Monday, December 19, 2011

Steve Jobs in 1980... and Cloud + Big Data Today

In 1980, Steve Jobs gave a talk (YouTube) about the early days of Apple.

It's fascinating because it describes events that lead up to the first commercially-available Apple computer, and presaged the movement toward "canned" programs (commercial software) rather than everyone writing programs themselves.

However, around the 12:00 mark, Steve made a really incredible observation regarding the true goal of Apple (at the time) and about how to use the new level of computing power available to Apple computer users:
"As we move into the '80's, the amount of computational power - the amount of raw horsepower - we can get into a small box for a reasonable price is staggering....   One of the things people always ask me [is] 'what we've got right now is just fine; VisiCalc runs fast enough. Some of the database stuff runs fast enough. What are we going to do with this extra awesome power?'

"The answer to that is that we're going to put it [computational power] into applying/solving that problem again: In other words, we're going to start chewing up power specifically to help that 1:1 interaction go smoother. And specifically not to actually do the number-crunching and database management and word processing. We're actually going to apply that power specifically into removing that barrier.... 

This statement struck me because, in essence, Steve was saying "it's not about faster, it's about easier". He was pointing to all that could be possible if the clunky interaction with the technology was relegated to the background and made invisible.

In today's enterprise context, my interpretation of this is "What could be possible if all of that data center technology could be relegated to the background and made invisible?" In essence I thought, what if we could mask the granular time-consuming operational efforts of managing servers, I/O, applications, networking, storage, security etc. and get back to why the data center is there in the first place?  Today, IT operations for 'keeping the lights on" consumes ~ 75% of IT's budget today, and only ~ 25% is left over for innovation and serving the business.

The goal of IT Transformation should be to "solve the problem again" and put the computational horsepower on autopilot. Let's get to the point where our interaction with the data center is the ability to ask for the resources we need, and in response, we instantly get fast, scalable, secure services.

If we can get to the point where 25% of the cost is to keep things running, and 75% of the cost is used for innovation about the data, what would be possible for business?

In my opinion, IT is still largely in the Dark Ages. We are obsessed with speeds and feeds, tuning and tiering. The purpose of the Data Center is the Data - manipulating and analyzing it for the business. If we found a way to direct 75% of our computational horsepower to THAT, what would be possible?

That needs to be the goal of IT Transformation. That is why Cloud is such a critical enabler.

Thanks once again, Steve.

Other Resources:

Thursday, December 15, 2011

Predictions: What You WON'T See in IT For 2012

While everyone is jumping in with 2012 predictions ('Tis the season) I want to impress on everyone that a dose of sobriety is in order.

IT vendors - and even analysts - are understandably eager to see new technologies and operating models (cloud) adopted quickly. But let's acknowledge the High Tech adage: Not very much happens in 2-3 years, although massive changes occur over a decade.

At any rate, I wanted to highlight things I believe we won't see happen in 2012. Perhaps because we're just being a bit over-eager, or perhaps other enablers have to precede them. Here goes:

An Instant Cloud
Despite claims from vendors, and a plethora of tools and technologies, you still won't see an "instant-on" private cloud solution in 2012. And you probably won't for years to come. I know many folks (mostly vendors) will vehemently disagree with me, but let me challenge you all with this: Cloud is an operational model, enabled by technology. Simply implementing a tool (BTW, most of which are still only tenuously integrated with each other, as well as with hardware and networking platforms) won't solve the problem for you. And definitely not for an enterprise-ready level of availability

Building infrastructure is one thing; knowing how to operationalize it, integrate it into your enterprise, and how to re-structure your service delivery processes are very different. 

However, I do believe that in the next 3 years or so, mainstream enterprises will come to realize that the issue is only partly technology-based. And I hope that change-management and organizational design models will mature to the level of technology models so that "turn-key" process change and skill development will accompany the product sale.

One Dominant Public Cloud
Despite growing use of AWS, Google and Azure, I believe that none of these will be a runaway dominant leader for 2012. Or ever.

Although there will likely be 2-5 very large public cloud leaders who compete on economies-of-scale, I believe the invisible hand of the market will instead cultivate many more "special purpose" community clouds. These players will develop based on knowing their specific market requirements - i.e. competing on security, compliance, use habits, special-purpose applications, performance, etc.  Take for example financial markets (NYSE's Capital Markets Community Platform), Healthcare (varied provider solutions), or even Federal, State, and Local Government (varied initiatives).

The next 2-3 years will definitely see more specific examples of these special-purpose cloud computing initiatives - and perhaps even the emergence of a few 'dominant' community clouds in selected markets. But in 2012, we won't see the community cloud market held back by the presence of large public clouds.

Broad Use of IT Chargeback
Financial chargeback (and showback) have been discussed for years, and are being implemented in greater numbers lately. And although I am a proponent of IT Financial Transparency, we won't see the broad-based use of chargeback unfortunately still won't go mainstream in 2012.

In my opinion, implementing financial transparency tools are second-level initiatives. They don't make sense to implement until and unless the IT department first has other financial controls and metrics in place. And they certainly don't make sense unless IT and the lines-of-business agree on what they're trying to achieve with better IT financial transparency in the first place. Is it really cost-recovery, or merely better knowledge of variable costs and consumption?  Is it an attempt by IT to become more 'competitive' and to measure itself against external providers? Is it an effort by the CIO and CFO to gather better build-vs-buy decisions? The enterprise has to ask these questions before forging ahead with a chargeback program.

So, while 2012 may not be the year of chargeback, it might be the year when IT begins to take a more evolved approach to measuring its variable cost, to metering consumption, and to implementing the goals and strategies it will need to begin these initiatives. Broad-based use of chargeback may still be a few years off, but I hope that IT financial maturity begins soon.

IT is Elevated to a Strategic Business Enabler
Unfortunately, in 2012, we'll still see the vast number of enterprise IT groups continue to report up to the CFO, to be pressured to keep-the-lights-on with less, and to simply be considered a cost center by the organization. As much as IT should be treated as a core enabler of the business, this just won't be so in 2012. It takes time.

But a transformation of IT is taking place, slowly. Saavy CIOs are thinking of themselves as "internal SPs", and beginning to relate to partnering with lines-of-business in a formal manner. To make this transformation, IT first has to adopt new models for services consumption, operations and technology.

In the new context of Business Enabler, IT partners with Lines-of-business to ensure that (a) services are quickly made available to support the top-line revenue needs, (b) IT works with business managers to educate them about potential new services and top-line opportunities, (c) IT adopts a 'consumerized' mindset whereby it supports an "any device" approach to endpoints, and (d) IT is as comfortable with brokering external services as it is with generating its own - doing whatever it takes to support the needs of business users.

But for 2012, let's push IT - and the business - to begin planning for this transformation.

Customers Catch Up to Vendor Vision
As a long-time marketer in high-tech, I've seen the tendency of vendors to push customers to adopt the Next Big Thing. And that N.B.T. is frequently disruptive (or at least discontinuous) with respect to the "legacy" approach to doing things.

To be sure, there are always customers who are leading-edge in their technology adoption. But the mainstream customer adopts technology *incrementally* and rarely if ever discontinuously. This is a byproduct of (a) the human tendency to mitigate risk, and (b) business' tendency to plan change - and budgets - incrementally.

So, for 2012 - and for the foreseeable future -  this trend won't change. Hopefully vendors will be more clear about what's "vision" or what's "for early-adopters", and maintain a healthy dose of sobriety about selling high-brow discontinuity to the mainstream market.

As this relates to Cloud computing, it's clear that more of the mainstream market is adopting virtualization, and has bought into the concept of cloud computing initiatives. But let's be clear: In 2012 the average IT infrastructure won't be completely re-built into a private cloud.  However, I believe that the majority of medium and large IT shops will all have begun their progression toward the private cloud eventuality.

Other (prediction-related) links

Sunday, December 4, 2011

Celebrating Innovation at EMC

Can innovation be fostered or facilitated? Can it be measured? Is it a formal or informal process? Can you have an innovation competition? Is it limited to Engineering? 

These were not the topics I was expecting at last week's EMC Innovation Conference. Rather, I thought it was going to be a geek fest with lots of greasy-haired, propeller-hat-wearing attendees drinking diet-Coke and speaking mostly PERL, C++ or Greek.

Instead, the event was attended by a few hundred local Silicon-valley EMC-ers, with hundreds more globally watching simulcast. It was business-casual with execs and technologists alike, and kicked off with none other than EMC's COO Pat Gelsinger (and former Intel CTO) and MC'd by SVP/CTO Jeff Nick. Rather than a siloed techno event, the importance of innovation at EMC was supported at the very highest levels of the company. Now that's refreshing.

Guest stars attended too - one Scott McNealy - in his usual humorous, politically- and sports-charged manner.

But First: Did the COO "Get It"?
Resounding yes on this one. Pat opened with a real command of what innovation means to the company - Innovation can mean disruption... and even destruction of old ideas. And that's good (if you don't agree, read the Innovator's Dilemma). But Pat also pointed out that without being coupled with commercialization/monetization, innovation doesn't count for much. And knowing how to commercialize an innovation can be just as challenging as coming up with the idea in the first place. More on that later.

Pat was also insightful in that innovation isn't just about technology. You can have organizational, marketing, and  process innovation, in addition to product-based innovation. And innovation can be top-down and/or bottom-up (EMC does both).  To help accomplish systematic, company-wide innovation, EMC uses the Golden Triangle: Organic innovation, University research/relations, and EMC Ventures. More on these later, too.

Can Innovation Be Measured?
Both Jeff Nick and Steve Todd (Director of EMC's Innovation Network) focused on this question, and there are many perspectives. Patents, for one, are a useful metric for organic innovation, but as Pat pointed out, they don't necessarily indicate that an idea has been monetized. Nonetheless, there is a rough norm of patents per employee per year that the company is striving towards. Intellectual Property in-and-of itself can be valued, as we've seen from recent acquisitions of IP by IBM, Google, and others.

Jeff also pointed out that another indicator of innovation is of course the monetization itself. What's the ROI on an organic development group? Did an acquisition recoup its cost?  Did a a venture investment create a useful return, or even better, might it have been a profitable acquisition? 

While measuring innovation might still be tricky, knowing how to facilitate indicators gets part-way there.  So when Steve pointed out that another indicator tended to be when small groups work together to solve a problem, there was an "aha" moment.  How to constantly encourage groups to form, and information to flow? Therein lies his "day job"...

Can you Compete for Innovation?
Thus, Steve realized that encouraging groups to form with a common purpose might spur organic innovation. And, for the past few years, he's been conducting an annual innovation contest within and across the company. Participation has been growing at a healthy clip. Topics can be many - some were geeky, but some were clearly for social betterment. 

To his credit, Steve also used a dose of data analytics to map every individual and every project across the company. He confirmed, in fact, that finalists tended to work in groups or "Cliques".Which begged the question, how to facilitate and accelerate this behavior in the future.

Formal or Informal?
The other aspect I really appreciated about the day was the intentional mix of formal and informal innovation the company strives to maintain. As I mentioned, the company emphasizes the concept of the "Golden Triangle". In addition to internal/organic innovation, we also work informally with external universities and research... and of course, formally with many prominent VCs and startups.

Mark Lewis, EMC's Chief Strategy Officer and leader of EMC's Venture group, gave a fantastic overview of how we regularly look at the startup community, and place measured investments with them. The ideas is to discover great ideas both ahead of the competition, and early enough such that we don't have to pay huge premiums to acquire should we deem them valuable.

The strategy behind these investments is broad, but what I heard was a 2-pronged approach:
(a) Pursuing the technologies that are complementary to EMC's product set - technologies and ideas that are net-new to the company, yet extend our leadership in the data center market
(b) Pursuing products that might eventually replace or cannibalize existing EMC products.  Like the Innovator's Dilemma points out, better to cannibalize your own products than have someone else do it for you.

Limited to Engineering?
Capping off was another surprising panel... of marketers. Could Innovation extend into how we market EMC and our products? Absolutely.

Two fascinating points stood out, and hopefully made a huge impression on the audience and EMC as a whole.  First, the panel chosen consisted entirely of recently-acquired marketing leadership located locally in the Bay Area. Scott McNealy had pointed out that "Silicon Valley DNA" was somehow different of that from the rest of the country - and EMC (based in Hopkinton MA) was making a concerted effort to graft that DNA into the rest of the company.

Second, was the attitude of corporate marketing. Do things differently. Use metrics. Understand the trends of the "new" buyers... (Hint: they don't pay attention to marketers using traditional email, snail-mail or traditional TV). Innovation here had to do with how social marketing, new/viral media, and fast-paced imagery was the way to both get attention as well as to change the face of the company brand.

For me, this day definitely served to change how I thought of the CTO's office, how I related to the concept of innovation, and how I think of corporate development.  As EMC further pursues these paths, hopefully other leading companies will as well -- it's good for them, for careers and personal development, and frankly for the economy as a whole.

More Resources
Jeff Nick (CTO): About  Article
Mark Lewis: About  Blog 
Steve Todd:  Twitter: @SteveTodd     Blog: Information Playground
Chuck's Blog: The Challenge Of "Innovating Innovation" In Large Corporations