Analytics – not so complex after all

Salim is Founder and Managing Director of On-Target Marketing Solutions, Malaysia, a digital and data analytics company and training ground, and given his immense passion for data has combined over three decades of Marketing experience, recently with a Data Science Certificate course. He is contactable at [email protected]

Recently, while addressing a small forum of senior IT professionals and marketers over a breakfast meeting, we went around the table over the ice-breaker and asked each of the participants to define what they were keen to derive or learn from the workshop. Interestingly we noted that a few of the responses were about wanting to understand what ‘Data Analytics’ was all about and the hype around it.

The last few years have undoubtedly witnessed a data explosion and the scope of this piece is not to dwell at length on that happening. To exploit from this hugely significant occurrence have evolved disciplines such as Data Analytics and Data Mining, Machine Learning, Artificial Intelligence and many more.

Not quite out of context, but I am reminded of a famous and old quote from the great Albert Einstein which goes as “Everything should be made as simple as possible, but not simpler”, which holds so relevant. Instead therefore it’s about distilling terms like Data Analytics which appears rather complex and overwhelming to some.

If we look back for a moment ‘data’ has actually been in existence since forever as we can recall, taking the form of lists initially which simply contained fields of information like identity and contact details of individuals. Even these basic and small pieces of information carried insights. For example, a simple name and address record could perhaps give us demographic and even socio-economic insights, based on which we could make certain conclusions and treat them accordingly.

The Data Loop

With the internet and social media, we have massive volumes of data coming at us through a variety of sources, with high velocity in real time and greatly accurate (a. k. a. ‘big data’), and thus carrying far more information and ‘insights’ which can be extracted.

As technology enables generation of high volumes of data, there are also fortunately increasingly sophisticated tools that form what I’d term as the ‘analytics layer’. This layer helps to explore and extract from this data very valuable insights for multi-disciplinary application across pretty much any category or specialised field.

There is thus the opportunity as well as need to extract, make decisions and act on these valuable insights in real time with equally high speed and accuracy, to derive the most out of them. Technology once again serves as a great enabler at the next ‘implementation layer’ to ensure these insights are addressed, through necessary action being taken only to generate even more data, and thus resulting in what could be termed as a ‘data loop’.

The Data Loop

From the above if we thus take a step back, the concept of Data Analytics isn’t all quite new but the scale (of data and accompanying insights) has multiplied and taken a quantum leap, rendering complexity to it.

Applying the Data Loop principle

If we were to illustrate the above fairly basic explanation, Google Analytics serves as a ready example among applications or tools that we marketers would relate to and use. If we stare at even a single frame of the output on the GA dashboard, it can give us some deep and fascinating insights to the sources of traffic to various pages on a website, level of audience engagement and interest with the page, exit patterns and more, thus prompting necessary action towards the content or sometimes the technical aspects.

Scaling up the above simple reality of data analytics to more advanced concepts, through Machine Learning, machines or devices are enabled or ‘taught’ to extract data, analyse it, draw the insights and go on to even make decisions and take action based on built-in algorithms, all of it being done in real time for an enhanced output, and result in a great ‘customer experience’. This is done with the help of training sets which are refined on an ongoing basis.

We do need to bear in mind that the pivotal piece or raison d’etre in the Data Loop is the ‘insights’ piece, from which arise the ideas or alternatives on actions and implementation. Whilst the key lies in having clarity from the start on what to seek from the data, asking the right questions and then extracting accordingly with the right tools in order to maximise insights through analytics, rather than being overwhelmed with the data tsunami and information overload and letting analytics get the better of us to lead our minds almost into a state of paralytics.