the growing need for
collaboration. That means a growing demand for data,
reports and analysis for a wider audience than just one
manager. “By integrating BI services with our portal and
collaboration technologies, information can be quickly
shared throughout an organisation,” advises Besnard. ?
Manufacturing Computer Solutions • March 2006 • www.mcsolutions.co.uk 37
finally comes
to the masses
Our Sponsor
Microsoft
Microsoft Campus, Thames Valley Park
Reading RG6 1QG
Tel: 0870 60 10 100
E-mail: [email protected]
web: www.microsoft.com/uk/business
? Enter 420 at www.mcsolutions.co.uk/enquiry
“BI tools have to be simpler, and more
accessible, deployed through a portal”
Renaud Besnard, Microsoft
operational intelligence sponsored by Microsoft
doc_111294955.pdf