Tuesday, February 17, 2015

Death of big projects?

A trip to Las Vegas is a reminder than not all big projects succeed and the story of failed big projects isn’t limited to construction as experienced IT professionals will quickly point out – but the issue today is whether small has become the new big!

This week has me back in Las Vegas for another event, the ATM Industry Association (ATMIA) US Conference, 2015. Under the overall banner of enhancing the consumer’s self-service experience, it has attracted the usual crowd of ATM vendors, payments solutions providers as well as all the service providers on the periphery of the ATM ecosystem. Among the sponsors are some familiar names, and it’s good to see a healthy representation from the NonStop community, and conversations started as soon as I left the registration desk. However, it was the familiar sites of the Las Vegas skyline that really caught my attention this week.

If you recall the last post to this blog, that of February 6, 2015, Of hubs and spokes; of niches, clouds and beyond the horizon; it all looks good for NonStop X! it contained the familiar site of the Las Vegas tower – home to the Stratosphere casino and hotel. The picture was taken from the window of my hotel, but this time, even though I am staying in the same hotel, my window faces the opposite direction and so the picture I have included above is of the view from this window. In case it is not obvious, the image represents an almost-completed site that has been standing abandoned for more than three years. Having changed hands a number of times, there’s still no indication as to when the development will be completed and yet, its primary building dominates this northern end of the famous Las Vegas strip.

It is not alone, however, as across the street and a little closer to the main attractions on the strip, is an even bigger vacant property sitting on 87 acres, the Asian-themed Resort World. With an access road already renamed Resort World Drive, there’s only a minimal number of structures visible with very little evidence that any work is being done whatsoever. At the other end of the strip, across the road from the Mandalay Bay complex, sit a couple of incomplete support structures for yet another Ferris Wheel for Las Vegas, even as the Vegas High Roller Wheel - the world’s largest observation wheel – has been completed and is operational. With rebar clearly rusting, these abandoned support structures are yet another reminder that not every major project in this city makes it to fruition.

When it comes to major projects floundering, many major IT projects have suffered similar fate. Back in the 1970s, during my trip to the US, there were advertisements placed by the Australian retailer, Myers in the American edition of Computerworld looking for experienced IT executives. Describing their need for additional resources for the biggest retail IT project ever, and offering starting salaries well above what any Australian IT manager had seen before, the project called for building two new data centers, equipping them both with some of the biggest IBM mainframes available at the time, and bringing online processing to every store, complete with online access to centralized databases. This was, at best, adventurous for the times (remember, 4800 baud communication’s lines), while at worst, a risky investment of corporate funds. Meyers did make it eventually, but after several delays and a move to distributed processing. More importantly, the financial overruns were fodder for numerous financial commentaries that made headlines even in the mainstream media.

Retailers, banks and government agencies are famous for the big projects they pursue. Many in the NonStop community will recall as I do the huge undertakings initiated by the Australian banks, Westpac (Core System 90?) and ANZ (a new platform proposed jointly by Dell and Microsoft?). These were equally famous for the many decision reversals that occurred along the way. I happened to be in the offices of a Westpac IT executive who calmly told me of the just-made, forward-thinking, decision to go with distributed IBM Series/1 computers at a time when their only use appeared to be to fill up empty warehouses (alongside, of course, the many unsold rebadged Stratus computers – the infamous IBM System/88).

Not everything big that we dream up comes to fruition, nor does the big idea develop into the game-changing, disruptive and innovative solution its proponents promise. However, apart from the news about failed outsourcing projects, of late there has been little news about the pursuit of the big project by an enterprise. So little news in fact that there’s been stories written about the death of big projects, given so few enterprises have the stomach anymore to tackle anything involving multiple years. Selling any project that may outlive the CIO’s likely residence in the corner office is becoming even more difficult to accomplish, so much so that for all intents and purposes, the big project has completely faded from the IT scene.

In the March, 2015, edition of the eNewsletter, Tandemworld, my Musings on NonStop column touches on the topic of big projects. Not so much about the death of big projects, but rather, the flip side of the topic – have all the big applications been written and has the need for big projects become a moot point? That is, with nothing big needing to be developed, are we approaching a time of change rather than a time to reboot with completely new processes? There’s no longer a situation where empty box outlines, describing functionality needed by the business, remain unfilled waiting, as it were, for a new product to appear on the horizon. Furthermore, I suggested, have all the applications been written? And just as importantly, is the science, or indeed art of large-scale application-writing, heading for the trash can? Will we all be witnesses to the death of the development of further enterprise applications?

Continuing with this idea, I then wrote that there will be exceptions, of course, but in general a strong case can be made favoring enhancing what we now have versus looking for what is new. As much as this may sound sacrilegious to the IT faithful, the dawn of the mini-applications, or simply, Apps, is hitting the enterprise and making large enterprise projects that most of us are familiar with, looking oh, so 1990s, to paraphrase recent comments made by a Google executive. Again, “in this age of mobility and information” the core of what we need to do as a business has been done, for the most part, leaving the tweaking of processing and presentation to better serve customers on the move – yes, the always-on world is forcing a modernization of our applications like we haven’t seen, well, since client/server computing, and yes, perhaps since the first distributed computer came onto the scene.

In a recent email exchange with WebAction, Inc. Cofounder and EVP, Sami Akbay, on this very topic, Akbay suggested that, “Traditional applications are dying. These applications used to sit between people and computers. They were optimized for human generated data to be the input. Humans also consumed the output.” And by traditional applications, for the most part these were the very big applications we had seen developed in the late 1980s through to the late 1990s, that anchored much of the IT processing of every enterprise. However, while they remain, Akbay suggested that there were complementary projects underway at nearly all of these enterprises. “The new generation of small applications sit between machine generated data and an infrastructure fabric,” said Akbay. “The input data is predominantly machine generated. Humans consume the output. Also other applications consume the output but that's the next generation ‘connected applications’”.

With so many discussions of late on the topic of modernization, it is often overlooked that modernization is more than just dressing up the look and feel of older, legacy, applications. What was the subject of the big project, decades ago, and remains operational, only requires a new GUI, or so conventional wisdom suggests. For many enterprises this is indeed the modernization projects being pursued. On the other hand, the advent of modern languages and frameworks has also seen incremental functionality being provided using some of these modern languages. This is the subject of an upcoming article in the March / April, 2015, edition of The Connection magazine that I am co-writing with InfraSoft Pty Limited CTO, Neil Coleman. In this article, the merits of using Node.js running on NonStop X and support JavaScript is actively promoted and there’s very little today that is more modern that Node.js.

The landscape of Las Vegas is liberally littered with the evidence of the failed big project. IT history tells us that Las Vegas isn’t alone in this respect – the numerous times that the IT big project failed likely continues to outnumber the IT big project that succeeded. It may be cavalier of me to suggest that the days of the IT big project are over, but by many counts, this may be truer than we realize. Embellishing the applications that keep enterprises in business seems to be the new reality and adding small applications, as Akbay observes, will likely play a bigger role going forward. Having a choice to run new platforms on NonStop X systems only makes going down this path seem even more realistic.

Big projects dead? Just look around your IT department and look at who has the appetite for more big projects and you will readily see that small is now the new big. It may not be as glamorous, and less likely to capture the media’s attention, but then again, as long as baby-steps remain in vogue, as we see today with small applications, then the likelihood of anyone in the NonStop community grabbing the spotlight will be close to zero. Put that down as yet another positive attribute for going with NonStop and one I have to believe every CIO will welcome!      

1 comment:

Mike Noonan said...

Maybe that it is just a lull. I read an article today listing 5 Calif BIG corps that are using their cash to buy back stock for the shareholders and to increase to executive's compensation. And they are laying off 1,000's.

The named corps are Cisco, Intel, HP, Oracle, and Chevron. If them, why not other corps down the pecking order? The article has a social, political leaning about how the middle class is being short changed. This is a 4 part series of problem, impacts, causes, and solutions written by multiple authors.