Did you know that Facebook has more than 1.2 million servers in operation? And that they are run by 100 employees? This means that the new standard for operations is 12,000 servers per employee. Does it sound OK?
Some IT-people I know call it exciting. ‘Interesting but not relevant’ is a more typical response. Followed by a tirade of reasons as to why this metric doesn’t apply to their particular environment. ‘A completely useless standard’ is one of them, ‘we’re not like Facebook, we have very diversified requirements’. Really? And Facebook doesn’t? Isn’t it at least interesting to learn more about what Facebook (and many others) think and do that delivers such incredible numbers? Are they really so special that the knowledge is worthless to us? Or are we the ones who are so special that we cannot reap the benefits of industrialized IT?
Such numbers scare IT – thus the defensive reaction. Of course we all have something to learn, inspiration to gather. In fact we’re all very much alike whether we’re running Facebook’s datacenter or still cling to our own infrastructure – or looking at entirely different contexts. We’re consuming electricity, breathing air, driving cars and using digital aids all day. Made possible by industrialized production and delivery of basic services and resources. Including IT services. However – as figures from Facebook and a number of really large industrial players indicate – general industrialization of IT is severely lagging. IT services cost too much, are too rigid, often architecturally old and ‘operated’ on platforms and in regimes that expired a long time ago.
It’s tempting to embark on yet another round of arguments about industrialized IT, service focus, resources on-tap, Cloud and real cost, but they are all the same as last year and the year before that. Facebook’s impressive numbers only reinforce them. Despite the strong business case for industrialization and the fact that it has been more than 12 years since Nicholas Carr published ‘The Big Switch‘ which presented the natural forces driving the change in a very understandable way. Still, a surprisingly big part of the IT market is moving slowly. Focusing more on ‘not rocking the boat’ than on costs and opportunity, more on stability than dynamism and flexibility.
This inertia doesn’t change by repeating old arguments, no matter how objectively correct they may be. As has been pointed out in other contexts, the problems have little to do with technology and a lot to do with mindset. Thus the simple message to the market – and the slogan I chose for this blog-site: You don’t need a new plan, you need a new mindset. A mindset stuck in yesterday’s ways and last year’s requirements, will block any attempt to synchronize with reality.
The essence of the new mindset IT must take on, is anything but revolutionary: Data. Forgotten over the years, but data is the reason IT exists, not the other way around. A misperception that was allowed to evolve when computers were ‘big iron‘, extremely expensive and far between, while data was (seemingly) cheap if not necessarily abundant. A perception that suited IT so well that it was kept even after it was obviously wrong.
Further, while infrastructure and tools are now commodities, data is now more than ever the real value and key to any business. IT is the enabler, the custodian of the data and responsible for aquiring, developing and delivering the services that make the data useful and available. It must take on that role fast – or risk getting moved aside by groups and forces that see and understand – and act on – what the organization needs.
Everyone lives off of data. No data, no business. Underutilized data, slow business. A fact that makes it even more odd that the basic resource, computing power and basic services – off-the-shelf resources – still seem to be the primary focus of many IT environments. The imbalance gets exacerbated by the fact that everyone is in a crunch for a different set of resources – short on time and heads, strained budget, too little flexibility (no headroom, constant overload) and wrong/outdated competence. And not the least – that everyone needs improved data flow, flexible integrations, faster adaptation to new requirements, better data utilization, a new generation of ML tools and a better understanding of data as a resource. Everything is data driven.
At the same time, as IDC observed in a mailing the other day:
80% of European CEOs admit they are under pressure to demonstrate value of digital investments.IDC
The numbers from Facebook’s data centers are not a metric or a yardstick, but a reminder: IT needs a new mindset. Data focus. My slightly provocative headline is seriously, if not literally meant: Goodbye IT, go data. And – in order to speed up the process, find a new name. ‘IT’ obviously does not work and is overloaded with history. If ‘data department’ doesn’t taste good enough, find something else, avoid ‘information’ and ‘technology’ which both give wrong associations. ‘Digital operation and development’ has been used – it’s a bit heavy, but clear. The ‘digital department’ is also used, but is confusing, ‘digital services’ is better.
In discussions with IT managers, we often complete a full circle and end up back in the ‘data department’ fold, which is short and concise.
In case ‘data dept’ triggers associations back to the 70s, don’t worry about the name. You need to worry about lowering the average age. Quickly.