This post originaly was published as column at ITSMportal
Predicting the future is a lot more fun than analyzing the past, but as Mel Brooks might say “A funny thing happened on the way to the future; it changed from what we expected.” And there have been plenty of predictions recently. For starters, Wired Magazine announced the death of the (browser based) web, predicting it will be replaced by dedicated locally installed desktop or mobile applications – those things we now call “Apps.” As you can imagine, this article prompted a large response by bloggers – and emotions were nearing outrage in some cases. Most of the reaction came from people who simply love their browsers, but one can imagine that many SaaS vendors also had a rough night. Being able to run multiple SaaS applications next to each other, while still offering a rather consistent, integrated look and feel –courtesy of HTML and the common web experience – is pretty fundamental to the long term success of SaaS.
Just a week prior, BusinessWeek ran an article by AT Kearney titled “The End of Outsourcing (as we know it)” in which they predict today’s outsourcers will be rapidly replaced by cloud outfits, in the relentless pursuit of economies of scale. They even go as far as to pick winners (Amazon and Google), potential winners (Oracle and SAP) and losers (today’s outsourcers, especially the midsized Indian companies).
AT Kearney sees today’s outsourcing champions, such as HP and Accenture, as hesitant to become cloud providers. Surprisingly (or perhaps not?) the authors do not mention IBM, by far today’s largest player (HP may be a bigger company today, but mainly because they still sell lots of printers and PCs) nor Apple. Now, you may argue that Apple is a consumer company, but as today’s innovations get introduced into consumer markets first, we could predict Apple to move their innovations into the enterprise market soon, offering enterprise versions of cloud offerings like MobileMe (maybe then called MobileInc?). That is, if the world indeed changes as fast as AT Kearney suggests in the BusinessWeek article.
But that is exactly the issue. Today’s big enterprise IT is just not that agile. Much of what is outsourced today still consists of code that was first written 20 years ago. We saw several companies try to “right-size” their pre-relational mainframe databases year after year, always concluding that it either did not have any ROI, simply was not worth the effort or the risk was too high. And as a SAP executive recently said, many large ERP requirements are still far away from anything cloudy.
Now don’t get me wrong, one sure prediction we can make is that tomorrow will be vastly different from today. In fact, today is already vastly different from yesterday, as Phil Nash pointed out “tongue-in-cheek” in a recent tweet “Welcome to the new decade: Java is a restricted platform, Google is evil, Apple is a monopoly and Microsoft are the underdogs.” But at the same time, companies with expansive IT operations will move slowly, as Brian Stevens CTO of RedHat seems to agree in a recent interview on Bloomberg TV: “It’s going to be several decades before the technology arrives and our [financial services] customers are using the capabilities of cloud more readily.”
We may not directly notice this dichotomy, because our magazines, articles and the enormous flood of social media almost completely focuses on describing new shiny projects (the 20% of the average IT budget) and hardly on the lights being kept on (the 80% of the average IT budget). In fact, the view may be even more distorted, because – as Marcel den Hartog recently described – some of these older systems are so efficient that they run the majority of the enterprises transactions at a fraction of the total IT cost.
Still not convinced? Have a look at Gizmodo’s animated history of the internet protocol or this infograph. It shows how it took almost 50 years to get to the internet protocol, (which by the way, is still changing). And now some are predicting cloud will change everything this year? (As an aside, the infograph made me realize I am officially one year older than the internet, which I guess is why I – unlike Gizmodo – still know what a coax cable looks like!)
All of this talk about predictions reminds me of the paperless office. Remember all the hype and anticipation around that? Funny thing, it never happened. In fact we now print more than ever before (making HP bigger than IBM) and only this year we finally see a device that may get us to this “paperless dream.” Yes, I mean the iPad, and it is not by coincidence (it never is at Apple) that the only function missing from iOS is … a printing function. The other major change attributed to the iPad (and its smaller sibling, the iPhone) is Wired’s announced return of the App mentioned earlier.
Personally, I believe Apps are a much preferred way to consume content, but the average knowledge worker is not paid to merely consume content. Wouldn’t it be great if you could spend your days reading blogs like this one, and be paid to do so? But we’re not. We’re expected to add value by analyzing, combining, mashing up and composing new content, or by putting this content in a new context. Capturing that into a single app sounds a lot like George’s job on The Jetsons. Just press one button and all the rest is automated! (A bit like James Urquhart’s vison of Self Managing Clouds , interesting and good to think/write about, but still far away.)
In the end, SaaS vendors can rest assured; it will be a while before they are rendered obsolete. Likewise for outsourcers. Sure, outsourcers should be thinking about adding cloud services – such as IaaS — to their portfolios. But at the same time, we see the main pioneer of IaaS, Amazon, taking a distinct step back last week – as David Linthium described– by starting to offer reserved instances. These are machines dedicated to one customer for anywhere between 1 and 3 years (which is longer than most modern outsourcing contracts).
Despite all of the ongoing debate, I am convinced cloud will happen for existing applications, but it will likely happen after we stop writing about it (see 4 p’s in a pot innovation ). Today, cloud will grow in new technology areas (for example, almost all social media sites are in the cloud) or with new things we simply do not yet do (like systems that help George Jetson make smarter decisions through massive data analysis and number crunching). And that is not a bad thing. If we need to choose between deploying cloud to make the systems we already have 5% more efficient or to do 5 new things we do not do today, I think we would all choose the latter. But of course, I am an evangelist and not a CFO. The remaining question is, what would or should today’s CIO do? What are your thoughts or suggestions?