By Dharman Shetty, CEO, Xenolytix
You cannot have a conversation these days with technology folks without words like Big Data, Cloud, Analytics, and IOT popping up within the first couple of minutes, and it is not surprising. There are considerable opportunities for businesses to leverage their data repository to create significant value, and business leaders are recognizing the gold mine they are sitting on.In fact, it is no more an option for many business to not do anything on the big data analytics front, as others within their industry are trying to gain a competitive advantage.
Advancement in technology and cloud-based infrastructure services have made it easier and cost-effective to extract valuable information that can be used to influence business outcomes. This, however does not mean that all Big Data analytics projects will yield positive outcomes; going about it the right way is key to optimizing value and return on investment.
PUBLISH YOUR ARTICLE
Define Business Outcomes: Yes, everyone knows this, but not retaining focus on the business goals has remained as one of the primary cause of disappointment inmany technology projects. It has not been uncommon for CXOs to get a bunch of operational reports each day, yet find it non-insightful to make an informed decision. Big Data projects are all about analyzing humungous amount of data. Ill-defined requirement, unclear expectation settings, and inconsistent understanding of business need have often led to less than optimal outcomes. To avoid churning the ocean, invest the time and effort upfront in defining the need as accurately and detailed as possible; it will save considerable time and effort in the long run.
Real-time Feedback or Strategic Decision Support: These two are distinctly different and cost variation can be significant. While Big Data analytics of any kind has to process large amount of data, real-time feedback systems is all about processing live streaming data at high speed and delivering instantaneous feedback. Time is of essence, and ROI is dependent on ability to correct or optimize performance in real-time. Strategic Decision Support Systems, on the other hand, typically do not need real-time processing but involve larger volumes of data, higher level of computing, with considerable drill down ability to gain critical insights. Clarity about the purpose can help optimize cost as the differential between the two can be significant.
Be Realistic: It is only human that we preferto know answers to all of our questions instantaneously. We have been spoilt by instant responses we get to our search queries on the internet, and by smart devicesthat constantly keep us posted on what are supposedly long lost best friend is doing on the other side of the globe. We bring the same expectation to our business requirement, and there is nothing wrong with that. However, business leaders have fiduciary responsibility to the shareholders. To optimize ROI, one needs to determine the opportunity cost of latency. Is there a meaningful loss in value if the response to a query is not in sub-second but within two or three seconds? It may be a nice-to-have criteria, but it has significant ramification on design, infrastructure, and thereby cost.
Be Selective: Do not churn the ocean; be selective about the data. It is important to know what data sources are included in the analysis, but it is vital to know what has been excluded.It will help make the right application.Leverage the Cloud: There are several third party cloud infrastructure service providers; but not all are the same. Take a strategic view of your infrastructure needs to determine the right service provider. The leading service providers have made significant investment to build best-in-class infrastructure and have introduced usage based pricing; why not leverage the same to reduce cost, increase flexibility and maximize return on investment.
Be willing to act: It is prudent to be cautious, but inaction at the right time would make the whole endeavor a waste of resources.