Does a Small Business Really Need Big Data?
In a recent business article, Forbes magazine urges small businesses to incorporate big data into their marketing plans. The implication is that businesses that ignore big data will be flattened by the competition before long.
Big data – the vast quantities of data about consumer behavior that one recovers from the social networks, the cellular networks and other records of consumer behavior – is touted by marketing consultants and data analysis conglomerates like IBM as the biggest marketing innovation since the Internet. Why are small businesses supposed to use it? Do the data analysis firms just want to find a way to profit from all the data analysis ability that they have?
Very few businesses appear to have any actual need for big data
Google is in the business of analyzing data from billions of webpages to rank and sort them. Certainly, Google needs big data and has the processing power to make use of it. Not every huge business that deals with millions of pages and accounts, though, can use big data. At vast businesses like Yahoo! and Facebook, much of the computing and data analysis done is relatively smalltime. While they might need their clusters (small powerful groups of computers) for the occasional high-level task, most of their big data analysis is handled by a single server class computer.
If big data at Facebook and Yahoo! is small enough to be done on a single server, the logical conclusion could be that they don’t use big data much. If this is all the data analysis that vast businesses need, could small businesses have any need for it at all? A research paper on the subject from Microsoft called Nobody Ever Got Fired For Buying a Cluster recommends caution in taking up big data, too.
Big data costs big bucks – something that small businesses don’t have
In many of the problems that small businesses face day-to-day, looking to big data for information could actually produce dangerously incorrect results.
The reason for this is that on the small business level, data analysis isn’t reliable enough to extract information out of very large data sets. Often, their analyses turn up false positives or false negatives. Businesses today simply don’t have the resources needed to tell which results are true and believable and which ones are not.
The expertise needed
Finding employees who can accurately and dependably interpret statistics can be one of the toughest hiring projects any HR manager undertakes. Big data is often full of gaps, biases and other flaws. Even data analysis experts with PhDs would have a tough time adjusting for all these flaws and coming by useful conclusions. Small businesses usually don’t have the resources needed to hire such expertise.
Should you choose small data or big data?
Small businesses often need to be agile – they need to quickly recognize marketing trends jump on them. Taking the time to analyze big data can hobble small business.
The major discoveries and inventions by great scientists of the past were made with nothing more than a few rudimentary analysis tools. Not only did they not have big data, they didn’t have small data, either. Often, vast amounts of data aren’t what you need for insightful decisions. Instead, you need a human in charge with enough marketing experience to recognize trends and make good calls.