Accurate and relevant data can be incredibly valuable in reducing risk and increasing revenue for the enterprise....
And a lot of companies are banking on big data, utilizing a variety of tools, from data interpretation with the likes of Hootsuite, DataSift, FullContact, Keyhole and Gnip, to data quality from Informatica, Talend, SAS, SAP and IBM.
Human data is the fastest growing and most collected type of data. It encompasses a wide variety of information, including geographical movement, pattern analysis and social media posts. This information is publicly accessible and feeds decisions for small and large companies alike; however, the data just being there doesn't mean much. Figuring out what makes all of that data "good quality" is where the process becomes a bit more complex.
- Defining the specific requirements for "good data" regardless of where it's used.
- Establishing rules for certifying the quality of that data.
- Integrating those rules into an existing workflow to both test and allow for exception handling.
- Continuing to monitor and measure data quality during its lifecycle (usually done by data stewards).
Quality data depends on the context in which it is consumed; and like anything in business now, it is constantly changing. Good data gets updated, modified, and added upon alongside your business. Stagnant data becomes bad data. The longer that human data stays in a silo, the less likely it is to be accurate.
Making a large amount of data more current was the driving force behind Microsoft's decision to purchase LinkedIn for $26.2 billion. Microsoft wanted to move further into the ever-changing job market sector and looked to benefit from the very active human data LinkedIn receives every day. Their vision is to become the primary platform for managing, curating and promoting their users' entire professional lives.
Satya Nadella, CEO of Microsoft, put it this way in his recent interview: "Having the entirety of what is your professional life be enhanced, more empowered ... being more successful in your current job and finding a greater, bigger, next job. That's the vision!"
Microsoft becoming the one-stop-shop for your professional life would be a very natural and profitable progression for them. Windows is already the unrivaled operating system of choice in business, as it accounts for 93% of the PCs used. This acquisition will help them achieve more market share in a space where they are already pretty well established.
So why is thinking about data as "living" a relatively new concept? Rolodexes and encyclopedias have been replaced by mobile phone apps and Wikipedia. We already know this -- and yet how many of us treat our data like it belongs in a vault instead of like a pet that's part of the family?
Trillium Software, another innovator in data quality enhancement tools, explains the problem: "The volume and variety of customer data available to modern marketers is growing exponentially, but incomplete, inaccurate, duplicated or inconsistently structured data is difficult to consolidate and use intelligently for personalized, one-to-one marketing. If data is not assembled into a clear, coherent and singular customer view, your marketers cannot identify preferences, deliver personalized experiences or accurately analyze results. The reality of modern day marketing is that available data is not necessarily actionable data."
That hits the nail on the head. Instead of using generic terms like "quality data," replace it with something more descriptive. What does "actionable data" look like to you? What data would support your sales team in building one-to-one marketing campaigns? What data would give you insight and a competitive edge during your next big meeting?
Start identifying what actionable data means to you by following these three steps:
- Identify a goal for the data -- I'm a fan of flexible reverse engineering. Using this approach, you can visualize a goal, create your priority structure to support it and then align people along with it. So start by identifying what you want your data to do. Be clear about your priorities and push yourself to go further than the generic "get more money."
- Review the state (quality) of your data -- This might involve a couple of different business units. What information do you have and how can it be applied? Are there privacy issues with sharing the data? How can this data be improved/joined with/streamlined? Ask these types of questions and identify the limitations or possibilities of what you already have.
- Weigh the benefits of improving the quality -- This is where the vendors we've mentioned would come into play. Many offer a myriad of services, all aimed at bringing your data into purposeful alignment with your vision. They services can increase the "quality" of your data by making it more "actionable."
If you decide to take the third step, consider choosing a reputable company with a process that makes sense to you. Gartner released a Magic Quadrant report for data quality services in November 2015 that gives excellent insight and in-depth analysis of the industry. I would highly suggest reviewing the report if you're learning about data quality in general or are actively choosing a company.
In conclusion, big data analytics now generates hundreds of billions of dollars every year and human data is the new gold of the information age. There is no doubt that the information we're collecting is valuable, but its value is highly subjective depending on its use. Define your goal and then find what data you need to accomplish it.
Data quality can be the dealbreaker
Good decision making hinges on good data
Data quality is dependent on good CRM design