Data analytics myths to avoid
Carlos Teixeira, Director, PBT Innovation at PBT Group
Every modern, digitally driven organisation needs to invest in managing its data and effectively analysing it to derive insights that can be of value for well-timed business decisions. Data analytics has for many years now played a key role in supporting businesses to do just this. However, some organisations still seem to be holding back on data analytics. Why is this, and what are some of the concerns and myths related to data analytics? To help get to the bottom of this, I recently came across an interesting and valuable Analytics Insight article titled: ’10 Data Analytics Myths That Can Ruin Your Business.’
While I am not going to discuss all 10 of these myths in detail, I will focus on what I consider to be the five most significant that are influencing some decision makers to shy away from the world of data analytics.
Data analytics is complicated
It can be. This myth comes down to how well a company knows its data, knows where it comes from, and knows what it is used for. If employees do not understand how concepts like System of Record and Sources of Truth fall into the process, then it can be very complicated.
However, this does not have to be the case. If the business understands these concepts and can track the entire lineage of the organisational data, then analytics becomes easier. While this myth can be considered busted, it is only if the foundational elements around data analytics are effectively put in place.
Data analytics is expensive
If a company does not set up its data storage, movement, validation and visualisation appropriately then costs can quickly spiral out of control. This also applies to selecting the right tool for the job. I would even go so far as to say that 80% of a company’s data analytics costs come down to not storing and visualising it correctly.
The best way to dispel this myth is to ensure that the organisation keeps a close eye on its data management and tool selection strategy. For instance, rather use a more affordable environment like Microsoft Azure SQL to store data instead of relying on customised – read expensive – tools. When companies try to rush things and buy solutions off the shelf that have to be highly retrofitted in order to fit into their analytics environment, they should not be surprised when the bill comes down the line.
Data analytics can solve all business problems
The optimist in me wants to believe that it can. However, if an organisation has poor System of Record in place which is caused by human error or gaps in the human process, analytics will be unable to provide true business insights.
Take a bank as an example. It might struggle with lengthy queues at a branch and not understand why this keeps happening. The truth is that it does not have enough data points in place to get the insights necessary to optimise operations. This can be corrected by giving customers an IoT-enabled tracking device when moving around in the branch. In this way, the bank can identify where the congestion occurs and trace points to why some tellers or employee points of contact are too slow.
Ultimately, data analytics cannot provide insights on what it cannot see. The business must therefore ensure that it has the correct environment in which to optimally collect the right kind of data to perform analysis on.
Data analytics is a one-time project
Athletes training for the Olympic Games are never measured just on one race but rather on thousands of them. Why then do companies think data analytics is something that is done once?
It comes down to understanding the benefit over time. Sure, if you perform data analytics as a one-time event, it can show a snapshot of what is happening but then there is no real understanding of the journey taken. For data analytics to be effective, companies must see insights over time. As part of this, they must also realise that not every business or industry sector shows insights in the same time frame. For instance, compare the fast-paced food retail environment with a more timeous car sales environment.
Data analytics is a living process. It must become part of the day-to-day operations of every business. Therefore, it should be entrenched in everything the company does to deliver the correct insights when required.
Data analytics requires a lot of data
The main issue here is understanding how wide (data types) and deep (data volume) data assets are. The higher the depth of data, the more accurate the analysis becomes. Relying only on ‘light’ data can be very dangerous. The width of the data is also critical as it provides potential for better data augmentation.
For instance, the data width can be seen as how many types of apples a company sells. The breadth refers to how many of the individual apples there is to sell. So, if a business only has one apple of each type and sells out, it becomes difficult to know which one is the most successful. Therefore, the more apples of each type available, the better a business can identify the ones that are popular.
Of course, data quality is also essential. If the quality is not there, then the depth and breadth of data become irrelevant. You can therefore have a 10 meter deep Olympic-sized pool to swim in, but if it is filled with sewage, then the size and depth does not matter – it remains useless.
In a digitally driven, data rich world, any business, irrespective of size or sector, can benefit from data analytics. Working through the associated myths and understanding where data analytics can add value are good first steps towards a positive data analytics journey.