The Next Big Thing
Posts about next generation technologies and their effect on business.

Are cloud failures different and more common?

Cloud failure.pngWith every technology there is a lifecycle and it cloud computing must be coming off the top of the top of the hype cycle with the number of stories similar to: The worst cloud outages of 2014.

 

And follow on stories like: Cloud failures will happen. Are you ready? It is prudent for articles to make statements like:

“Even if you only use the most reputable cloud services and products, things are bound to go awry from time to time so it’s crucial to be prepared for failures.”

 

And the ever popular: Why some cloud projects fail? Granted there are some staggering failure rates for cloud projects, but there are significant failure rates for all technology related projects.

 

Many of these patterns of failure are not unique to cloud. It usually gets down to a few issues core to every IT project:

  1. Know the business expectations/requirements and how to measure them.
  2. Have clear executive support
  3. Start small and make adjustments based on facts – iterate. If you don’t get what you expect make changes.
  4. Keep the big picture in mind (in many dimensions). After all, you’re trying to address the needs of the enterprise and not usually just a silo, at the end of the day.

Moving to cloud implies moving from operational thinking to a services mindset. Too many companies still bother about the underlying technology, forgetting to realize they now buy and integrate a service.

Data and decision latency expectations shifting

Analytics time.pngOne of the issues I’ve talked about many times over the year is the need to shift our understanding and expectation of latency and action. I came across this post: Analytics Time Lags Result in Lost Opportunities. It also discusses the fact that the data gathering and analytics that were great a short while ago are now viewed as insufficient, stating that “72% of analytics and business leaders surveyed were dissatisfied with the time it takes to get data-driven results”.

 

Unfortunately, it didn’t really do more than imply that solutions exist. Fact is most solutions deployed today are based on hindsight. There is little doubt that the vast amounts of data available are going to require an exceptional command of information, far beyond just hindsight. It will require a refocusing of skills and perspective that are based on generating value from the abundance of computing and data available. This will require new techniques for computing as well as data gathering and integration. The work going on in HP labs related to The Machine will help address these needs, when this platform is released.

 

We have data coming in from sensors and mobile devices creating an ever increasing amount of Dark Data where value can be generated. We can also build context from the other data about what happened when, who or what was involved or happened at the same time. This derived data or metadata can sometimes be more valuable than the raw data itself, since people don’t really make decisions off the data but the context the data describes. Organizations are recognizing that all this data will provide a depth of understanding about what happened in the past, present and future that we’ve not really taken advantage of before.

 

We can develop a greater depth of understanding about what is happening right now that can enable us to automate decisions or concentrate that rare resource – employee attention - on those areas that really need it. There are relatively new technologies that most teams have not even looked at like software defined networks… that can operate on data on the fly instead of just data at rest. This will eventually enable a more active, organizational approach to tackling opportunities.

 

Finally, over the years we’ve learned that getting to zero response time is very difficult. It may actually be easier to move to a negative response time, where you predict what is likely to happen and adjust to be ready to address it or even shift the outcome. Tools to address all these various perspectives of data and enabling right-time decisions are available to improve your ability to optimize time-to-action are available today.

Marketing in 2020

marketing.pngThere have been a number of industry specific version of HP’s 20/20 effort but I just saw the most recent one focused on marketing. The subtitle for the release is Welcome to a new reality of split-second decisions and marketing by the numbers.

 

When they were pulling together this release, they took a number of subject matter experts and allowed them to discuss the key issues they see in the marketing space. It has a number of articles and perspectives such as:

  • An overview of marketing macro trends
  • Real-time marketing
  • Buyers in control
  • Insights from information
  • Too much information
  • Challenges of marketing in 2020
  • The CMO of 2020

In any case one thing that is clear – as marketing becomes more information and context rich, it will become measured by actual and modeled performance more than ever before.

Automation and business restraints…

any ideas illustration.pngThe concept of ‘golden handcuffs’ has been around for a long time. This is a job that has such good benefits or salary that it can be very difficult to leave, no matter how much it frustrates you.

 

As we move into a work environment that has ever greater use of automation in knowledge management roles, there is the likelihood of a ‘steel handcuff’ scenario, where a business benefits greatly from automation but eventually loses the ability to effectively maintain their rules and models. They maximize the financial benefits and reduce the workforce to the point where they no longer have the critical mass of industry knowledge to adjust to future business demands. They essentially get locked into the current model.

 

This issue can be overcome by understanding the skills and value of the industry expertise they have as well as the dynamics and value of the automation components. If they feel they can’t afford to have that expertise available full time, they need to devise a consulting supported approach to keep the expertise available, since it will be needed sometime. Tweaks and experiments should be part of the automation strategic model, since it should never be considered done.

 

Some organizations have experienced this scenario with their existing COBOL environments, for example. Their systems work today, but most of the people who understood it have either retired or been let go. To make a change may require a total redesign/rewrite, to develop a new crop of people who understand the business needs.

 

My view is that the foundations and rules should be viewed as a starting point for continuous adjustment and understanding and not a final product that can be declared complete.  On-going interaction with the business will be needed.

Why is the IoT viewed with such potential and confusion?

Internet of things.pngThere is a fundamental shift underway from dumb devices where organizations guess about how their being used, when and by whom. Now a physical product (from almost any industry) has the potential to be a first-class participant in its own value chain. It can talk back to its creators in engineering and manufacturing as well as those who service it, cutting downtime and improving its use. It has the potential to talk with (where everyone seems to be focused) those that actually use it, making their life better and more productive. There is also the potential to collaborate with sales and marketing to share what users are thinking based on where, when and how it’s used. Devices/products are becoming members of an environmental view of the context that surrounds them. Although it involves information technology, it is about a shift in business value.

 

This challenges the foundations of many of our existing products and services. Devices can have an active role in CRM and marketing. We can shift the analytics view from the past to the future. We can use the information to gamify processes and shift behaviors. As this understanding increases, what is measured and the decisions made will shift.

 

As I mentioned last month, the impact on our definition of services will shift as we understand and embrace the potential. This change will shift much of what exists (people, products and services) in our environment/industry.

Search
Showing results for 
Search instead for 
Do you mean 
Follow Us
Featured
About the Author(s)
  • Steve Simske is an HP Fellow and Director in the Printing and Content Delivery Lab in Hewlett-Packard Labs, and is the Director and Chief Technologist for the HP Labs Security Printing and Imaging program.
Labels
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.