Big data has swept the world as a buzzword but, at this point, it is mostly just a cliché.
While the hype has led to increased investment in the sector and a renewed interest in data matters, big data itself is still high in the hype cycle. The view from the front line and the back end is that it will take some more time before the technology matures.
While marketing budgets churn out big data solutions as the panacea to all enterprises’ data management needs, within the enterprise there is a need to make changes to how we conceive, manage and operate data in order to unlock the big potential.
What we find at Equinix is that business outcomes need to be wedded to any data initiative. By defining the business objective first, the potential of big data can be strategically unlocked to suit that purpose and derive maximum benefit. When all the hype settles down, what is revealed is that not all data is important and to paraphrase George Orwell, some data is more equal than others.
According to Cisco’s Visual Networking Index, global IP Traffic will reach 120,643 petabyte per month by 2017 at a compound annual growth rate (CAGR) of 23% while worldwide mobile data will reach 11,157 petabyte per month by 2017 at a CAGR of 66%, meaning that more data is being created than what we can truly make sense of using traditional methods. Therefore, finding what part of the data piece is relevant to your business is of utmost importance. Data may be growing at exponential rates but at the end of the day, the business owner must be able to decode which fragments of information are relevant to the longevity of the enterprise.
For businesses, the question that underlines every big data conversation has to be: ‘what is it about this data that is important and will it give my business a competitive edge?’
Increasingly, it is not only the depth and qualitative understanding of big data that is of value to enterprises but it’s also about getting the information in a timely manner and using it to compliment existing ways of making data-driven decisions, not as a substitute.
The key issues to look for in the big data ecosystem include:
Latency: Many enterprises need fast access to incoming data in order to make quick, actionable decisions. If there is a time delay in accessing data, this could severely impact the deal or decision.
Proximity: If the speed of information flow is critical to your business, it’s important to be in close proximity to a data center where you have close access to multiple data feeds, ensuring instant information.
Multiple Data Sources: Today’s businesses are looking at making decisions based on multiple data sources – and some of these sources may not be internal. Therefore, access to external data sources becomes key and locating your systems close to these various data sources is quite important.
Increasingly, new data intensive sectors in the digital economy require low latency, close proximity and immediate access to other businesses to be viable. For example, in the emerging sector of real-time bidding (RTB) for online advertising, these three “big data” factors are critical for the success of buying and selling online display advertising. Key to the success of the online and mobile advertising ecosystem is the need for targeted advertising to the right audience to drive the best returns – all within a few milliseconds! Deciding where to locate your data and servers, then, is a crucial factor.
In the more traditional telecommunications space, big data analytics creates significant potential for a telco to leverage their massive customer data to customize services, thereby creating a common ground to promote greater collaboration between telcos and OTTs.
Two critical aspects to close this out: 1. Identify the data that matters to your business. 2. Ensure infrastructure is deployed to store, access and process the relevant data to deliver the desired business outcomes. Big data may live in the cloud but it cannot be placed in a silo. Businesses need a scalable infrastructure to extract the best out of their data-whatever size it may be.