Recently, I was asked to present my personal thoughts around what enterprise IT might look like in 2020.
That's an interesting horizon: only a bit more than six years ahead of us, and during a period of intense change in our industry. In some regards, six years is a relatively short time in enterprise IT evolution: you're pragmatically limited to ideas, technologies and concepts already evident in the marketplace. Even if something radically new and game-changing were to be invented tomorrow, it would take many years for it to be widely adopted.
For me, the more interesting discussion is the enterprise IT function itself: how it will evolve, what new roles will it play, and how it becomes the essential core of any modern business strategy. The technologies are interesting; how people use them -- even more interesting.
Let's Go Back A Bit, Shall We?
A reasonable starting point might be to go back to 2006-2007 and reflect on the last six or seven years. We're talking a world that would be missing several key concepts that we're all familiar with today: cloud, mobile, big data, XaaS, app factories, etc.
We were discussing utility grids, not clouds. SOA was the sweetheart of enterprise architects. VMware was steadily working its way out of smaller test/dev environments and taking over more critical apps. Data warehouses with tidy models were in vogue, vs. the present unstructured world of big data.
The enterprise IT groups I met with at the time were largely focused on improving availability and intensely focused on lowering costs across the board. For example, labor arbitrage -- in the form of lower-cost IT workers -- was a very popular conversation. The desire for agility in its modern interpretation rarely manifested itself. And the great recession of 2008 was still a year or more away.
You can color your personal recollections to suit your tastes, but the point is clear: an awful lot has changed in the last six or seven years. And it would not be unreasonable to assume similar magnitudes of change (or more) in the next six or seven.
Starting At The Top
Through a business-school lens, our discussion should ideally begin with views on how businesses and organizations are evolving, which in turn drives a better exposition of how enterprise IT organizations will need to respond in kind.
There are many good discussions about how business itself is evolving to become essentially information-centric.
"The Third Platform" is a reasonable model; my personal preference is to describe future business models as digital business models: characterized by key business processes that have been re-invented entirely around shared information flows vs. merely being a bolt-on to established processes originally envisioned in the physical world.
A different way of saying the same thing: in the emerging business world, IT becomes the modern factory at the core of each and every value proposition the organization ever hopes to offer: differentiated products and services, superior customer experiences, effective sales and marketing, lower operating costs, better regulatory compliance, etc.
Every business imperative is ideally re-constructed entirely around new information flows and data models.
Frequent readers will recognize that I believe the analogies between future-state IT and modern manufacturing are quite powerful.
There are amazingly strong parallels in organizational models, key processes and metrics, notions of value-add and much more. I think the next six years will put this view in the forefront where it belongs: IT seen as the factory powering newer digital business models.
If we boil down a future enterprise to two key assets: brand and information, enterprise IT plays the essential role in maximizing the second, and directly impacts the first.
Modern Enterprise IT -- Re-Imagined?
Taking this view, enterprise IT could then be seen as the maximizer of information assets, much the same way that finance maximizes financial assets, human resources maximizes people assets, R&D maximizes engineering assets, and so on.
Without any disrespect to these other disciplines, I believe getting information right will end up being far more important.
Given this view, applications might be seen as simply portals for acquisition, value-adding and dissemination of organizational information (wrapped in a bit of business logic), delivered in the context of specific business processes.
The information persists, applications become disposable and infrastructure is ultimately fungible.
Organizationally, this shift will dramatically impact how enterprise IT is viewed by the business. In this emergent information-centric view, the most valuable IT groups are seen as in-house consultants: working with business stakeholders to fully exploit information assets and designing entirely new workflows and processes around information -- and not simply delivering a well-defined project.
Client-facing IT pros will want to be adept at understanding business specifics, suggesting information resources and application models in context, helping the client move from whiteboard to proof-of-concept as quickly as possible, and scaling when needed.
This is most certainly not the norm today, but I'm betting it will be within the next six years.
Current Themes Moved Forward
Clearly, big data in all its forms becomes a relevant discussion, as does mobile, rapid app development, cloud-like infrastructure, etc. That much appears obvious -- both now, and into the foreseeable future. But there's more going on here, I think.
When I have an opportunity to fully consider a core business process re-imagined in a digital business model context, I am always stunned by the complexity and scope of the "application". They are not neat and tidy standalone entities, more like vast organic molecules that interact with other applications and business processes.
They are frightfully complex entities, and I'll think we'll see many more of them.
A competitive enterprise tends to be fully integrated across multiple business functions; it makes sense that their application chains and workflows will inevitably reflect their business model -- and that has all sorts of interesting implications.
One conclusion: the "big view" of how all the pieces fit together will reside in only one place -- enterprise IT -- so, in essence, IT is uniquely positioned to understand the underlying business blueprint behind the organization's operational model.
Want to understand how the business really operates? Talk to IT.
Not the norm today, but likely to be so before too long.
In this view, information becomes a critical asset for any business stakeholder: everyone is going to want access to everything for any purpose. As a result, information governance (and security) becomes far more important in this view -- needing the same sort of organizational maturity as we'd find in, say, finance or human resources.
The trick will be to balance value-creation against managing risk -- not an easy task, but it will inevitably fall to future-state IT organizations.
Own Your Customers, Or They'll Go Elsewhere
A big theme in current enterprise IT discussions is ITaaS -- IT as a service.
The idea is simple: give your internal customers easy-to-consume IT services that compare favorably to external alternatives. But that implies that you've invested in knowing what your internal customers want, and making sure they know how to consume it intelligently -- sales and marketing, if you will.
But the strategic challenge gets a bit more difficult when your internal customers start looking outside for nice chunks of external business logic -- SaaS if you will. Salesforce.com is an excellent example, as is Workday and many others.
While these are all wonderful services, they do equally well for all comers. It's hard to get a differentiated competitive advantage from these external services, *unless* you're prepared to consider them as just another hunk of external application logic, and are prepared to own both the information as well as the integration back into your own differentiated business processes.
Progressive enterprise IT organizations will need to be able to draw the big picture for individual business stakeholders, and show how their use and generation of information from within their applications interacts with the business as a whole.
It won't be enough to own the infrastructure -- actually, that becomes less interesting over time -- it will become essential to own the relationship between their standalone business logic (and its associated information) -- and the organization as a whole.
Agility Becomes A Way Of Life
Digital business models are inherently agile, with time-to-react measured in hours/days/weeks vs. months/quarters/years. Long-term planning becomes investing behind educated guesses about likely scenarios and required capabilities, vs. waiting for well-articulated requirements.
This required agility will reflect itself throughout the enterprise IT function: the ability to meet at short notice with business stakeholders (your clients), the ability to quickly agree on short-term steps to validate hypotheses, using quick sprints to gain feedback and knowledge, adding integration, scale and robustness as needed, and then tearing it down and doing it all over again.
Sounds more like consulting than traditional IT, doesn't it?
Business empires have been built around simply being more agile and responsive than the alternatives; it makes sense that this will apply to progressive enterprise IT organizations as well.
Core Vs. Context
The other side of this coin is to discuss what enterprise IT organizations will be doing *less* of in the future, as they focus on a more business-critical role.
Over the next six years, I beleive vast tranches of traditional IT responsibilities will be up-for-grabs for externalization through one mechanism or another.
First up: messaging and collaboration, as well as the whole "office productivity" thing and sourcing the devices we use: desktops and mobiles.
Other potential candidates: infrastructure resources, security monitoring, big hunks of enterprise application functionality, monitoring and reporting, and so on.
Compared to other modern organizational functions, I see that most enterprise IT groups try and do too much, and -- if they're going to rise to the level of business consultants -- something has to be left behind. Expect plenty of emotional discussions over the next few years.
I would offer that you can't move forward unless you're prepared to leave something behind.
The Lines Get Blurred
The best integrated business models result in environments where you can't really tell who works for what function -- people are coming together to solve a business problem, bringing expertise from their respective disciplines.
In this view, the very best IT functions will become essentially invisible and transparent. They will provide unquestionable business value, and it's just assumed that they're the best available. They will be seen as business leaders with strong IT backgrounds vs. members of a standalone IT organization.
Good consultants have this way of blending in with the team. I think the same will apply to client-facing IT pros.
The New Control Points?
Another way of looking at future enterprise IT is in terms of control points: where the new leverage will likely emerge, and where it won't be.
Clearly, access to information assets in context becomes an essential control point -- one where both extreme value can be added, or ugly friction created. Owning the interrelationships between key business processes and their associated information flows becomes a closely related control point.
Designing and delivering the platforms that make IT services (and their information) easy to get to, easy to develop applications on top of, easy to scale, easy to protect and govern -- also becomes a key control point. And, finally, understanding the business blueprint becomes a relatively new unique value-add that IT can now bring to the table.
But many of our familiar control points become less useful in this view. Budget, for example. Or being the only source of application development. Or providing access devices. Or making it painfully hard to access information for legit business purposes.
But still, trade-offs will have to be made ...
The Evolved Mediation Role
In any marketplace, there's a need to balance demand and supply.
Envisioning a future-state organization chock-full of digitally-literate business stakeholders who have no shortage of wonderful things they'd like to do with information, and applications and supporting infrastructure -- well, there will no shortage of demand.
But the supply side will inevitably need to be constrained. Just like we can't spend infinite resources on headcount, etc. the business has a legitimate interest in constraining demand, and applying resources where they'll deliver the greatest business impact.
Yes, most enterprise IT organizations do this today -- prioritize requests, etc. -- but this will need to be a much more dynamic and complex function in the future. For example: I need 200 nodes to analyze this data *right now*. How important is the request? What should be de-prioritized to accommodate?
We'll have much more agile infrastructure in this timeframe (think fully-evolved SDDC) that will potentially enable us to make low-latency decisions as to where the resources (infrastructure, people, etc.) need to be applied right now to gain the biggest impact. And, of course, that requires an intimate understanding of the business picture -- micro and macro.
And no one will accept that you have to have a bunch of meetings to make a decision ...
So What Will Be The Key Roles?
So what IT roles become are likely to be more important in 2020, given this view?
First, anything that faces directly against a client becomes much more important. Without demanding internal IT clients, there's progressively less rationale for the internal IT function.
These people should be prepared to bring both non-trivial business knowledge, as well as a broad context of what's available and how it's being used.
They are problem-solvers for their clients -- part of the extended team.
Second, the platform builders take on a special role. We no longer can tolerate isolated environments, nor data stovepipes. If it doesn't integrate, it won't be of much use.
Whether we're talking data fabrics, infrastructure, security, governance, mobile applications, etc. -- the people who can create re-usable and extensible value through platform engineering will likely earn exceptional favor.
Third, enterprise architecture will move sharply in the direction of information repositories and information flows: largely separate from application, infrastructure, device, etc. A core set of people will have to have their heads wrapped around the underlying business blueprint for the entire organization, maintaining its integrity and extending its functionality across platforms owned and consumed externally.
Perhaps think of this team as owners of the information platform?
Continual process re-engineering will become a prized skill set -- this is true in other business settings, and will likely be the same in enterprise IT. Some of the processes will be within IT itself; the more important ones will span the boundaries between IT, their clients and their suppliers.
If you didn't see yourself here, not to worry. I believe there will still be plenty of demand for knowledgeable practitioners -- but breadth will likely be given precedence over depth.
If you're looking for a more technology-centric discussion, I don't want to entirely disappoint you. A few of my default assumptions over the next six years look something like this ...
Infrastructure becomes largely transparent and fungible: whether a private, public or hybrid cloud. By 2020, we'll all be much more comfortable with this model, with many great choices to consider and robust integration planes between domains. Behind this, SDDC concepts will be well-evolved and integrated: dynamic control of infrastructure and supported services, aligned at application boundaries.
The internet of things becomes de-facto conventional wisdom. Big data moves from something new to something proven, powering multiple aspects of the business, and fully integrated as part of the application landscape: call it the analytical enterprise or whatever you choose. Sourcing rich data streams and harvesting them in near-real-time will become a discipline unto itself -- a whole new meaning for the familiar "data mining" ...
Mobility becomes the expected norm; static knowledge workers evolve into a very specific use case. The basic mobility model will likely extend into wearable tech (glasses, watches, etc.) but that's simply a variation on a theme from my perspective.
If you're into hardware, the components will get inevitably faster/cheaper/better/denser. I think we'll see more of both dis-aggregated architectures at larger scale, and tightly converged at more modest scale. From a storage perspective, we'll pervasive use of flash for primary storage, with multiple flavors of flash in multiple locations, and disk acting as the persistent capacity store vs. primary storage.
There's much more we could talk about, but that's not the focus here :)
An IT Leader Perspective
One of the most rewarding parts of my career has been working with strong IT leaders who see themselves as agents of change. They understand the model is broken, a new one is needed, and they see it as their role to make that happen.
We inevitably find plenty to talk about ...
My argument here is simple: you're not done yet. Yes, you've re-invented the basic mechanisms by which IT produces services, but that's just the beginning. The road ahead has new potential and few new perils, and it doesn't show any sign of ending anytime soon. So enjoy the ride.
And for those of you who haven't yet dived into the deep end of IT transformation yet -- well, time is short.
A Personal Perspective
I now work at VMware, and am exposed to no shortage of powerful technologies being developed, both in isolation and across disparate domains. Rich, heady stuff.
One of my personal goals is to attempt to figure out the new operational models that are inevitably implied by the new technologies, and -- more importantly -- how these new capabilities can change the equation of what is achievable in enterprise IT settings.
But -- ultimately -- it boils down to people's willingness to use these new tools to remake the world around them. And, when you get talking to people, there appear to be no shortage of barriers to change -- real or imagined.
A quick story?
We own four dogs, and use an invisible fence to keep them on our property. A buried antenna triggers a progressively nasty shock from a battery-powered collar. You quickly learn not to carry the collars around when you're working outside.
Three of the dogs always respect the implied boundary, whether they're wearing a collar or not.
The smart one has learned that when the collar isn't on (and no one is looking), she's free to romp through the neighborhood and have a wonderful dog adventure.
Boundaries are what we make of them. Don't assume the collar is on. Personally, I'd like to see many more IT leaders go for an adventure in changing the world around them.
The time is right.
Like this post? Why not subscribe via email?