data management
COMBO EXERCISES
David Abramson of Logi Analytics looks at how organisations can combine data sources,
and also embed analytics within workflows to get a more complete picture of their business.
D
ata fuels us. Nearly
every organisation
today runs on data
– from marketing
automation tools and
human resources Apps to financial
management tools and Salesforce,
and the list goes on. Whether
you’re a consumer with a FitBit or a
business user, it all comes back to
the data.
In fact, IDC’s Digital Universe
study predicts the amount of data on
the planet will grow tenfold by 2020
– from around 4.4 zettabytes to 44
zettabytes. That’s a lot of data, to
put it mildly. And harnessing it for
analysis is now a major hurdle for
businesses everywhere.
Traditionally, IT departments
have managed data for their
businesses in a centralised manner
– and that data came from only a
few applications. Now, business
intelligence (BI) is a completely
different ball game. Not only are
companies collecting ever increasing
volumes of data, they’re collecting
that data from a dizzying number
of sources and in multiple formats.
This makes it very difficult to get a
complete picture of the business.
30
Companies can no longer flow all
their data through a master system,
store it in a data warehouse, and call
it a day. Today, we face the complex
task of tracing the origins of our data,
blending it, and getting it into the
hands of users for analysis – all at the
speed of business.
while census or demographic based
data is useful in the retail sector.
These are all new classes of data
that can be very valuable – but only
if you have a plan for combining data
sources, storing them, and building
analytics on top.
Combining the data
Following the data
The landscape of where data
originates is vastly different than
it was a decade ago. In many
companies, individual departments
manage their own applications
and generate their own data. And
everyone in the organisation has to
not only figure out how to access that
information, but also decide what to
do with it.
On top of that, some of this
data isn’t ‘owned’ by any single
department. For instance, think about
data from social platforms like Twitter
and Facebook – or, in the case of
manufacturing companies, data that
comes from machines and sensors.
Some companies also use public
data – in other words, data they’re
not even generating themselves. For
example, weather data is useful to
logistics or manufacturing companies,
There’s no sense in simply storing
your data and then doing nothing with
it. Once you’ve found and secured
your data, you need to combine it and
prepare it for analysis. In the past, a
‘data mashup’ meant displaying all
the data from different sources on a
single dashboard screen. Today, we
call it ‘data blending’, and it’s a much
more sophisticated process. Not only
do we display the data, we join (or
combine) it to find common values,
and then we query it to gain insights.
The key here, however, is to get
these capabilities into the hands of
your users. The idea of a central
warehouse where you store blended
data is just too difficult to manage.
Data today is updated constantly, and
different people will want to combine
data in different ways.
For instance, your marketing team
may want to blend its Marketo data