The explosion of data being managed by organisations is
driving demand for solutions to help them extract value and better
understand their business and customers. But as competition to
extract intelligence from data intensifies, companies will need to
be cleverer about the questions they ask if they are to get useful
answers.
For the past few years, organisations have been struggling to
manage a truly staggering
increase in commercial and corporate data volumes. Now, as it
becomes clearer that there may be money in these bulging
datacentres, something of an arms race is underway to uncover
it.
"There has been an order of magnitude increase in the amount of
data companies are attempting to manage. Now there is a fear that
'if my competitors are doing more with information than me then I'm
at a disadvantage'," says Gartner research director Gareth
Herschel.
The volume of data being generated online alone is encouraging
organisations to get faster and smarter at processing it, while new
and emerging tools increase their chances of finding hitherto
undreamt of connections between things.
The drive to improve customer service has led to call centres
becoming data collection points in their own right as companies try
to match, for instance, conversations with customers with online
behaviour. Advances in
speech recognition technologies are even allowing companies to
throw other information, such as emotional content and triggers,
into the mix.
A business basic
Deloitte strategy consultant Tim Willey says that the field of
data analytics has sprouted from something of a niche industry,
catering to only the most market-savvy companies, to become a core
business function. "We are seeing data analytics employed for real
competitive differentiation rather than merely a back office
function."
The number of companies offering various data analytics services
has mushroomed.
Gartner estimates that over the past five years, the volume of
"structured" data, such as spreadsheets and databases, being
handled by organisations has risen by 220% year on year.
Unstructured, or emergent, data types, such as information on
online behaviour, are growing even faster.
Herschel says companies such as Amazon have seen their total
data volumes increase by around 1,000% year on year, as more and
more detailed information on customers is collected.
Poster child for business intelligence, hotel and casino chain
Harrah's realised
early on that the better it understood its customers the more money
it would potentially make. More than 10 years ago it deployed a
solution from WINNet Systems which allows it to compare data on all
of its customers across all of its businesses.
By analysing information such as gender, age, place of residence
and types of casino games played, Harrah's achieved a $100m
(£60.7m) increase in revenue from customers who visited more than
one property. "It found out that happy customers spend more, and
that different customers will have a different point of unhappiness
compared to others," explains Willey.
Such data modelling has proved to be especially useful for
industries such as insurance, as well as airlines, for which
understanding human behaviour is so critical.
The potential value of customer data is informing the way in
which companies in industries such as retail develop their club
cards and rewards programmes. A few years ago, there was debate
within the retail sector about the value of managing rewards
programmes and club cards.
So successful has UK retail giant
Tesco's Club Card been,
however, that it inspired a best-selling book,
Scoring Points. The retailer has announced plans for a major
investment to relaunch the card in a bid to stave off competition
from rivals, including ASDA and Sainsbury's.
Strategic
planning
Gartner's Herschel feels that data analysis will soon play a
central role in informing the strategies of most significant
organisations. "An inflection point is now being reached. Smart
companies are starting to ask 'what are the wide open spaces we
haven't looked at before?'," he says.
Recent developments in the hardware space are also creating new
possibilities, especially with the emergence of in-memory
analytics. The idea behind this is that all the data resides in the
memory of the server, with queries therefore not requiring any disk
input/output, a development which promises immense increases in
processing speeds. While generating activity among manufacturers,
established business intelligence companies, including SAP and
Microstrategy, are also making serious investments to add in-memory
analytics as part of their offering.
Tracking behaviour online is a key driver for data analytics,
with the perceived potential value of tapping into
Web 2.0 services pushing up demand for technologies that might
tap into and reveal connections within and between "viral networks"
such as
Facebook, Myspace and Twitter.
Behavioural advertising company
Phorm is yet to set a
date for the UK launch of its
online content gatherer Webwise. The idea behind the
application is that regular internet users indicate their various
tastes and preferences, which Webwise uses to present the viewer
with information about things that might interest them, such as
shopping and entertainment. For example, a user who has been
looking at a website advertising digital cameras is afterwards
presented with adverts from companies selling cameras or other
photographic products and services.
The company has come
under fire from privacy advocates, but stresses that users can
opt in or out. Further, it states that users are assigned a random
number and that no IP addresses are stored, while page records are
destroyed once links have been made. Phorm is believed to be in
discussions to license its technology, with Virgin and BT two names
so far mooted.
Data gathering tools
Of course data analytics is not all about customers. The ability
to collect and interpret data within organisations is increasingly
viewed as an important tool for understanding worker behaviour and
improving productivity.
Following a minor exodus from its senior ranks - including
advertising sales boss Tim Armstrong, engineering director Steve
Horowitz, and search quality chief Santosh Jayaram - Google
recently set about developing an algorithm that would help it
predict when someone within the organisation might be planning to
leave. Data analytics is becoming a more important tool generally
for HR professionals, especially now that many companies are
looking to identify their most talented staff for fear of losing
them once the economy recovers.
Google is seen as the inventor of what is now referred to as the
semantic web, or semweb, which is now the dominant search
methodology on the internet. The company's model for linguistic
searching revolutionised the internet experience, both for users as
well as the companies trying to market to them.
A recent addition to the semweb, US movie distributor
Netflix is currently running
a competition to improve development of a system designed to make
people's movie choices for them based on their past viewing history
and other criteria.
While the ideas now being presented by the semantic web are yet
to have any serious impact on the corporate world, they may have
implications for business in years to come.
The right
question
Launched amid much fanfare in May, the science and academic
search service Wolfram
Alpha has been criticised for being overly particular about how
questions are worded. Whether or not this criticism is
justified, it nevertheless raises a fundamental issue with regard
to how organisations think about information.
According to Tim O'Neill, co-founder and director of business
solutions provider Avolution
- a Gartner "cool vendor" - it is no longer adequate for firms to
be merely passive consumers of data. To truly harness its value,
they need to be more thoughtful and aggressive in finding out what
it is that they need to know.
"Often this comes down to IT executives asking the questions
that are normally too painful to ask," he says. "For example, Where
do I cut 30% off my budget without affecting my service
levels?"
Formed out of the University of
Technology in Sydney, Avolution markets a unique 3D-enterprise
modelling system called Abacus.
3D visualisation is an important new trend in information
management, and Abacus is designed to provide organisations with a
deeper understanding of the relationships between the myriad
components that comprise their business and the emergent behaviour
of those components.
O'Neill says that for companies to achieve this, they need to
get better at asking the right questions and better at interpreting
their answers. "It is not good enough to just draw a few box and
line diagrams and hack together a spreadsheet. You have to look
deeply into your organisation and manage it with specific metrics
and rich visualisation techniques."