Controlling and optimising Business Intelligence applications can be a dry topic even for some of us who work in the domain of BI & Analytics.  This shouldn’t limit consideration of the problem area as there can be significant challenges for organisations with associated costs if the challenges are not addressed.  When we think of various analogies that could be used by way of illustration one scenario to consider involves an imaginary restaurant, “Chez Metrique”, from inception to maturity.

OK, following the commitment to invest in Chez Metrique and a period of concentrated effort by the team to design, develop & prepare, the restaurant launch has gone well; customer reviews are positive and bookings are encouraging.  Looking beyond the bright, busy dining room into the kitchen the stations are pristine and organised and processes well-defined.  The staff are appropriately skilled, have a clear understanding of their roles and communicate effectively with each other.  The output is of high quality, delivered promptly by front-of-house and well received by appreciative diners.

So far so good. Chez Metrique is buzzing.

Of course, from this point in time there are many possible outcomes.  For this scenario, let’s take the common path and peer through the restaurant windows 5 years hence.

Unfortunately, it’s not an encouraging sight.  Only a few people are dining – and with a long spoon.  The once-bright and appealing décor has faded and the buzz has gone.  Pan through to the kitchen and there’s a picture of disorganised endeavour.  The remaining cooks, although doing their best, are working without method or direction and producing meals that should carry a health warning.  Regulators (Environmental Health), if aware, would close the place down in a heart-beat. The once-promising enterprise has turned sour through a managerial inability to adapt to shifting demands and sustained under investment.  Chez Metrique is done.

What a waste of potential, resources, effort & time.

There’s a good chance that another restaurant / café will start up on the same site within a few short months.  Will they repeat the journey of Chez Metrique?

And so to the parallel with Business Intelligence & Analytics programmes.  Often there is an initial directed focus with the aim of supporting a strategic business initiative.  The capability is promoted by positive executive sponsorship and transitioned successfully into business-as-usual (BAU) for on-going support & maintenance.  However, all too often that team is under-resourced and, over time, with the pressure of new initiatives being prioritised the quality levels degrade.  Inconsistencies creep-in to data, metrics, reports & dashboards, it becomes harder for BI consumers to locate, understand and trust the integrity of the information they need. The catalogue of objects expands with increasing duplication and redundancy.  Control slips away and what was a key information asset is now a liability.

Most BI applications start out like this restaurant kitchen.

Many are somewhere in between... could benefit from a deep clean, revised menu and restoration of trust rather than being left unvisited and destined to be shut down.

The application may persist for some time as the effort and cost to regain control, clean-up & optimise is judged to be too great; but it’s in a fugue state.  At some point though, after long-term decline, the zombie app will be replaced with a shiny new set of tools at not inconsiderable expense.  Will the organisation’s approach change to investing in people and resources too?  In many cases, not.

It’s clear that chronic technical debt and poor governance contribute to bad outcomes, be it a promising start-up restaurant or an organisation seeking to use BI & analytics for competitive advantage.  Rinse, repeat & fail.  It’s a path followed by many; but, thankfully, not all.

What are your experiences as a consumer of BI or analytics, developer, analyst or manager? What kind of fayre is being served up in your organisation? I welcome your thoughts.