Business Analytics: Friend or Foe?

Ever since I got involved in the information technology sector back in the 1970s, the promise of business analytics has been a siren song.  Mine the transaction data under your management to detect trends and extract insights that will give you competitive advantage.  Who would not want to do this?

And it only becomes more compelling with every year’s exponential expansion of the universe of accessible transaction data (we’ll leave the discussion of access and privacy rights to another forum).  Internet enablement of IT systems has driven this resource to unimaginable dimensions from just a few years ago.

Of course, sorting through all the hay to find the needles is non-trivial.  But God bless the math geeks and their indomitable imaginations (see BusinessWeek’s cover article of its January 23 edition for an overview of some of their recent accomplishments).  From the relational calculus that brought us report-writers and spreadsheets to the correlation algorithms of regression analysis to the neo-geometries of semantic space mapping, there are tools that are adequate to task—or promise to be so shortly.

So why is business analytics always a bridesmaid and never a bride?  For that has been its status ever since I can remember.  It is never a bad business, but it has never come close to fulfilling the potential that everyone involved in it believes is latent.  What is the problem?

The problem is closing the loop.  To complete the journey from generating the insights to using them to drive operating procedures that can systematically capitalize upon them in time and at scale has been the exception rather than the rule.  Mind you, when the exception happens, as it has with programmed trading on Wall Street, and as it must and will in anti-terrorism intelligence, the results are transforming.  But for the most part we see a landscape of intermittent connections, flashes of insight, but no systemic gain.

Despite the fact that some of these tools are way too hard to use, this is not in the main the vendors’ fault.  It is the customers.  In most companies, business analytics represent a highly evolved form of corporate entertainment.  Its core practitioners generate insights without accountability.  They communicate those insights to business managers, who do have accountability and who are moved to act, but who are unable to do so in time.  As a result the insights become part of a growing library of great but lost opportunities, supporting a culture eroding into passive aggressive despair.

What is the right answer here?  It is to grow business analytics in closed-loop systems where operationalizing the insight, making the bet, and keeping track of wins and losses, is inseparable from the generating the insight.  Deny yourself the guilty pleasure of insight without accountability: it will only delude you and your colleagues into thinking you are smarter than you really are.  Force yourself instead to contain the scope of the system and let it take you places you had not idea were even there.  Then and only then will you generate the economic miracles of which the sirens have always sung.

Comments

[Carlos Leyva]

Geoff,

Awesome insight! That is why technology only solutions fail more often than not, because the people that could operationalize the insights don't, either because they do not have the organizational position power to do so, or the whole organizational decision making process is too slow, or some combination of all of the above.

However, I recently consulted at a mid-size company where almost all the business intelligence was coming from a monster spreadsheet (not the best of analytical tools) BUT the executive team was operationalizing the crap out of the insights they gathered and changing course (mostly direct marketing strategy into large markets) on a daily basis. In short, they were masters of "closing the loop" with phenomenal results.