Analytics: Hang the imposter!
People want to replace the term business intelligence with ''enterprise
analytics.'' I don't buy it. At least, not yet.
First, nobody agrees on the meaning of the term ''analytic'' or ''analytics.''
I've polled people in our industry and everyone has a different definition.
Some think ''analytics'' are reports, while others believe they are report
components, such as calculations, pivot tables and data access functions. Other
people think analytics are key performance indicators or metrics. A few believe
they are fancy dashboard graphics (for example, ''stop lights'' or ''car gauges''),
and old-timers think analytics refer to sophisticated statistical algorithms or
data mining functions.
In my opinion, when something can mean anything, it means nothing.
Then we come to the term ''enterprise analytics.'' Proponents use this to
describe next-generation business intelligence capabilities, such as real-time
monitoring of key performance indicators and closed-loop processing where the
output of analyses are fed back into operational systems. But if this is what
''enterprise analytics'' is, why do we need a new term? Why can't we just call it
''next-generation business intelligence'' and leave it at that?
By now, some of you must be rolling your eyes. So many buzzwords have come
and gone in our industry that many of us are numb to them. We mostly ignore
them, or we translate them to something with which we are familiar and move on.
So what's the big deal?
Despite the industry's constant churning, semantics do matter. Ever since God
put Adam in the Garden of Eden, man has gained control over his environment by
classifying and labeling things. It's no different today.
Vendors spend a lot of time and money evangelizing terms for new technology
sectors, hoping to assert their leadership and steer the market in their general
direction. In general, he who coins the terms of the debate defines and controls
Ideally, vendor-neutral organizations, such as The Data Warehousing Institute
(TDWI), or analyst groups should classify vendor products and market directions,
placing vendor organizations on a level playing field and bringing clarity to
the market. (Of course, problems occur when analyst organizations offer
competing terms and definitions.)
We all benefit if we can standardize terms and definitions. Many of you are
fighting this war within your own organizations. You know the value and business
benefits of creating a single version of the truth. What's true for user
organizations is also true for markets.
Why business intelligence?
At TDWI, we like the term ''business intelligence''
because its meaning is self-evident. The goal of business intelligence products
and services is to make businesses ''more intelligent.''
TDWI formally defines business intelligence as ''The processes, technologies,
and tools needed to turn data into information and information into knowledge
and plans that drive profitable business action.''
For us, business intelligence is an umbrella term that spans IT-oriented
processes and tools to prepare and integrate data for analysis; and
user-oriented processes and tools that enable business users to analyze and use
information for business gain. To this end, business intelligence encompasses
data warehousing and content management, collaboration and knowledge management,
ETL and EAI, data mining and text mining, OLAP and portals, and so on.
Unfortunately, most of you may not agree with our definition. Or you may
agree, but you don't adhere to it in practice. In conversation, many of us may
slip into using the term ''business intelligence'' to refer to front-end analysis
tools, such as those from Business Objects or Cognos. This is common, but
unfortunate, because it underutilizes the potential of the term. These end-user
tools can't work without the underlying data warehousing infrastructure to
support them. We can't make businesses more intelligent just by implementing
Business Objects -- far from it.
TDWI uses the term ''business analytics'' to refer to end-user analysis tools.
We now have a ''Business Analytics'' track at our conferences that contains
courses on reporting, data mining, OLAP, visualization, geographic information
systems and so on. We think the term Business Analytics is appropriate since the
dictionary defines analytic as: ''pertaining to or proceeding by analysis;
skilled in or habitually using analysis.'' People use business analytic tools to
analyze business data. This makes semantic sense.
However, TDWI is not deluded into thinking we can convince everyone to use
our terms and see the world as we do. But we have a professional responsibility
to educate our audience and sharpen any fuzziness in the marketplace.
Our task is harder because we (the industry as a whole) have been sloppy in
our definition and usage of the term ''business intelligence'' and now we're
paying for it. Some want to replace one sloppily defined term with an even
sloppier one -- ''enterprise analytics.'' I'll only stomach this semantic
sea-change if the definition of 'analytics' becomes crystal clear and there is
an undeniable groundswell of support for it.
Until then, come visit me in my Garden of Eden. We'll have plenty of time to
enjoy the scenery.
Wayne W. Eckerson is director of education and research for The Data Warehousing Institute, where he oversees TDWI's educational curriculum, member publications, and various research and consulting services. He has published and spoken extensively on data warehousing and business intelligence subjects since 1994.