Primary tabs

For the last couple of decades, savvy businesses have been using both Management Information and Business Intelligence (MI and BI) technologies to maintain better control over operations, and to make better business decisions.

More recently, new tools have emerged to help them make sense of the huge swathes of amorphous, unstructured data they can now access as a result of social media take-up, increased employee mobility, and further automation of manual processes.

Meanwhile, the leaders of large, private and public sector organisations are showing a resurgence of interest in the use of scientific methods to support strategy and policy. There’s a chance that evidence-based decision making will become accepted as superior, and in consequence we’ll see a decline in the number of decisions made in haste, or with undue bravado.

With so much current interest in analytics, CIOs can expect support from various quarters when they start to plot their organisation’s course. But with all eyes on this subject, they have to make sure that their course is the right one.

Approaches in analytics

Right now, we see four dominant approaches in analytics:

  • Technological: analytics activity is stimulated by technology products, and emphasis is placed on what interesting data exists, and what interesting insights can now be gleaned.
  • Scientific: an evolution of the technological approach, which seeks also to experiment over time, monitoring and comparing alternative outcomes.
  • Informed: the emphasis is on seeking-out the questions, which would generate the greatest business benefit if answered with analytics.
  • Learning: the most mature approach, which puts the right questions ahead of the most interesting insights, but also uses experimentation to create any necessary missing data.

Approaches in analytics

Once you’ve understood what’s useful to the business, it’s time to check the viability of answering the questions you’ve framed. Reach this point soon, because value is sometimes overwhelmed by costly complications which may take time to address, or may even prohibit further investment some areas. They include:

  • collecting, cleansing and repairing data
  • applying structure to unstructured data, and quantifying qualitative data
  • commercial and regulatory sensitivities
  • infrastructure cost from running complex algorithms on large data sets
  • skills shortages
  • attachment to the existing decision-making culture
  • executives’ loss-of-face if reports are at odds with their instincts.

There are usually solutions to these problems, but it sometimes takes creativity and experience to find them.

Advice is at hand

CGI’s CIO Advisory practice has addressed these challenges before.

We understand the marketplace of analytics tools and high-performance-computing suppliers, and the potential they present.

We’re rational and uncompromising, so we uncover the real questions before seeking to answer them.

We offer creative possibilities to clients, but also have decades of deep engineering experience, so were early to recognise what can reasonably be achieved through the manipulation of data.