In my quest to continuously learn more about data-related roles, I have recently come across two very interesting articles that suggest that data-related roles will be changing throughout the remainder of the year. These pieces share some great insight, and I couldn’t resist the opportunity to share my additional views and so in this, the first of a two-part blog series, I will examine the Chief Data Officer (CDO) role, the Chief Analytics Officer role (CAO), and the evolving roles of data engineers.
CDO and CAO
The first piece published on Spiceworks, ‘Top 5 Ways Data Roles Will Change in 2023’, addresses the changing use cases of data to predict that the previously distinct roles of CDOs and CAOs will likely become one role. As part of this, CDOs will increasingly begin working on how companies leverage information and insights.
The author, Anuj Mudaliar, quotes a 2023 report by the Harvard Business Review which states that more than 60% of CDOs today are working on initiatives involving AI and data analytics. I agree with this assessment especially given how many businesses are trying to cost-justify the appointment of a CDO, never mind having two separate data-related roles. I believe that as organisations mature into their analytical insights journey, only then will they attempt to create a CAO role. However, until then the CDO will be expected to cover both areas.
The second article was published on Technative. Entitled ‘2023 is the turning point for data roles’ and authored by Andy Palmer, the co-founder and CEO of Tamr, it builds on the notion that ‘the CDO and CAO have developed as relatively separate roles, each with its own responsibilities and remit. But the ongoing evolution of data means these roles must converge.’
Palmer bases this on the fact that CDOs must embrace a more holistic view of the way data is consumed across their organisation. In turn, CAOs have begun realising that clean, curated, continuously updated data is essential in allowing them to deliver meaningful and valuable analytical insights to the business. This aligns with the “garbage-in, garbage out paradigm”. In other words, you can develop the most amazing analytical models, but if they are trained and executed on sub-standard data, the usefulness of the insights produced may still be questionable.
Mudaliar further reckons that CDOs will also have to take up the roles of CAOs in cleaning and categorising frequently updated data to meet organisational objectives, in order to mitigate the issues arising from poor data quality. It is here that I must note that I disagree with Mudaliar as I think data categorisation and cleansing sits squarely in the remit of the CDO. My view is that a CAO is responsible for producing insights that increases business value. This is dependent on good quality data, but not responsible for it. They may produce insights to report on the data quality and the impact it may have on the business and the decisions made, but the cataloguing and cleansing is not their responsibility in my opinion.
Data Engineers
Another interesting overlap is that both authors stress how much more important the role of data engineers is becoming. I think this is refreshing at a time when sceptics believe AI and bots will take over all programming jobs in the next five years. For me, this is a key area that will require a substantial skills pool and will rely on companies driving skills development internally.
I agree with Palmer who writes that ‘while the number of data scientists has exploded in recent years, many companies have realised that without next-generation data engineering, it’s becoming increasingly difficult to realise all their data science and AI-related promises.’ After all, if you want your data scientists to produce valuable insights, you need to enable them with the appropriate data with an appropriate level of data quality.
It is a waste of a valuable and highly skilled resource to expect data scientists to extract, wrangle, and clean data, before they can apply their analytical and machine learning processes to it. The data wrangling is much better handled by skilled data engineers.
Both authors also accentuate that the data engineers’ skills will have to evolve to include a fair component of business knowledge and understanding too. Palmer highlights how business analysts are getting frustrated with the lack of management and quality of the organisation’s data. To me, this creates another opportunity to fill the data engineering skills gap especially considering that there are more technical business analysts that can be cross-trained into the data engineering roles. Their advantage is that they already know the business meaning of the data they would be managing. In my mind, the technology aspects are easier to train on and faster to learn than the acumen of business knowledge that may take years of walking the passages to build up.
Of course, I’ve just touched the tip of the proverbial iceberg when it comes to this topic. Look out for part two of this series next month as I delve even deeper into this exciting topic.
