Businesses, like humans, not always realise that they are intelligent.
Unlike us who can test ourselves, a company may not realise how much valuable a
single column of information is. Business Intelligence (BI) is not new but it is
certainly not easy. “Traditionally BI was complex, difficult to implement,
architected for selected analysts and hard to use. After having invested a lot
in back-end systems, less than 10 percent of their employees actually touch it,
or get access to the data,” reflects Mr Sanjay Mehta, Chief Executive Officer,
MAIA Intelligence Pvt Ltd, Mumbai. His company is small but it has got its own
key, literally. ‘1KEY Agile BI Suite’ is a comprehensive and robust BI
application, tested and certified by Microsoft, that MAIA offers for catering to
strategic, tactical and operational data analysis as well as reporting needs.
“BI was once considered a domain of statisticians and analysts; not anymore:
Earlier, only the
elite had access to business intelligence. High cost and manageability issues
restricted enterprise-wide BI roll-outs,” Mr Mehta told Business Line in
course of an interview, conducted over phone and email. Catch him in this Q&A on
how businesses are becoming intelligent...

Edited excerpts.

Tell us, what does MAIA do?

Founded in 2006, MAIA Intelligence is a young innovative company in powerful
business intelligence reporting and analysis software product space. The company
aims to democratise BI and take it from few expert users to the operational
managers, frontline executives to back office. Quarter after quarter, we have
released new product features and product enhancements to address the business
and technology challenges that companies face today. An integrated offering, we
feel, should be one with a choice to pick and choose modules, enabling
organisations to deploy BI framework with minimal investments.

The kind of work that your software does, you must be competing with MNC
firms…

Globally there are companies in BI like Business Objects, Cognos, SAS, Hyperion,
Microstrategy which are known as traditional BI. Such deployments in any company
are expensive and used by only top tier management level, expert users –
analysts; the rest of the 85 per cent of the business user pyramid is deprived
of a BI for MIS, analysis and monitoring / gauging performances, which if
provided can help them gain visibility into the business and drive performance
and get everyone working towards a common goal. BI markets continue to evolve.
There’s still plenty of room for growth in the BI market. The BI market has
only penetrated 10 to 15 percent of the known user base, but there is a vast
opportunity for BI well beyond today’s known markets.

Affordability doesn’t sound glamorous; but that’s what it matters in
enterprise roll outs: Implementing BI with few experts is easy, but rolling out
to 000’s of users within the enterprise and beyond is a different story.
Traditional BI systems are expensive, open budget projects.

During the telecon, you told us about an unusual client of yours where your tool
found that each employee was wasting 53 working days each year. How did you go
about the exercise?

One of the leading financial services and broking companies in India had
deployed our BI tool for their internal MIS reporting. The HR head was smart and
with help of the CIO he took the challenge of finding out the productivity loss
due to the late coming of employees. The results were startling.

They had this show card attendance system which was recording the data of
incoming and out going of all the employees. We found that company had 53 days
of working productivity loss. Sliced into various departments, data showed which
department employees were coming late.

Then the same data were sliced further to male-female, with various role
hierarchies, and the results were startling. The same was then applied to check
the attrition and movement of employees within department, male-female,
designation-wise, period-wise. The senior people were the ones who were coming
late often. The HR department could actually resolve and pinpoint and resolve
the issues related to attrition and productivity loss. I am sure this kind of BI
experience will really make any management sit up and improve corporate
performance management.

Does it help if data were accessible from close-ended business software?

Yes. For example, Tally data are not open; connecting such data for doing
reporting and analysis is not everyone’s cup of tea. In-depth knowledge of
Tally data structure (flat file system) and TDL (Tally Definition Language) is
required to extract thousands of records from multiple companies of Tally
database within seconds, allowing analysis on the same in a BI. So, 1KEY talks
to and connects to Tally database and lets users create dynamic reports quickly
and easily. We helped have one such customer, Gujarat State Petroleum
Corporation (GSPC) who had challenge in reports from Tally.

What are the data that CEOs most prefer to see on dashboards?

CEOs mostly prefer to see the data which give the macro view of everything
needed to run the show, and these data have to be accessible at a glance. The
dashboard of a CEO contains critical details of a business that are often buried
deep within a large organisation.

CEOs want their dashboard to be powerful enough so that they can see key changes
in their businesses almost instantaneously and take quick, corrective action.
CEOs can choose from metrics for their dashboards, like sales, production,
profitability, RSS feeds etc.

A CEO’s dashboard should contain more useful information than just data. The
reports should offer early warning, so that CEO can proactively take steps to
avoid a serious problem. When driving, an early indicator of, say, low fuel is
useful so that one can obtain fuel rather than wait for the car to suddenly come
to a stop; it’s also nice to know that you need gas, and not an oil change.
Otherwise, to diagnose the problem can take time away from solving the specific
issue that’s dampening performance.

A CEO dashboard can offer the CEO information, in context, regarding the
performance of a crucial area of the business system, flagging trouble before
the problem fully manifests itself. Context is easily established around a
business process once the steps of the process are understood along with their
cause-and-effect relationship on each other and on other areas of the business.

Have there been instances of failures - when BI didn’t deliver?

In case the data quality is not good, BI can fail. Technically it works fine but
functionally it may give wrong results. For example in ‘city’ column in few
records it is ‘Mumbai’, in some it may be ‘Bombay’; then in city-wise
reports Mumbai and Bombay would be considered different. Technically it is
correct, but functionally the report it might be wrong.

Secondly, if there are no data in some fields – for example ‘city’ field
is blank; then the city-wise report generated out of such data may be wrong.
Cleaner data allows better reporting and analysis.

What are the key challenges in BI?

Organisations must understand and address few key challenges for BI success. BI
projects may fail because of the failure to recognise BI projects as
cross-organisational business initiatives, and to understand that as such they
differ from typical standalone solutions. Unengaged business sponsors (or
sponsors who enjoy little or no authority in the enterprise) can cause failure.
Issue such as unavailable or unwilling business representatives, lack of skilled
and available staff, or sub-optimal staff utilisation also can be dampeners.

Failures can be seen where there is no software release concept (no iterative
development method), no work breakdown structure (no methodology), no business
analysis or standardisation activities, or no appreciation of the impact of
dirty data on business profitability.