Democratising Data Science and Analytics
For decades, enterprise IT has been on a constant trend of taking responsibilities from business units into a central IT function who provide technology "as a service" to the rest of the organisation.
This worked for a long time, when enterprise IT was all about cost control, stability, governance and managing outsourcers. In todays world though, it's a hard sell to continue centralising control.
Today, almost anything that business units and product teams want to achieve involves software. To depend on a central IT team to deliver is often too slow and unwieldy, so they are increasingly hiring their own developers, creating their own product aligned technology groups, or buying technology directly. This was previously called "Shadow IT", but is increasingly an approved way of doing things.
Over the next decade, I think this trend will continue and centralised enterprise IT will unwind further in the pursuit of speed and agility. Business units will do more of their own technology work, leveraging cloud and SaaS as they go rather than the centrally mandated offerings.
As central IT shackles fall away, this will add momentum to the trend of the "citizen developer", a person within far corners of the business who can build their own applications and make them available to other people to help them do their job or better serve customers. With some support, they will be able to deliver their software with appropriate guardrails and have it accessible to other people in the business.
Moving from a situation where all software is centrally mandated and planned, to one where thousands of people on the ground can develop solutions is surely a win for innovation. For this reason, I'm really bullish on tools such as Airtable and Mendix as we are really early in this evolution.
Data is an area which will should have the same evolution. Where before it would be central teams of analysts looking at data, enterprise data will increasingly be opened up (with appropriate controls and anonymisation) to people within the business to analyse and test hypothesis with. As with software, the tooling is in parallel coming through such that power users can begin to do really powerful data science type work using higher level tools.
Businesses full of people writing code and innovating with software and data is an interesting future, and it's a development that we need to plan for today, whether you're a buyer, vendor or enterprise IT department.