Analytics is new mantra in data management

WE LIVE in an environment infested with data. There is big data (and by corollary, small data) but data is not tantamount to insight and more significantly, data does not mean "measure". Management guru Peter Drucker said: "If you can't measure it, you can't improve it". That begs the question: What do we measure?

From countries to corporations, there are clear trends towards outcomes and sometimes, qualitative (and therefore less measurable) goals. This brings with it a challenge for traditional mathematicians, accountants and economists.

There is also the reality that in many situations, the input is measurable, while the output is not. Marketers battle all the time with trying to measure sales or market share effect from a marketing or advertising campaign. Some ROI (return on investment) techniques are in vogue, but nothing is perfect.

Measurement of social programmes and sectors are now receiving attention as large private and public funds pour into these activities. If you want to measure the healthcare status of a public health programme (which most advanced nations have), the ideal outcome that needs to be tracked could be longevity and disease prevention or reduction. Multiple factors impact such an outcome and hence the performance of an X-ray unit, for instance, cannot be directly linked to this. Thus far, the X-ray department would be monitored based on the number of X-rays done in a specific period. Does this really signify anything? NSW Health, the state health programme of New South Wales in Australia, is trying to move away to the larger healthcare goals for measuring and rewarding healthcare professionals and organisations.

Countries too have expanded their palette of measurement goals. Bhutan has coined the Gross National Happiness Index (GNHI) to help track its progress rather than the ubiquitous GDP (gross domestic product).

Singapore's national bonus computation for civil servants is based on four socio-economic indicators each accounting for 25 per cent of the total. These are coated with finer sophistication to the traditional metrics, and link outcome to the human life quality and not just dollars and cents.

Several years ago, manufacturing companies (and increasingly service companies) embraced the "Six Sigma" concept that measures not the output, but the quality of output in a reverse way by pinning down maximum defects in a given population of output.

This was when companies realised that cost wastages that arise from poor shop floor quality outweigh the gains from faster productivity and quantitative pursuit. If the cost wastage data was not analysed, this would not be possible. Thus analytics, the modern day statistical czar, did exist in the past.

Companies are including social metrics into management rewards - ecological footprint, human development index etc. - and not just shareholder-centric financial numbers.

Unilever is pursuing an altruistic goal of "help a child reach 5" by focusing on early childhood hygiene and attacking factors that cause childhood deaths due to poor hygiene. If this could be measured and correlated to specific steps or usage of hygiene regimen (including products), it would be a good impetus to measuring progress in human health.

Analytics is the new mantra in data management as there is realisation that not all numbers matter. Burgeoning Internet usage led us to the wrong metrics initially. Eyeballs, clicks, views, unique visits, traffic, etc. were marketed as the end goals for online efforts.

It soon dawned on us that if these did not translate to actual business outcomes, do they really matter? When DoSomething.org put out a YouTube campaign seeking donation of used sports equipment to the poor, it received 1.5 million views and zero equipment!

As computers track, store and process everything that we do, there is rightfully, an urgent yearning for a distilled picture. The same technology companies that once sold the virtues of capturing "data" are now busy converting voluminous data into a handful of metrics that are usable as dashboards and can lead to meaningful action or conclusion. Along with this, we will see more research into what constitutes a reliable metric that correlates to outcomes. Simplicity has a knack of prevailing.