Big Data: a big train to catch (Part 1)

By Laurent Duenas, April, 7th 2013 


Reminder of the business stakes

Big Data is a new approach of business intelligence exploiting multiple data sources, creating value to companies by having a higher awareness of customer’s needs or the weaknesses in business process.  As compared to previous decisional applications is the requirement to manage large scale and diverse amounts of data originating from a variety of sources that “Data Scientists” have to manage through complex data models.

With this new technology, companies expect to sell more products by interacting with customers, utilizing enhanced information and contextualized proposals related directly to client identity.  Big Data is where .Net companies are currently competing to be predominant. Loyalty programs, online-selling propositions, are based on rich and unlimited knowledge tanks fed from mobile and geo-localization systems, online applications, or social networks.  The sources are unlimited: even images and sounds can be converted into interpretable data. Everything is useful to model customer habits and define successful go-to-market approaches.

Big players in the internet such as Amazon, Google, Apple, and many others, made a lot of investments in Big Data and they have enjoyed significant return on their investment. The online shops surpass traditional stores in better understanding of how purchases are made. Big data is also used to identify root causes of deficient operations such as supply chain, maintenance, or quality issues. For instance, Airlines companies have begun to combine weather information with selling forecast, in order to optimize their ground operations which were, until now, the origin of unpredictable costs and service outage.

Big Data domains have no other limit than the intelligence of the data model. This is a new IT “el dorado”. Major IT vendors have understood and embraced this trend providing cheaper storage and servers. New data analysis software is frequently released into the market which appeals to companies both large and small.  Nothing can reverse this movement.


How ITSM can support Big Data?

 

What are the challenges?

The efficient and profitable management of large amounts of data remains a key challenge for companies. On one hand, we have a variety of IT services which can feed in Big Data containers. On the other hand, we have on-demand, modelization works which require prompt, but significant power for functionality.  All of this stuff for potentially short- term results that probably will be repeated several times before finalization.  The very challenge is to manage the resource provisioning at low cost with a flexible scalability. To do so, IT Service Providers must reach a mature industrialization proceeding with orchestrator and request fulfillment portal for Cloud provisioning, key for efficiency.

It points out that agility will be one of the most strategic capabilities of IT Departments. Companies will redesign some of their core business processes to integrate Big Data outcomes into their front-offices systems (online shopping services, order entry or CRM systems).  Big Data recommendations have to be implemented on an on-going basis following the customers pace, which will certainly be frequent. Responsive “Change, Release and Deployment Management” processes, along with suited change models and release policies, are mandatory if IT departments want to actually contribute to the competitive advantage of their company.

If we extend our purpose to others warranties, a global approach of the Big Data value chain must be taken into account. Only focusing on Big Data components is not realistic. What provides a competitive advantage to a company  is not only the intelligence of the Data models nor the power of Big Data infrastructures, it also relies on the IT departments capacity to embrace the benefits of “Big Data” that make front-office IT Services superior. If they cannot be accessed within the appropriate response or updated time frames, “Big Data” component performance is useless.

 

 

Copyright © 2013 - PRACT Publishing


Other articles

►  No link


External links

►  No link


Related Products

►  No product