fbpx

Shed New Light on your ‘Dark Data’ with a Data Transformation

By | 2020-01-10T11:18:27+00:00 January 10th, 2020|

If your business could realize substantial savings, gain additional revenues, and boost operational efficiency, wouldn’t you want to?

Then you’ll need to make sure your business gets its data transformations and quality right.

A successful data transformation, and the related big data implementation to business processes, can unlock massive benefits, according to global consulting firm McKinsey & Company, which notes one US bank expected to save more than $400 million in savings after rationalizing its IT data assets – and a $2 billion gain from new revenues, lower capital requirements, and operational efficiency.

But for all these success stories and all the focus on data-driven initiatives, many organizations are still struggling to realize actual value, according to McKinsey.

What is data transformation?

We spent a good portion of the last decade talking about the exponential quantities of data being generated as more users took pictures on smartphones, we created exponential volumes of data – and then grew that daily, and we frequently enter search queries and input tracible digital records.

In other words, people – and the businesses they work for and buy from – generate a lot of data. Every day.

But the majority if this data isn’t actually using, spurring Gartner to ominously call this information that is collected, but unused, ‘Dark Data.’ This is data that isn’t generating value for businesses – until it is transformed, that is. Imagine big data in the legal industry – much of it has to be stored for compliance but is often otherwise unused.

Data transformation is the process of taking raw data, cleaning it, validating, and spitting it out in a ready-to-use format that enables your organization’s big data analytics implementation.  

In practice, this requires significant upfront planning, and then the data transformation team must transform the data into the target format (called ‘conforming’).

What about data quality?

In order for an organization to realize real value from a data transformation, the data must be:

  • Easy to access
  • Consistent
  • Secure
  • Reliable
  • Oriented to its users

One key step in this process is cleaning the data. Take dates for example, which are sometimes imputed MM-DD-YYYY or others DD-MM-YYYY. For data quality, it’s imperative the data fields are uniformed, or records will quickly become muddled.

Some records may also have missing data, and you’ll need to determine how to handle them.

You’ll also want to record any audit or data quality metrics used during the data transformation process, such as:

  • Where there any columns with the null field, junk data, or values outside the bound?
  • Are facts supported by records? For example, does a sale column entry have a table recording the actual sale?
  • Is today’s data load larger than yesterday’s? (if it should be.)

Including facts and audits help support data validity if a stakeholder asks. (Remember: this data transformation is going to enable leaders to quickly tell stories with data, and if the news is very good or very bad, their first question might be “Where did this data come from?”)

How to design a data transformation with value in mind

Your business has a wealth of data that can help inform business decisions – in real-time – if that data can be harnessed into a usable output.

This of it this way: if you were going to ask an investor for funding, you wouldn’t hand him the login to your Salesforce and expect him to mind through your customer purchase data and pipeline. Instead, you would harness that data into a digestible presentation, and present the highlights.

Your enterprise data needs to do the same thing for business leaders using company data, which is why it’s critical your business designs a data transformation for the end game: valuable, useful insights.

Follow these practices to kick off your data transformation with a plan in place, and get the most out of that wealth of data.

  1. Set (Your) Expectations. Understand exactly what the business plans to get out of a data transformation before you embark on overhauling your data warehouse. Know what exactly you hope to achieve, and what impact that will reasonably have on the business.
  1. Break it up. Next, take that overarching expectation and break up the data strategy into its various uses. For example, this might be fixing gaps in pricing data or establishing a new reporting service and assigning a priority to each. 
  1. Overhaul your architecture. Now is the time to revamp your data architecture, avoiding what McKinsey calls a “data swamp.” The visual the term conjures up says it all: you don’t want your valuable information sinking slowly into a murky pit. In fact, you want just the opposite – crisp, clear data with efficient access and seamless API integration. (Now might also be the perfect time to entertain a move to cloud computing). 
  1. Ask for input – often. Weave the people who will be using the data into the entire transformation process. Gather their feedback, repeatedly, and ensure they can – and will – use the final product. 
  1. Place someone in charge of the data. Officially called ‘data governance,’ this is the process of ensuring your data remains quality. While its likely one person is responsible for overall data standards, processes, and management procedures, business units can be tasked with the data quality coming out of their business unit. Most data quality errors can’t be blamed on technology – it’s the people behind the tech causing the issues. So, it’s fair to ask the customer service unit with ensuring customer data is inputted correctly, and hold the sales department accountable for accurate pricing data. 
  1. Champion new ways of working smarter. Once the data transformation is complete, ensure your business reaps the benefits by anchoring the business in the applications to use the data for. If previous pricing inconsistencies cause customer issues and revenue loss, engrain the now accurate (and real-time) pricing data into the process of sending customer quotes. 

Are you ready to make the most out of your big data? Learn how our Genies can consult with you every step of the data transformation process, and ensure your final product is a quality one that can help yield immediate value for your business. Contact us today at info@techgenies.com to get started.