Data is now the driving force behind successful business. Enterprises that are able to
interpret and act quickly on insights gleaned from data generated by their activities are
more likely to come out on top in an increasingly competitive business climate.

And there has never been so much data available to an enterprise. Data is now routinely
recorded in petabytes, each equivalent to 500 billion pages of paper. However, the
nature of that data has changed considerably. Traditionally, information stored in digital
systems has been highly structured. There was no other way—every bit of information
was expensive to store and process. Architects of these early systems had to work with
only the most essential data, meaning that it was down to humans to make sense of
the information.

Today, companies that want to remain competitive must take a much broader approach
to the kind of data they gather. The need for humans to arrange data into machine
readable structures is now going away. Formerly, to get a public opinion, field research
had to be undertaken , and then rely on some form of “data entry” to translate scribbles
on paper into machine data. Only then could you start with your analysis. Today, you’re
no longer bound to just a small sample group of people to get your information. You
can poll Twitter and Facebook and process all available opinions, at once. Nuggets
of market intelligence are better mined from what customers, fans or citizens say
voluntarily using outlets like social media where they are more likely to speak their
mind. The downside — it’s people. They talk in pop‑culture references, use satire, slang,
and make grammar mistakes. All that means the information is captured in its raw and
unstructured form, challenging anyone who wants to tap into the riches to employ
modern interpretation techniques and accept much larger volumes of resulting data.

Data is also generated in vast quantities by automated systems such as those that
run factories, manage vehicles, monitor energy usage in your home and even control
vending machines. Add to this business data from e‑commerce and more traditional
sources and most organizations have a wealth of data at their disposal that they didn’t
think could be useful in the past.

This so‑called big data (well, wait a bit and it will, again, be just “data”) is not only
available in large volumes and in many different formats; it may be generated over very
different time periods — a mixture of data stretching back years and data generated
every second by a production process. These variables are often referred to as volume,
variety and velocity. However, big data by itself is useless unless it can be analyzed and
interpreted in a timely fashion.

Without an enterprise data architecture — an amalgamation of models, policies, rules
and standards that enable your organization to collect and analyze data — you will
not be able to get a full picture of your business. This potentially affects the ability of
a business to make crucial decisions, such as how to improve your products and services,
the best way to speed up manufacturing or where to site new warehouses. Your data
will lay dormant, scattered, unused and often perceived to deliver no extra value.

There are many different models of enterprise data architecture, all with different
benefits, drawbacks and challenges for a business. The aim of this white paper is to
guide decision makers towards an enterprise data strategy that will get maximum
benefit from their data while staying realistic given timelines, skills available and budget.