The Future of IT: The Next Five Years.

Enterprise CTOs or CIOs often ask me about how IT will change in the future, about the “Disruption of Technology in 3-5 years”. Most CTOs and CIOs realize that that is primarily a strategic question rather than a question about a certain vendor or technology. I usually explore the specific challenges together with the executive in five areas.

1. Radical Simplification

We are at a point of changes in production (different server and application architectures) and changed in consumption (on demand, on mobile, as-a-Servíce). We have built layers upon layers. We have built fixes for fixes that were fixes for other fixes. We will enter a period of radical simplification in production, customer journey, and data management.

2. Hyperscale Infrastructure

Hardware and software not only have to radically simplify. We will move away from separate organizations for Server Ops, Network Ops, Storage Ops, and Application Management / Developers. Instead, we will have a new ITOps organization. Applications will have to adjust as well to different server architecture as well as runtime architectures. Containers will be the new norm, and the transition will take place rapidly, in the next five to six years. Edge computing will not be a topic anymore: Everyone will implicitly mean edge computing and distributed computing when you talk about computing. Streaming data analytics — with and without machine learning — will be part of all applications.

3. Reduced Transaction Costs

Radical simplification together with hyperscale infrastructure will lead to reduced transaction costs. IT leadership should re-evaluate business areas or operational models that were previously cost-prohibitive.

4. Machine Learning plus Automation

Machine Learning will be everywhere. You don’t try out “ML-enabled applications”, you implement solutions for meaningful business results — all of which will operate within an architectural framework of machine learning and automation. If you didn’t build out your organizational machine learning and automation muscles yet you will be left behind within three years: Your organization will be puzzled how to PoC machine learning over traditional automation, you will be hesitant and too slow to test-drive machine learning features, you will select non-machine learning solutions where machine learning can provide you with significant competitive advantages, and you will select machine learning solutions where your adoption and transformation friction will have no discernable advantage. Competitors that have developed machine learning and automation skills throughout their whole organization will outperform, out-invest, out-operate, and outpace you. Machines are the new non-collar workforce.

5. Enterprise Agility

Successful businesses are business-driven, not IT-driven. But they find a way to communicate the potential of IT agility and translate it to the potential of enterprise agility. It’s not as if a business ever said “we want a slow and inflexible, monolithic infrastructure that has a strong vendor lock-in and would take a decade to transform or displace.”  Now your operational model will embrace self-organizing networks and allow for self-organizing enterprises. IT will become monolithic in a sense that it allows flexible arrangements within policy auditing frameworks (versus policies) and within global privacy frameworks (versus database rules) and within security frameworks (versus security checklists and provisioning processes).

Bonus: Closer Relationship with Vendors

All of the above will create a much closer relationship with vendors who need (and want to) understand your business and support your constant transformation processes. Successful IT organizations will have learned to work closely with vendors and startups. They can deal with the fluidity of platforms, and changing ownerships of products and smaller vendors. Successful IT organizations explicitely expand their workforce into vendors and thus treat some of their vendor budget as workforce and product development, versus licensing and NRE cost.


This is what you want … and this is what you get. Remember “Hardware”?

Thoughts? Opinions? Comments? Corrections?