|
|
|
|
| Management Briefings
|
|
|
|
|
|
'B' is for business: Rick Anderson, A de V (October 2008)
|
|
|
|
|
|
The SAP Business Warehouse (BW) comprises three main elements – data extraction, warehousing and reporting. But which
of these is most useful to the business?
Consider a conventional retail or supermarket warehouse. However well-designed the delivery and storage process is behind
the scenes, ultimately success depends on whether customers buy the products.
Similarly, however elegant is the technical design of your BW-based warehouse, it is the end users and business leaders who
will judge its success. And their evidence for success will come from answers to questions around how many users there are,
what new insights are being gleaned and what decisions are made with the information – basically, are the reports working, not
just technically but in the widest sense?
In line with this, the recent release of BW is called BI: ‘warehouse’ has become ‘intelligence’, but ‘B for business’ is still the
same.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The need for mature migration: Philip Howard, Bloor Research (August 2008)
|
|
|
|
|
|
Data migration is a major part of the data integration business – according to Bloor research,
the market for data migration products, services and staff was significantly in excess of
$5 billion in 2007.
But not only is data migration big business, it is also in need of review. Historically, the task has
been accomplished using conventional data integration and data quality tools, or by means of
hand-coding. However, research undertaken by the Standish Group in 1999 found that more
than 80% of data migration projects ran over time or budget. When Bloor Research conducted a
survey on the same topic in 2007, we found that this figure had not changed.
In other words, seven years of general-purpose enhancements to the tools available made no
difference to the success rate of such projects.
One of the conclusions to be drawn from this is that traditional tools, on their own, are not enough. Companies need data
migration methodologies, they need experienced professionals, and they need special-purpose tools; or at least add-ons to
conventional tools that have been specifically designed to support data migration.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
In search of the truth: Chris Butler, Aspective (June 2008)
|
|
|
|
|
|
As business intelligence continues to feature highly on the IT agenda of companies around the
world, it seems appropriate to look at one of the key tenets of the subject – the single version of
the truth. Is it possible to implement and maintain one without bankrupting your company?
The theory is simple: in order to manage and develop any business you need to understand
how that business is operating, and a key insight into this operation is provided by data. But for
this to be of use, the data presented at all levels of the organisation must be consistent,
allowing all departments to make the right decisions at the right time.
Everyone can tell an anecdote of the meeting from hell where 90% of the time was spent
establishing where the data came from and a mere 10% on making a decision. The single
version of the truth solves this by providing a single trusted source of data for the organisation
from which all decisions can be made.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Getting real: Peter Scott, Rittman Mead Consulting (February 2008)
|
|
|
|
|
|
My car does real-time computing (or at least I hope the engine management, traction control and ABS systems are real time!).
My mobile phone is real time, even some of my home entertainment gadgets are real time – but what about business
intelligence systems?
In the early days of BI (or was it called decision support back then?) reporting data for the current month was an achievement.
As techniques and technologies evolved, companies moved to report on the previous day but were still reliant on an out-ofhours
batch process to move data from the source systems and store and aggregate it in their data warehouses and data
marts.
But with 24-hour business days and the need to report across multiple time-zones, the traditional batch window is being
squeezed into near non-existence. New data extraction and load paradigms have also been developed to trickle-feed reporting
systems. But are these really real time – and does it really matter?
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Three paths to pervasiveness: Vuk Trifkovic, Datamonitor (December 2007)
|
|
|
|
|
|
Business intelligence (BI) has been attracting a lot of attention in the last few months. The
continuing vendor consolidation stepped up a notch as two of the market leaders, Business
Objects and Cognos, were acquired by SAP and IBM respectively. At the same time, Oracle
announced that embedding BI directly in enterprise application transactions is one of the core
design principles behind its forthcoming Oracle Fusion applications.
Large acquisitions are always noteworthy. In this particular case, however, they were not
entirely surprising. They are indicative of broader trends in the BI market identified in
Datamonitor’s recently published report Economic Outlook: Business Intelligence – in particular,
commoditisation and convergence.
Through ongoing technological development, the bulk of traditional BI functionality has become
increasingly commoditised. As a result, competitive differentiation within the BI market is
shifting.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Analyse this, and that: Fern Halper, Hurwitz & Associates (October 2007)
|
|
|
|
|
|
Text analytics has recently burst on the scene as an important technology to help companies understand and gain insight from
their unstructured information. The technology’s value proposition is compelling and companies are starting to sit up and take
notice.
The technology is rapidly moving out of the early adopter stage into the early maturing stage, with vendors reporting growth
rates of 30-50%. But as with any early maturing technology, questions still remain as to exactly what text analytics software is,
how it works, and how companies are using it. Also worth noting are the challenges that early adopters have faced when
implementing analytics applications.
As part of a major research initiative in this area, Hurwitz & Associates conducted interviews with text analytics software
vendors and companies using the technology. We also conducted an online research study in April 2007, involving large
companies across three categories: those who had already deployed the technology, those who were planning to deploy the
technology, and those who had no plans to deploy the technology but were at least somewhat familiar with it.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
MDM: the task ahead: Mike Ferguson, Intelligent Business Strategies (Jul/Aug 07)
|
|
|
|
|
|
Many organisations have recognised that in order to make full use of their information, they need to extract business
intelligence from their whole complement of applications. To do this, the data held across the company has to managed in the
same, standard way. This is known as introducing enterprise-wide master data management (MDM).
In fact, organisations can deploy four kinds of MDM systems – a rules-based MDM synchronisation system, a virtual (aka
registry-based) MDM system, an MDM data hub, and an enterprise MDM system (see ‘Master Data Management: Creating a
Single View of the Business’ by Colin White and Claudia Imhoff at www.beyeresearch.com/study/3360).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Go configure: Keith Inight, Atos Origin (April 2007)
|
|
|
|
|
|
Configuration and asset management is rising up the IT agenda. Why is this? There are
three main drivers:
1. Compliance: including software licensing, where over 30% of companies can expect
to be audited by a vendor each year; security, where it is necessary to show that the
production environment includes the correct versions of all software including patches
and other security requirements; business control, to continuously demonstrate that
business processes are not being impeded by IT failures; and asset reporting, a
laborious but necessary fiscal process.
2. ITIL. The IT Infrastructure Library is becoming a de facto standard – together with
implementation to ISO 20000.
3. Business alignment. Infrastructure consumes over half the IT budget and business managers are increasingly
expected to understand the links between what they pay and the quality of services they receive – involving a complex
relationship of services and IT assets.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The danger that's over-looked: John Morris (December 2006)
|
|
|
|
|
|
The failure rate when companies introduce new software or systems should be enough to make
people stop and think. Some commentators suggest that as many as 80% of IT migration
projects either fail or overrun their planned budgets. Indeed there is an almost fatalistic
resignation to the inevitability of difficulty and failure. When asked about the success of their
latest programme, many a project manager will shrug and say: “It was a success – there were
the usual data glitches of course.”
Does it have to go on being that way? Do our expectations have to be set so low?
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
It's not what you have – it's the way that you flex it: Rupert Booth (Oct 2006)
|
|
|
|
|
|
Consultants’ articles on corporate performance management often begin by advancing the
latest ‘solution’ and then move onto a proposing a set of startling benefits. This article takes a
sceptical approach by considering recent initiatives in CPM and examining the evidence for
claimed benefits, then summarising what has been shown to work – and on this basis indicating
where to invest in solutions and how to prepare a business case.
There are two sources of evidence of the benefits of corporate performance management and
measurement – the more robust is retrospective academic analysis, and the more persuasive is
feedback from executives on their experiences.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Beyond the hype: Simon Bell, BPM Partners (May 2006)
|
|
|
|
|
|
Within the data warehouse and BI marketplaces, some key topics head the list of what is
hot and what is not. Data quality; integration in its various forms; real-time (or maybe
more appropriately right-time) architectures for warehouse delivery; offshoring or
rightshoring - all these trends have received their fair share of column space. But master
data management (MDM) in particular is a topic that has been warming up for a couple
years and is now pretty hot.
The question is: how much of this is hype and simply the vendors of technology and
services looking for the next sale - and how much is valuable reality where managers
should be spending their time investing and developing capability?
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Triple whammy: Chris Howard, TLCC Global (March 2006)
|
|
|
|
|
|
Businesses in the 21st century are under greater pressure than ever to perform, both in terms of competing with their
global rivals and satisfying the demands of their shareholders. Technology plays an ever-increasing part in enabling
those businesses to operate efficiently and effectively – so you would expect it to have a significant role in delivering
corporate business performance (CPM) information.
In many organisations, corporate performance management is indeed undertaken using information technology. But for
many businesses, it’s a labour-intensive, time-consuming, manual operation, often undertaken with spreadsheets. And
due to a lack of common definitions for key business metrics, it can be prone to error or subjective interpretation.
So how can data warehousing and business intelligence (BI) technologies be used to deliver corporate performance
management in an automated manner?
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Dead or alive?: Ian Gotts, Nimbus (December 2005)
|
|
|
|
|
|
Are your managers acting as company doctors or coroners? That may sound like a bizarre question, but ask yourself
whether your executive management, managers and supervisors are using business information delivered through
reports and scorecards to take the pulse of the organisation, or to conduct a post-mortem of last month’s performance.
Put another way, are the key performance indicators (KPIs) they are using to make their decisions ‘leading’ indicators or
‘lagging’ indicators? (see examples in Table 1). Because things can start going wrong in a business well before the
performance measure turns the traffic light on the scorecard red. Using metrics that measure past events is like driving
whilst looking through the rear window. It’s easy not to see an opportunity or threat on the road ahead until you’re upon
it.
So if leading indicators are clearly more valuable than lagging, why do many (most!) projects seem to deliver reports
and scorecards full of lagging indicators?
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
A supply chain of data: Haydn Durrant, Phusis (November 2005)
|
|
|
|
|
|
The return on investment to an organisation from data warehousing lies in the way the business users interpret reports
generated from the data warehouse and from the actions that they subsequently take. It can
therefore be argued that inaccurate reports can be worse than no reports at all because they can mislead business
users, causing them to make poor decisions that result in missed opportunities, wasted resources and lost revenues.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Looking DAPR: Michael Mainelli, Z/Yen (June 2005)
|
|
|
|
|
|
As business gets ever more competitive, finer and finer distinctions in timing and price
make all the difference. Leading firms are increasingly looking at DAPR – dynamic
anomaly pattern response – systems for their new finery:
dynamic – adaptive and learning from new data in real-time; anomaly – identify unusual behaviours; pattern – reinforce successful behaviours; and response – initiate a real-time action.
DAPR systems are being successfully applied in a number of markets, from
manufacturing through logistics to finance. Such systems are essential in dynamic
environments where the ‘rules’ cannot be precisely specified in advance because the environment is ever-changing.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Round the data mountain: Adrian Hepworth, Atos Origin (February 2005)
|
|
|
|
|
|
Most IT systems now in operation have been designed to retain more data than they actually need for their day-to-day
functions. Companies seem to find comfort in retaining the last n+1 transactions when only n is required to perform a
business function. They often justify this as risk reduction, the ‘just in case’ scenarios that no architect can think of when
designing the system.
But how often is the spare capacity that is programmed in actually used? Data warehousing was born from this desire
to keep data beyond its immediate useful life; it allows users to compare a single event with others, so that patterns in
business activity can be observed and analysed.
|
|
|
|
|
|
|
|
|
|
| |
|
|
|
|
|