It would be foolish to suggest AI isn't massively transforming how we analyse data or the broader data analytics field. This super-power of intelligence and grunt is revolutionary - what hope do mere humans have? For the foreseeable future, a lot. Leaving AI uncontrolled with our complex information and insights would be reckless. Rather than succumbing to a digital coup the analytics professional has a critical role in not only taming and containing, but grounding, guiding and inspiring this uprising.

​

Data analytics is essentially the science and art of transforming data to insights. These may be insights about the past events, the reasoning and drivers of these, and of course predicting the future. This does not necessarily need to produce actionable insights, but certainly in a commercial context it should, or at least be driving decisions.

​

This AI-led transformation will come across many fronts. AI will mean more power, speed, intelligence and accuracy. This could conceivably shrink the analytical role, although demand will be raised as data and application scenarios surge. Furthermore, AI will democratise data analytics, allowing the layperson to indulge in quite sophisticated data interrogation via mere chat interaction. This is great for everyone.

​

AI may be able to drive our cars and fly our planes, but what about our cross-tabs? Such developments press the question what the data analytics profession actually is. It is not simply number crunching. This part of the task will increasingly be delegated to machine analytics or the layperson - perhaps a relief to everyone. The danger lies in trusting these channels for deeper analysis, uncovering insights and driving strategy.

​

It is here the analytics professional steps in. The role will become more and more focused on guiding and applying the technology whilst to a lesser degree driving it. This evolution will provide temptation to reshape data analytics. This is not the case. Despite the evolving landscape the fundamentals of successful analytics remain unchanged, merely adapting to this AI-intrenched world.

​

So what are these fundamentals?

 

• The first is recognising effective analytics is a team pursuit. A team of business experts, engineers, data scientists, broad expertise and of course machines. Just as important as raw skills, the most effective solutions come from pooling the minds of different people and experiences. This stresses the importance of diverse teams, not just in terms of gender, age and culture, but also experience and personality. Often it is how people think just as much as what they know. Analytics at its best is finding the right road to the destination, not driving the predetermined route.

​

• Second is recognising the most important period of a project is the beginning. Absorbing perspectives early allows thinking through many potential solutions based on the experience and expertise of team members (even those not directly working on the project). It also allows being able to see early the impact end to end and what will work to achieve goals given the nuances and restrictions faced. Likewise pitfalls regarding data, methods or application can be avoided early. It is far easier to take the right road from the beginning than having to turn around or end up at the wrong destination (see also my previous post of unmatched ability of the human mind to work through complex and nuanced problems).

​

• Analysts must always focus on the problem. Too often analytics can evolve into creating a solution for which a problem must be found - this is rarely an effective strategy. Rather be clear on the issue at hand and going ahead to solve it. AI's are not distracted so easily. A lot of solutions will have multiple applications, but the essence of analytics is unique solutions to unique problems. This requires steadfast focus and plenty of creativity.

​

• A simple one but know the data! We are becoming increasingly accustomed to the word 'hallucination' with respect to AI. A big limitation of AI in analytics is its inability to fully grasp the context of data, its uniqueness and nuances, let alone the broader issue at hand. This is exasperated by the breadth of data from vast sources which need to be sorted into the good, the bad and the ugly. Fully outsourcing analytics to AI leaves significant risk of misinterpretation. Human analysts are just as prone to this limitation. Data will get more and more abundant and complex, and analytics teams must relentlessly remain on top of it.

​

• AI will bring a whole new level of sophistication, but it will remain critical for analytics teams to understand the tools and the processes they use. For most people analytics is a process of mysterious magic - taking data from which comes beautiful chart. To some degree it is, but magic can go wrong. If data analytics tools, be them broader AI or specialist data tools, are misunderstood it can introduce unintended bias, faulty assumptions or a raft of other fallacies. No analyst wants to make the news! These tools indeed are becoming increasing difficult to fully grasp, but analysts need solid understanding on how AI tools are built, what they are doing and how the statistical processes function. People rarely want explanations but would like to know they would get one if they dared ask!

• Analytics is largely communicating data and insights. The better this communication, the more effective and appropriate the outcomes of it will be. AI has the potential to do this very well. What analytics teams can add to this is the value of context and relationship. it is important to understand the real world context and relate to the audience. These are the soft skills analysts must continue to hone and expand.

​

• It remains important to recognise data analytics is a critical task. Analysts need little persuasion of this but this is paramount within broader organisations. With advances in technology, analytics can easily be seen as an area to save a bit of budget and pass on to the machines. In reality it is the opposite, the increasing complexity, volume and nuance of data and processes mean it is now more important than ever to have people at the table who can decipher, understand and apply it effectively.

​

The potential of AI in data analytics is very exciting and there is no better time to be in the field. It will lead to better decisions and actions, and everyone should benefit from that. But to optimise this a strong partnership between machine and human is needed. For humans, despite all these advances, the fundamentals of analytics remain more or less unchanged. Remaining focused on these will mean data analytics will continue to have a vital role in organisations, society and beyond.

​