The Bad and the Good
As our ability to collect more and more data grows, new learnings are also emerging… however are we taking these onboard quickly enough?
We know that businesses can now collect ever-growing quantities of data. When it’s done well, big data collection and analysis can help us identify business opportunities. Becoming more efficient; reducing costs and helping with better decision-making for instance. But are we reaping these benefits or do we still have a long way to go?
Bad or Rogue Data
Let’s examine “bad” or “rogue” data. Basically, this term means that a business isn’t making proper sense, or best use, of the big data they’re collecting, says iConnect CEO, David Godfrey. Bad or rogue data is a general way of referring to data that’s somehow not useful, accurate or true. It could suggest that the method of gathering the data was not done in a scientific way, to begin with. Leading to over-simplified findings. When this happens, it’s usually because the parameters that were chosen to collect the data weren’t correct. Put simply, a parameter is an important component in any statistical analysis. See below for more on this*
At other times, rogue/bad data is also used to describe the findings extracted from that data. Even if the data was gathered well, but are not being analysed scientifically or in a way to assist the business goals. These goals being the reason the data was collected to help achieve, in the first place. Many businesses collecting big data, abroad and locally are making these mistakes every day. Which explains why data scientists are in demand and universities are offering data analytics courses.
Experts now say that bad/rogue data can cost a business between 15 and 25 percent in lost revenue. So, if there’s bad data happening in yours, make sure you use best-practice to collect and analyse data.
Your Data Team must:
- Become disciplined about collecting data logically and measuring appropriate parameters? – If not, the findings won’t really be useful or accurate and can be easily misunderstood. A common mistake is simply merging big data streams. Instead of taking the time to first identify the patterns emerging and then work out which are significant, and which aren’t.
- Analyse the data correctly with your specific business outcomes in mind?
- It’s important to balance your findings with how this will impact the end user’s experience of your product or business?
If your connectivity solution isn’t enabling effective big data collection, contact iConnect.
*See https://spotlessdata.com/blog/saying-good-bye-rogue-data. For more about the definition of bad, dirty or rogue data.
* See https://explorable.com/parameters-and-statistics to find out more about how the word parameter describes a numerical value (number or percentage). And, at the most basic level, scientists usually begin measuring data with two important parameters – i.e. they find the average parameter; and then work out how many (and how far) most people (or whatever’s being measured) deviate or differ from this average. And then use this information as insight on how best to improve or influence a situation or achieve a goal.