Last week I brought up the concept of ‘the next killer app’ and asked you what that might be. Many of you weighed in with your own thoughts, which are always my pleasure to share.
The perplexity of life arises from there being too many interesting things in it for us to be interested properly in any of them.Gilbert Keith Chesterton
Dear Vorticians,
Last week I brought up the concept of ‘the next killer app’ and asked you for your thoughts on what that might be. I posited that complexity is the killer app – both because it is a killer of apps and innovation, and because reducing complexity and helping customers manage it represent a major business opportunity that a wide array of companies are embracing.
Many of you weighed in with your own thoughts, which are always my pleasure to share.
Vortician Dave Burstein replied with a short, intriguing message and I won’t spoil your click. Let me know what you think the implications are. “John, next killer app? Look at https://upload.video.google.com/ and think about the implications.”
Interesting question, Dave. What do you think are the implications?
I got the longer following note from Vortician Umang Gupta, CEO of Keynote Systems and one of the classiest folks I’ve had the pleasure to meet in this business. (You’ll just have to trust that my compliment is genuine and not in any way related to how Umang launches into his thoughts.)
“John, great missive!
“I’ve always believed that the true reason for ‘software as a service’ is not just to achieve lower costs but to reduce complexity (hence lower total costs of ownership).
“When I started in the business 30 years ago, with IBM and a plethora of also-ran hardware vendors all selling their own proprietary IT hardware/software combos, the reason people found IBM a safe choice was because the company was able to throw in a lot of free technical personnel (read: SEs) to help its largest customers make the complex software go down easier. Then along came the independent software package industry, and their value was obvious — they eliminated the complexity by offering easy(ier) to use software packages running on bare-bones DEC minicomputers at a fraction of the cost of what it took to customize the proprietary hardware/software solutions from guys like IBM and Burroughs. Then came the PC and client/server revolutions and the software became cheaper and even simpler by an order of magnitude.
“But the enterprise software package industry took a wrong turn in the mid-1990s. In an effort to get a leg up against competitors in an increasingly brutal software battlefield, software companies large and small started to extend their software into ‘platforms’ not ready for prime time combined with in-house consultant services — with the whole thing being touted as a ‘solution’. Of course, they forgot to tell their customers that the customizations would cost way more than the original software in consultants fees and in-house IT personnel.
“This trend was fuelled by the large consulting companies who saw an opportunity to make lots of money from this complexity, and the Internet-driven venture capital explosion that gave birth to a whole bunch of software features masquerading as companies. It wasn’t too long before the standard business model of an enterprise software company had stopped delivering on the promise of easy-to-use low priced software packages. No wonder the software industry is in trouble today. It is falling of its own weight brought on by years of binging on feature-creep and complexity.
“The software-as-a-service model promises a return to the original promise of the industry — low upfront costs, flexible usage, no shelf ware, and innovation that truly has value to a customer. That’s why I love doing what I’m doing at Keynote!”
Another CEO Vortician, Bob Dutkowsky of Egenera, weighed in with this: “John, kudos on your conclusion that reducing data center complexity is our industry’s next (or even current) killer app. We hear the same feedback from our customers and prospects. At the end of the day, all they want to do is run applications that make their people more productive and their businesses more competitive. Trouble is, they spend incredible amounts of time and money trying to coax this value out of today’s IT infrastructure.
“For example, we recently analyzed a 1,000-node [storage-area network-] connected grid for a prospect and found that, in addition to the 1,000 servers, nearly 18,000 other pieces of equipment were deployed. It’s this supporting ‘glue’ that is choking today’s data centers and creating the fault-prone, high-maintenance, costly and unresponsive IT environments you describe.
“In terms of where the answer will come from, I like your phrase ‘cashing in on concepts’ to describe where IBM, HP and other major vendors are today. As you no doubt know, Sun recently downscaled its ambitions for N1 yet again, HP killed off UDC a while ago, and IBM has yet to deliver On Demand. From what we’ve seen and heard, these visions are more ‘slideware’ than anything else, and customers are disappointed. The problem, we think, lies in these vendors’ refusal to acknowledge that the underlying architecture of today’s servers is flawed. (Since they collectively sold enterprise IT shops the billions of dollars worth of gear they’re now tripping over, it’s not surprising they don’t concede this point.)
“We would argue that you can’t foster simplicity by adding yet another piece of hardware or software to an already complex environment. The killer app must be more than an overlay that tries to mask complexity. Instead, we maintain that simplifying the data center requires a new, fundamentally different server design purpose-built to solve the complexity conundrum. As we’ve seen time and again in our industry, innovation rarely comes from large incumbents. Instead, it’s smaller companies like the four you mention (Egenera, VMWare, Azul and Cassatt) that are unencumbered by legacy installed bases, and can drive the disruptive change needed for significant improvements. When it comes to unraveling the complexity knot and therefore delivering the killer app, we expect history to repeat itself once again.”
And, finally for this week, Vortician Bill Baker had this to say: “(We attribute) way too much importance to individual killer apps (1-2-3, VisiCalc, mobility, etc.). The real killer app for technology has not changed since the invention of the wheel, and that is: staying ahead of whatever is chasing you by any means necessary. You can say the same thing for the car, radio, TV, printing press, the pill, etc., etc., etc. Applied innovation, that’s it!”
In reference to my comment about my limited skills at identifying killer apps – I cited my forecast of killer appdom for asynchronous transfer mode (ATM) – Vortician Baker added. “By the way, even though ATM did not succeed as a network architecture, it was quite successful as an embedded technology (i.e.: most routers used ATM chips somewhere in their design).”
Bye for now. As always, speak to me about killer apps or anything else that’s on your mind, at jgallant@vortex.net.
