What’s Coming Out of the Wall?

by Andrew McAfee on May 20, 2009

I’ve spent the past three days drinking from the geek (remember, that’s a term of praise) firehose at MIT. Monday was a Center for Digital Business workshop on IT-fueled business experimentation organized by Michael Schrage and Erik Brynjolfsson (with a tiny bit of help from me). I’ll blog more about this event later; I’m still wrapping my mind around what I learned.

Tuesday was the CDB’s annual sponsors meeting; I gave the lunchtime keynote on Enterprise 2.0, but had a lot more fun listening to presentations on the wide variety of research conducted at the Center.

I tweeted about both these events, as did McKinsey’s Michael Chui; our hashtags are #mitee for the experimentation workshop, and #mitcdb for the meeting yesterday.

Today is MIT’s annual CIO symposium, and I just listened to a panel discussion among Erik B., Tom Malone (founder of the Center for Collective Intelligence), and Jeanne Ross (director of the Center for Information Systems Research) on “The Future of IT.” The panel was moderated by Gary Beach of CIO Magazine, who asked more than one question about whether cloud computing was that next big thing.

The cloud seems to be on everyone’s mind these days; the symposium has a panel dedicated to it this afternoon. Most folk, myself included, think that whether or not cloud computing deserves the title of Next Big Thing it’s a pretty big deal, and will have a large impact on IT producers, IT departments, and CIOs in the years to come.

But what about CEOs and other senior line executives? What, if anything, does the cloud mean for them? A question asked by Gary this morning highlighted one view of the cloud’s impact on top management. He asked attendees to raise their hands if their company had a CEO. Of course, most hands went up. He then told them that the ‘E’ he was referring to was ‘Electricity.’ We laughed and lowered our hands.

Gary’s point was that stuff that becomes a commodity –  that becomes undifferentiated, universally available, and cheaper and cheaper over time –  ceases to deserve attention from the very top of a business. Electricity is vital to business, and at one time let to a massive restructuring of manufacturing industries, but we no longer have Chief Electricity Officers. Why not? Because the time when it ‘changed the game’ of manufacturing is long gone, and it’s now just a commodity that comes out of the wall.

Many cloud computing advocates consciously or unconsciously use the imagery of electricity to describe the shift now taking place. Companies are ‘plugging into the cloud’ and computing is becoming a ‘utility’ delivered from a ‘grid’ of service providers. The conclusion is clear: that IT resources like processing power, storage, development environments, memory, and applications (even integrated ones) will soon come out of the wall reliably and in the required amounts, just like electricity does now. And the consumers of these resources will need to give no more thought to them than I do to the current that cooks my toast in the morning.

This imagery is simple, intuitive, and seductive. It’s also, I believe, deeply unhelpful and counterproductive. The problem with it is that it will lead to conversations like the following:

“Hey boss, this powerful IT stuff now comes out of the wall, just like electricity does. Should we get some of it? It’ll be more reliable and cost less than what we’re doing now.”

“Yes, make it so.”

“OK, what kinds and how much?”

“Huh? There aren’t ‘kinds’ of electricity, and I thought the whole point was that it would come out of the wall in whatever quantity we required. So just plug us in and stop bothering me.”

My main problem with the IT-as-electricity metaphor is that it provides a lovely justification for business leaders to spend less of their time and attention on IT-related issues. I hope all readers are by now thoroughly sick of hearing me say that I believe this to be a very, very bad idea, and a great strategy for getting throttled by your IT-savvy competition at some point down the road.

But cloud computing is definitely moving us toward having something come out of the wall. If electricity’s not the right metaphor for it, what is? As I was sitting listening to the panel this morning I challenged myself to do better. Here’s what I came up with:

Cloud computing is moving us toward a world in which some kinds of business change come out of the wall. Not all types of business change will come out of this wall, but some will. A few of these will be easy, most will be at least somewhat difficult. Some will be profound, some trivial. The different types of cloud-enabled change will affect various constituencies, both internal and external.

The reason I like this imagery better is that it will (I hope) lead to conversations more like:

“Hey boss, thanks to cloud computing some types of business change are now coming out of the wall, just like electricity does.  I’m going to go ahead and plug us into the cloud, OK?”

“OVER MY DEAD BODY!!!”

“Why not?  Don’t you want to change?”

“Of course I do, but I want a BIG say in defining the changes this organization signs up for, and I also want to lead them. Now, what exactly is coming out of the wall, and what’s it likely to do to us?”

“Well, that’s a longer conversation.”

“Well, let’s start having it.”

The imagery we use to talk about big trends like cloud computing is going to shape perceptions, so we’d better be careful with it. I am deeply uncomfortable with imagery that portrays IT as anything like a commodity or utility, because busy businesspeople in most industries don’t give a moment’s thought to commodities and utilities — to stuff that comes out of a wall.

They all care, though, about innovation and improvement and refinement and disruption –  about the many flavors of business change. I think we technology enthusiasts would be well-served if we started talking in those terms, instead of using language that will relegate our offerings to the utility closet.

What do you think?  Do you like the IT-as-electricity imagery?  If so, why?  And are there any problems with my IT-as-change imagery?  Leave a comment, please, and let us know.

  • Ian

    Your over simplifying the electrical model – I think us practitioners do understand that after maybe 20 years we will have standardization of functionality provided by “The Cloud” – ex. functionality you want has already been built and we simply need to “plug” it in and use it (for a fee). The electrical model does work but remember we are myopic in our thinking that all electricity is 110v at 60Hz. The keys to the model are being held by vendors who pervert the cloud and SAAS vision with license models which do not support transaction or instance usage.

  • http://www.meetingtrader.com/ Stephanie Smith

    I like the IT-as-change imagery much better than IT-as-electricity. As you explained, the IT-as-electricity implies that the use of the next step in IT will be simply to use it – not why or how in relation to the greater goal of the business. Cloud computing will be useful for many businesses, but in different ways. Will there be a fundamental framework that evolves out of cloud computing, as would follow the IT-as-change imagery? i.e. All offices need electricity to power the same things: lights, telephones, appliances, computers….but not all offices have the same equipment, and may have additional things needing electricity based on industry (i.e. manufacturing warehouses). I think this will be the shape of cloud computing, once the initial framework of its use is determined. How it will be determined, however, is another conversation!

  • Adam Prewett

    Spot on sir. “Cloud Computing” does makes IT available in a new format, but does not commoditize it. The winners will be the companies who can leverage on demand compute and storage for the correct purposes (accelerating development and trial, using it for seasonal compute needs) will benefit the most.

  • Alex

    I think the IT-as-electricity image is useful if you’re more specific about IT. The parts that are quickly moving towards being a commodity are everything on your list (processing power, storage, development environments, memory, and applications) except applications. The “plug-in-and-get-it” certainly applies to bandwidth, processing power, storage, and even remote-hosted code and apps, but it breaks down with the actual use of those resources. We all get electricity from the wall, but we all plug different things into it. My spouse needs a hair dryer. I’m bald, and do not. BCG needs a certain type of accounting software, but a mom and pop just wants a ledger.

    I think that the ease of app development vs. the difficulty of manufacturing material goods will preclude software from being turned into a commodity; it’s just to easy, and having just what you need is priceless. Consumer appliances, however, have been hemmed in by manufacturing restrictions to only have a finite number of options: pick 1 of 5 hair dryers, and make do. Choose one of 28 ovens for your bakery, and build your business around that. They’ve forced the behavior to adapt to the appliance. We’re lucky enough that our software can easily be adapted to us.

    As an app developer, here’s hoping that there’s always a need to bridge the gap!

  • Gordon Haff

    >>Do you like the IT-as-electricity imagery? If so, why?

    No. For pretty much the reasons you state. In addition, the whole cloud-as-electric-utility metaphor bakes in the assumption that the external utility is always going to be cheaper than the internal utility and suggests that you get increasing economies up scale up to some very high scale point–which remains IMO an unproven assumption.

    Processing cycles are closer to the electricity analogy (with some problems) but that sort of “peer-to-peer computing” isn’t really where cloud computing is focused and isn’t really true to the idea of buying standardized services that you don’t need to worry about anyway.

    I’m not sure about the “some types of business change” term. Maybe business processes is closer to the mark. For one thing, business processes imply a mapping to internal organizations and workflows. It’s also a fairly logical extension to outsourced services such as payroll or 401k administration.

  • docgeek

    I think I appreciate the problem you are addressing by reframing the metaphor as you did.
    My initial thought was to think about 'your toast' – where does that fit in the new metaphor?
    Or prehaps the 'toaster' – neither image of something coming from the wall is very effective without the object being 'plugged' in. The IT-as-change for me is IT-as-enabler so plugging into the cloud is giving an enterprise radically more capabilities to change.

  • splint

    I would agree with you on the IT-as-electricity metaphor. However, I would take this a step further and point out that are businesses losing competitive advantages because they treat electricity solely as a utility that just “comes out of the wall”? When thinking about data centers, why do we worry about power redundancy, backup generators, UPS systems? Because in these cases we cannot treat electricity as just a utility. The same thing could be applied to IT as well. Cloud computing does offer some great advantages, but if you treat it as just “coming out of the wall” and do not take an active role in it, then when it fails to “come out of the wall”, it could put a business at a competitive disadvantage.

    I think your metaphor fits well with how we should treat IT. That is, businesses should evaluate their positions and decide the cost and benefits of a cloud solution versus a non-cloud solution. Those areas that are best served by cloud computing should be placed in the cloud. The areas where it doesn’t make sense to have cloud computing should not be put into the cloud. Also, businesses should be constantly evaluation their positions and change if necessary.

  • Kishor Desai

    As you have rightly pointed out, everything follows a cycle from being ‘specialized’ to ‘commodity’ over a period of time. What is special today will become commodity tomorrow. Some business supporting applications (not just infrastructure like processors, memory) will be commodity on day 1 of the cloud (Day-1 is symbolic). While applications that support more core business processes will continue to be specialized (for whatever reasons like competitiveness). But, once someone finds out a cost effective (read commodity, mechanical, de-skilled) way of achieving the same business objective, without compromising on competitiveness (read profitable, faster to market ahead of competition, superior servicing capability), they too will become commodity. And, this cycle of innovation continues (it has to because it is got to do with the way humans are built. We always like to solve new challenges. But, the day a challenge is conquered, it looses interest and becomes commodity. We move on to the next one). So, in my opinion, both imageries are relative (and not absolute) and hence will have their own place of existence in ‘time’. Technologists will have to stay ahead of this and be able to advise business leaders on how to make something ‘electric’ by using/leveraging ‘change’.

  • eriklue

    Mr. McAfee,

    I thoroughly enjoy reading your blog. I'm 24 years old, planning on going to business school in a couple of years, and work at salesforce.com. I really enjoy reading your continuous discussions and theories of where Cloud Computing will drive us.

    I agree with you that businesses who adopt Cloud Computing cannot flourish without a healthy IT organization. My current role at salesforce is to ensure that business development is being driven by our customers to take full advantage of our platform and resources that we provide in our software-as-a-service model. That being said, our successful clients are those with a great IT group, that have resources and knowledge and help the business achieve their vision. Without a strong IT group to understand the capabilities and limits, the vision is convoluted and never really clear, thus often times not fully achieved.

    We at salesforce encourage our client's IT groups to concentrate on innovation. Because the cloud computing model offloads the traditional responsibilities of an IT group to the cloud, their main focus should be on development. Cloud computing offers a scalable solution that benefits all, but you can only get out what you put in.

  • http://www.macroprinciples.com Julien Le Nestour

    Hi Andrew -

    Long-time reader, first-time commenter.

    I'm in charge of IT Innovation for a large group. I'm using the following metaphor (Harrods with some very specific twists) to illustrate the changes implied by the various technologies under the “Cloud Computing” umbrella. I would be interested in any comment as well (here or on the original post):

    “In the world you operate today, imagine that literally all business managers, from CEOs of public companies to sole business owners, including would-be business creators, have a huge supermarket just round their corner. A special supermarket in fact. It sells a tremendous array of products and services, for every industry, need, or activity. It is also ever expanding: if you make two trips at just 1 hour interval, new products will have already been stocked and available for sale, in addition to the old ones and on new shelves constantly being added. On top of that, the price of each product is incredibly cheap: new cars for example, are selling for 1/1000 of their normal price, or even cheaper.

    The last characteristic of this store: you don’t own what you buy, you just lease it. It can be a one-time fee, regular payments, or any other scheme, you lease it. What’s more, if for any reason the provider of your product disappears, then your product disappears as well. Let’s illustrate this with the car example: you can lease a car for, say $30 a month, and use it as you own it. You would then leave your “stuff” inside the car: some books, some papers, perhaps a watch, etc. If the producer or your car disappears, your car vanishes as well, with your stuff inside.

    The trade-off begins to appear: $30 a month for a car is a really good price, but you are at risk of this disappearance. You can mitigate it however: choose a well-known and established vendor, or just leave only non valuable “stuff” inside. Many parallels can be made with this metaphor.It is for example really difficult to find your way in an ever-expanding, huge store, and find the exact product that would fit your needs. Then, how do you know if its producer is not on the verge of disappearing? You can stick to the big names only, but what this product taht would really fit your needs, would you try it?

    You can also see why small and large companies would react differently, with different risk aversion. Large companies would prefer to buy their cars, for example, and benefit from the certainty. Small businesses would buy the $30 a month car, and if it disappears, just buy another with a little it of trouble. You can already see how competitive dynamics would evolve, with small early-adopters buying everything at these stores and competing on equal standing with large companies, while having a risk profile much higher as well.

    In the IT area, Cloud Computing represent such a shift: each business, no matter how small or remote, now has access to such a supermarket instantly. Now, think about your business, and examine how your competitive advantage(s) would be sustainable in this scenario. Some won’t resist.

    At its core, Cloud Computing is nothing more than an irresistible way of enabling distributed economics of scale, that benefit organizations of all size. A second order consequence is a spur in innovative products and services that executives can take advantage of. Conceiving it as a huge, ever-expanding supermarket across all areas can help frame this new context.”

    - Julien

  • http://www.6fusion.com/ John Cowan

    Andrew -

    Great blog post and so timely given the hype in the industry. My company is very passionate about this very subject. I started typing a comment a few days ago and figured it was getting a bit too long, so I made it into a full blog entry over on our site.

    http://6fusion.wordpress.com/2009/05/25/computi…

    Cheers,

    John

  • http://paologalli.wordpress.com/ Paolo Galli

    We have some effect of IT-as-electricity from a consumer perspective, I've got my yahoo account since 1999 and a lot of services are now available, over Internet, the firs big cloud.
    My boss explained that idea on IT about four years ago, we were talking about the concept “on-demand” proposed by IBM. He ask me and I told him “well, there a lot of technical difficulties, but will be a day whe you'll get something similar, and the way we think about IT will radically change”.
    Now the concept of IT-as-electricity is more near to be real, I think only for processing and storage capacity and for some kinds of no-core application, even if there are a lot of problem and questions about security, responsability of the provider, legal questions about storing data ” I don't know where”, which kind of law apply, and so on.
    In my opinion the big stuff (or question) is how (and if) will change the way IT can be a competitive advantage, as you stated in some of your post.

  • http://andrewmcafee.org/blog/?p=851 The Diminishment of Don Draper : Andrew McAfee’s Blog

    [...] got to thinking about these questions during the conferences I attended at MIT last week, which were all about IT’s impact on the business world. The presentations I [...]

  • Joe Pelletier

    Only until the CEO ('E' for executive) can rationalize IT — to the extent a CEO currently understands the use of electricity in the workplace — can technology be depicted through the IT-as-electricity imagery; essentially, marginalized as a commodity in the enterprise. Certainly, some facets of IT, like software and workstation deployment, have become more commoditized over time and give the company little competitive advantage, if any. However, most of today's emerging technologies (like cloud computing) are still rather fragmented (in the sense they are less standardized relative to business processes like software and workstation deployment); therefore, I do not agree with the IT-as-electricity imagery and continue to look at IT as change.

    IT-as-change makes the most sense to me. Sure, I think it is understandable to see future commoditization of cloud computing services like Amazon AWS, let's say, given the economies of scale associated with on-demand infrastructure. However, these are public clouds and I believe much of the shift in cloud computing will move toward private cloud usage. As others have said in this comment roll, companies will use IT and cloud computing to their benefit in different ways; but crafting this strategy is where the CIO's role begins to evolve, rather than disappear (hence, my support for IT-as-change). Data is the true commodity here, but using this data to engineer the company's next competitive advantage is where IT must, and will, persist.

    Though only a recent graduate, possessing (relatively) limited professional experience through internships, I have seen companies design, develop, and implement their own private cloud services in-house to achieve greater efficiencies through collaboration and mass distribution of on-demand, highly-custom software. This is not new, but incorporating the right balance of public-private services and SaaS applications is critical. Obviously, just plugging into a cloud platform without any purpose or direction will not yield many benefits for the business — and this is where CIOs see their next challenge.

  • http://vzach.de/blog Valentin

    I believe you oversimplify by treating all kinds of cloud computing the same. I would like to differentiate between IaaS, PaaS and SaaS.

    First, Iaas – infrastructure as a service. Here you buy computing power and this is well defined by open standards and easily quantifiable. In most organizations this kind of computer power is already delivered by 'commodity computers' (supercomputers are a different story) and it makes perfect sense to treat it as a utility.

    Second, PaaS – Platform as a service. This is a bit more difficult, but in some cases it may also be based on open well defined standards that are implemented by multiple vendors and can also be realized in house (think J2EE) – this too, can be seen as a utility.

    Finally SaaS and specialized PaaS – yes, here the commodity metaphor leads us astray. However even there I'm deeply sceptical about the “IT-as-change” image, since in my experience technology alone never changes an organization, the organizational will to change has to be there first (although this may be driven by the effective IT use of competitors).

  • boston movers

    Cloud computing is moving us toward a world in which some kinds of business change come out of the wall. Not all types of business change will come out of this wall, but some will. A few of these will be easy, most will be at least somewhat difficult. Some will be profound, some trivial.. :)

  • paulbjaylee

    we started to get more nervous when we researched what was going on in Austin. The city is universally recognized as a leader in the area of sustainability and green-building practices. And everyone in the fidelity 401k community is involved: the utilities, the government, local businesses and individuals.

  • http://www.mysecuremovers.com/ boston moving help

    What a great move! I'm lucky that I've visited this blog. I'm looking forward for your next post, so keep on posting!

  • chinesedemocracy

    Thankyou for the article about how business is going digital, I found it really interesting and I might have to implement some of the changes you mentioned in my own business,

    chinese democracy

  • http://www.temporaryhaircolor.net/ Chi Silk Infusion

    I think it is understandable to see future commoditization of cloud computing services like Amazon AWS, let's say, given the economies of scale associated with on-demand infrastructure.

  • http://www.temporaryhaircolor.net/ Chi Silk Infusion

    I think it is understandable to see future commoditization of cloud computing services like Amazon AWS, let's say, given the economies of scale associated with on-demand infrastructure.

Previous post:

Next post: