Apples and Oranges – How Different Data Types Can Make the Perfect Match

Apples and Oranges – How Different Data Types Can Make the Perfect Match

By Aim's data protection experts

March 2021

 

We have written before about the issues one might encounter when trying to merge datasets, for example, company A acquires company B and wants to merge their CRM systems, but both have different naming conventions, different field lengths, one does business only in the US, and the other worldwide. These are not insurmountable problems, they just require a careful analysis to establish what goes where, what types of data there are, and how the final product is to be used, so they can be joined, or not, if that’s what the analysis says is best (remember, don’t be held hostage by your preconceptions).

 

The data we get as a result of the merge is potentially more valuable than the two separate datasets, because even though there are the same number of contacts, or maybe fewer if we discard duplicates, the combined data may show trends or links we could not have possibly known by only looking at one of the sets of information.

 

I mention this because the oft quoted “data is the new oil” demonstrates that data has a value, and maximising that value is key to new business, growth and a healthy bottom line. The question is how do we maximise the value of our data, especially when that data does not seem to be connected? Simple paths include having a Data Scientist, whose job it is to analyse your data and find connections, you could have knowledgeable employees who’ve seen it all before and can come up with solutions, or alternatively, you might get lucky. But if you don’t have these options, what else is there?

 

As with most problems, we need to go back to basics, so starting from scratch work out what data you have, categorise it, look at the ways it can be grouped, see if there are clear connections, and then look wider. What I am most interested in here is the ability to collate large volumes of disparate data, in many different systems, and use a consistent categorisation method to structure that data and provide clues as to how to combine it. Even better if we can then assign monetary values to that data, and taking it a step further, what if we can utilise machine learning to identify and classify objects, and then use this knowledge to find new sales channels via the analysis of publicly available information?

 

This is all possible, for instance Aim’s dataBelt® can scan and index data from many sources, categorise it, and assign value to that data. In addition, its machine learning capability makes it valuable in virtually all verticals, maximising value, whether that be monetary or not (remember, law enforcement want maximum value from the data they have, even though that value is not measured in money). So, in a data rich world, don’t sit on separate islands of data. Instead, use technology to integrate them, analyse them, and make them work for you.