Predictions for 2018 by Barry Delvin

4 min read

What was, in your opinion, the most important thing that happened in Analytics in 2017?

My view is that little enough of major importance happened within analytics in 2017! However, a number of things happened around analytics that will have major impacts over the coming few years. AI and GDPR, as discussed below, are probably the most important. At a lower level, 2017 saw a renewed focus on data structure – driven by the enormous growth – ongoing and predicted – of IoT. Data from IoT devices has a much higher level of structure than social media, and is significantly dirtier than operational data. In addition, its volumes are growing extremely rapidly. All of these factors contribute to new approaches to analytics and new/revamped tools to support them. In these circumstances, the distinction between data warehouses and data lakes has been subject to reassessment.

What is, in your opinion, the most important thing that will happen in Analytics in 2018?
The increasing scope and power of AI techniques and tools will challenge how we think of analytics, much the same way that the emphasis on analytics in the past decade changed our way of thinking about BI. While analytics was very much a background activity of a business, AI is evolving very much in the public view. Given the overlap between analytics and AI in terms of outcomes, public concerns about the impact of AI will likely spread to analytics, causing more customers to question what businesses are doing with the data they collect and if or how the public actually benefits from this process.

On a personal note, 2018 marks the 30th anniversary of the publication of the very first data warehouse architecture by myself and Paul Murphy in the IBM Systems Journal of February 1988, http://bit.ly/EBIS88.

There was a lot of talk about robotization in 2017 – how do you think this will affect businesses in 2018?

Robotization is simply a specific case of automation (replace human labour) and augmentation (ease or enhance human labour) of the work currently done by people being taken on by hardware and software, especially AI. In 2018, I see it continuing to increase as it has long done. Businesses see it as a very effective way of reducing production costs and so will continue to increasingly adopt it. Robot / AI vendors will emphasise augmentation – it is seen as less politically contentious. But both aspects will continue to grow with increasing speed… until the link with decreasing employment rates becomes too obvious to ignore. How politicians deal with this issue in 2018 will be critical. Based on current direction and speed, I suspect we’ll see increasing social push-back to AI and analytics emerge in 2018.

Big Data is said to be a dying term, and Business Intelligence appears to go the same direction. Do you agree with this, and if so, what do you think is the best term that can replace these?

The idea of dying terms – whatever they might be – has long irritated me. It suggests that fashion drives the IT world, a concept that makes no sense. Mainframes were declared to be dying in the 1980s – they still exist. Similarly, data warehouse, relational databases, and BI have been declared dead. They too still exist. All these things and concepts still exist because they have value and refer to fairly specific and well-defined components of our reality. Big data, on the other hand, was a vague and woolly concept from the start. Not only should it die, it should have been killed at birth 😉 Big data can and should be simply called «data». As for BI, I believe it will live on, although the use of the term «analytics» (as here), and the way that many vendors are misusing or extending its scope, means that confusion between the two terms is causing problems for many customers in defining what they want to do.

What new concepts and buzzwords do you think will appear in Analytics in 2018?

As my last answers might indicate, I have little time for buzzwords! Furthermore, I can’t recall when last we discovered a truly «new concept» in the world of data warehouse, BI and analytics. I do believe some older concepts will become prevalent again. First will be data modelling and structured data stores, driven in part by the 25 May compliance date for GDPR. Second, data security and privacy concerns will become more prevalent – customers will demand that businesses respect and protect their data. Third, and this is perhaps more of a hope, ethical thinking should become a central focus: about how we use data in analytics and AI, what are the biases and unjustified beliefs we embed, what unplanned side-effects in society emerge as we try to use data to drive everything… and how we can address these important issues.

About Barry Devlin

Dr. Barry Devlin is among the foremost authorities in the world on business insight and data warehousing. He was responsible for the definition of IBM's data warehouse architecture in the mid '80s and authored the first paper on the topic in the IBM Systems Journal in 1988. He is a widely respected consultant and lecturer on this and related topics, and author of the comprehensive book Data Warehouse: From Architecture to Implementation.