Blog post

Counterpoint: Why We Should Not Use the Cloud

By Andrea Di Maio | April 20, 2012 | 4 Comments


The IT world has embraced the concept of cloud computing. Vendors, users, consultants, analysts, we all try to figure out how to leverage the increasing commoditization of IT from both an enterprise and a personal perspective.

Discussions on COTS have turned into discussions on SaaS, People running their own data center claim they run (or are developing) a private cloud. Shared service providers rebrand their services as community cloud. IT professionals in user enterprises dream to move up the value chain by leaving the boring I&O stuff to vendors and developing more vertical business analysis and demand management skills. What used to be called outsourcing is now named cloud sourcing, while selective sourcing morphs into hybrid clouds or cloud brokerage. Also personally, we look at our USB stick or disk drive with disdain, waiting for endless, ultracheap personal clouds to host all of our emails, pictures, music.

It looks like none of us is truly reflecting about whether this is good or bad. Of course, many are moving cautiously, they understand they are not ready for prime time for all sorts of security, confidentiality, maturity reasons. However it always looks like they have to justify themselves. “Cloud first”, some say, and you’ll have to tell us why you are not planning to go cloud. So those who want to hold to their own infrastructure (without painting it as a “private cloud”) or want to keep using traditional delivery models from their vendors (such as hosting or colocation) almost feel like children of a lesser God when compared to all those bright and lucky IT executives who can venture into the cloud (and – when moving early enough – still get an interview on a newspaper or a magazine).

Let me be clear. I am intimately convinced that the move to cloud computing is inevitable and necessary, even it may happen more slowly that many believe or hope for. However I would like to voice some concerns that may give good reasons not to move. There are probably many others, but it is important to ask ourselves – both as users and providers – tougher questions to make sure we have convincing answers as we approach or dive into the cloud.

If data is the fuel of the 21st century, why should I fill somebody else’s tank?

Social data, open data, big data: I think it was UK Minister Francis Maude who said yesterday that open data is the new raw material, and Gartner’s head of research Peter Sondergaard used the expression “oil of the 21st century” almost two years ago at one of our symposia.

One could argue that it does not really matter where your data is hosted, as all you need is cheap storage and powerful analysis tools. But is that true? Why do countries build their own oil reserves? Do I risk to concede a competitive advantage by letting my data being hosted somewhere else?

If IT is a key driver for economic development, why should I let vendors in another country thrive?

The cloud is already displacing traditional roles in user IT organizations, and will progressively hit smaller IT providers (including cloud ones) through an inevitable process of market consolidation. The few large remaining vendors will be headquartered and operating in certain countries and not others. So what are we going to do with all those IT specialists that many digital agendas say we need to modernize the country? Will they all find a job in technology, maybe moving up the value chain or inventing uses of IT that we cannot even imagine now, or will they be underemployed or unemployed? Should a state or a nation or a province maintain local infrastructure, local software development, local data management as a way not to lose its technical edge?

If large cloud provider become target of terrorist attacks or just go out of business, should I put all my eggs in their basket?

Many debate the actual or perceived reliability and security of cloud providers. But as they grow bigger and we rely on them for more and more of our IT needs, they become a possible target for physical or cyberterrorist attacks. In the past enemies would try to neutralize your manufacturing and transportation capabilities, but in the future they may simply coordinate attacks to cloud providers to hurt multiple countries at once.

Even in a more peaceful world, events over the last few years have shown that no organization is “too big to fail”, actually do. Could a series of incidents or a financial crisis affect the viability of cloud service providers? And even if we could manage to retrieve all our data, would we ever have enough space to store it all, after years that we have relied on cheap and apparently limitless storage? Would we have the bandwidth to transfer petabytes of data when everybody else is trying to do the same?

What if the Internet becomes way more expensive in the future?

When looking at the business case for cloud, very often people forget the cost of the network. That’s seen as a real commodity, so it does not make a great difference on how much we pay per employee per month. But, in reality, we are assuming that the current model is sustainable, while – with the explosion of remote access due to wide cloud adoption as well as due to market and regulatory changes – it is not impossible that the pricing structure changes when we are already stuck with our virtual assets in the cloud.

The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.

Comments are closed


  • Lou Cypher says:

    The force behind ‘Cloud First’ is financial accounting. It demands that someone provide a financial forecast of IT costs, unit-ized by the org’s operational metrics of choice (e.g. per-worker, per-worker-hour, per-project, per-project-phase, ad infinitum), instead of the metrics of general asset and labor management. It also wants the ability to negotiate risk-exhange with the service-provider, again, according to the business contract (e.g. premium for short-stay usage, discount for long-term occupancy, ad infinitum), not driven by the metrics of general asset and labor management.

    Science is running waaaay ahead of both management and accounting right now. Management, finance and accounting disciplines are attempting to quarantine the financial minutuae/volatility of science-intensive IT operations, and Cloud First has become a convenient market metaphor.

    The dirty work is in getting service-providers to engage in more-complex risk-exchanges with subscribers, rather than the standard No Warranty No Fitness For Purpose Disclaimers that currently pollute market offerings. Separately, Cloud First is not geo-agnostic: in addition to each sovereign nation, The Euro Zone, The Patriot Act Zone, et al, all represent unique operational areas, with many other jurisdictions overlaying these. Cloud First makes geography a real question: where will access-points be operated, where will data be stored, et al, are net-new questions for subscribers to ask and service-providers to answer.

  • Ken Post says:

    In my opinion anyone who puts mission critical data that is required for daily operations in the cloud without a fail safe local operational alternative is not much of a businessman. There will be failures. One must add in the cost of halting your business for one to five days while a failure is fixed.

  • Christie Wagner says:

    I totally agree with Ken Post’s (April 24, 2012 at 3:52 p.m.). intelligent analysis.

  • AS says:

    I think that the choice of computing environment (in-house, private cloud, etc.) for a particular service should be based on a set of explicit rules. For example –