Big Data is the next big solution looking for a problem. Faced with falling revenues from lower ICT spends and clients with reduced consulting budgets Big ICT looks to Big Data for Big Revenue. Customers will be left as Big losers if they fall for this parlour trick.
The central premise of Big Data is this: Companies, organisations and government can spend a lot of money cleaning and storing data through traditional means. This takes a long time and costs a lot of money. On the other hand, the same bodies can buy a supercomputer and software to crawl over their data and run sophisticated algorithms and come up with the same answer. The answer will be about 80% correct and based on statistical analysis rather than absolute truths but it’s still a good answer. Therefore, don’t worry about information management. Just buy really expensive hardware and get a bunch of clever people to use it and voilà! instant brilliance.
There are 2 problems with the aforementioned premise, namely: (a) Most businesses cannot do small data well, let alone big data. Secondly, (b) there is no proven link between ICT spend and business productivity.
IT companies would have you believe that businesses are drowning in data. Well, to continue their analogy, a person can drown in 6 inches of water. To go even further, an infant can drown in 2. The point is clear. If management do not know what they wish to do with the data then no amount of computing power will save them. In the absence of clear action companies should structure their important information so that it can be stored and found easily. ERP systems do a good job of this and there are plenty of EDRM systems which are convergent. More sophisticated users need more sophisticated solutions and this is where Big Data can assist. More importantly, with the capital costs of Big Data solutions it will not be feasible (at least in the short term) to use it on standard management information. Big Data solutions will only be cost effective for advanced business intelligence-type analyses. If a business is looking for mainstream solutions to support business intelligence it will need to look for the more mundane.
Secondly, the link between ICT spend and business productivity is not made out for any other sector other than Financial Services & Insurance (FS&I). This sector is now so heavily automated in its business processes that there is a clear dependency between IT and profitability.
For the rest of industry and government nothing much has changed since Strassmann’s work in the mid-90’s (see below). This is to say that there is no point buying ICT to automate a bad process. Expensive IT will not help poor management. This is why banks, Tier-1 consultancies and advanced process oriented firms (such as GE) will likely be the only beneficiaries from Big Data.
Big Data is somewhat of a misnomer because it is not like standard ICT spend. The Big Data will almost inevitably be run in conjunction with technical services because part of the benefit is determining veracity through market sentiment. For instance, if an investment bank is trying to determine the initial share price for an IPO they would like to run trend analysis on similar companies but also see how this fares against potential investor sentiment. Big Data used in this way is certainly a unique offering if not a competitive advantage for smaller companies.
Big Data is not a game-changer. As computing power and processing speed increase (according to Moore’s Law), along with ubiquity, then the benefits and services of Big Data will become mainstream. In the meantime, however, Big Data will be a competitive advantage for many of the smaller and 2nd Tier businesses.