By: Elizabeth Autumn
As our technological capabilities grow (especially with the explosion of artificial intelligence, machine learning, and Software as a Service, or SaaS), some are starting to think of computing differently.
I am one of those people.
And I’m not alone in my enthusiasm for tech.
According to Pew Research, nearly half of the workers whose primary job function includes computer use feel that technology helps them be more productive at work. While some popular business publications will have you believe that technology will erode focus and productivity on the job, those who embrace technology find that it affords them more flexible work arrangements and wider networking opportunities (Purcell and Rainie, 2014).
As a data advocate searching for business insights in the bottomless “big data” ocean curated by financial services, my computer is not just a tool; it is a valued co-worker.
Computers as Co-Workers
But, of course, this analogy is overreaching. Humans and computers are not the same.
We have different triggers for different events. We have different responses to different stimuli. We are driven by different processes.
What we both have are seemingly endless possibilities to create positive outcomes through a mutually beneficial partnership with each other. If we are to unlock the full potential of our abilities, a symbiosis must exist between players (automated or otherwise) in the advancement of industry. For that symbiosis to exist, we must address the needs of all players.
It is easy to know what human players need. We are humans ourselves, after all. We interact with other humans every day. Some of us are even responsible for caring for other humans as well. We know what they need because we all need the same things. But what about the computer players? What do they need?
The Hierarchy of (Data-Aware) Needs
To arrive at the answer, we will work backward. The most basic human needs (according to American psychologist Abraham Maslow) are physiological; these are at the base of the human requirement structure. They include sleep, sex, shelter, overall physical condition of the environment, water, and, of course, food. These foundational needs underlie all of the human experience, from critical thinking to emotional processing to physical wellness (MasterClass, 2020).
So, what is it that underlies the full experience of being the world’s most powerful supercomputer, or a rocket that will take a new generation into space exploration? What is the primary input — basic need — for my computer co-worker to be able to help me uncover valuable insights and generate information for important decisions every day?
But will just any ol’ data do the trick? Of course not!
Can we just feed as much data as we can possibly scrape from every source into one place, and let people format it and change it for their particular use? Eat ten giant Chipotle burritos for lunch and tell me how efficient you’re feeling afterward. Overloaded data systems commonly lack consistency or conformity between sources. They may also lack an adequate audit trail, and, depending on how tightly access is monitored, could leave your firm open to data mishandling or even a breach.
What if we give computers only the data we think they need at any particular moment? Well, unless you have a company policy that your employees are only allowed to eat two peanuts at a time in two-hour intervals until 5:00 p.m., this method is faulty as well. Starving your data reduces your analytical capabilities, narrows the toolset you can work with, and drastically slows your time to insight.
Success Is in the Journey
As with many things, data quality is a journey, not a destination. It’s not about having the right amount of data, just like we don’t often think about what is the right amount of food. We eat when we’re hungry, we don’t when we’re not. Computers don’t have an appetite for data. They don’t crave it. So, it is up to us as their operators, their co-workers, and (sometimes) their caretakers, to craft a data diet that contains data that is fit for use, meaning it is:
- Complete – We know where we have missing or unusable data, and we know who to contact with data quality questions or issues surrounding incomplete data.
- Consistent – Data values are disseminated according to standards laid out by company data governance rules, such as a data dictionary or business glossary.
- Unique – The data is not redundant or does not create duplicative processes.
- Accurate – The data is up to date and correct.
Why It's Worth the Effort
For what it’s worth, I get it. Data quality is boring, and unless you live in that world 24/7, it can be hard to understand what all this means and why it matters.
But I am sharing this with you because YOU have the power to leverage your data in a way that adds benefit to stakeholders across the value stream. How?
- Your employees will see the returns when tedious, manual processes can be automated, allowing them to create a better work-life balance and maintain lower stress levels in their day-to-day.
- Your borrowers will see the returns when your systems can correctly time and generate approvals for lending requests and notices regarding their business with you, creating a better communication pathway and reduced noise.
- Your business owners will see returns in retention, sentiment, and operational efficiencies that could produce between $8.2M and $100M annually in savings from the placement of data quality initiatives (Pitney Bowes, 2014).
All of us, as humans, require food to function. Similarly, our computer counterparts require data if they are to function as expected. And, just like the quality of the food you eat impacts your performance, the quality of the data you “feed” your processes will impact your analysis, your decision-making, and ultimately, your business results.
Elizabeth Autumn is a data advocate with experience in operational reporting and analysis, workflow efficiency, report design, project management, and process automation. She works with operational data for the service center and her insights seek to address common client pain points and create value for State National’s entire portfolio. She also enjoys animals, heavy metal music, learning about new technology, Python programming, and data visualization.
Purcell, Kristen, and Lee Rainie. “Technology’s Impact on Workers.” Pew Research Center: Internet, Science & Tech, Pew Research Center, 30 Dec. 2014, www.pewresearch.org/internet/2014/12/30/technologys-impact-on-workers/.
“The ROI of Data Quality.” Pitney Bowes, 2014, www.pb.com/docs/US/PDF/software/customer-information-management/wp-roi-data-quality-93734-amer-1401.pdf.
MasterClass. “A Guide to the 5 Levels of Maslow’s Hierarchy of Needs – 2020.” MasterClass, MasterClass, 10 Apr. 2020, www.masterclass.com/articles/a-guide-to-the-5-levels-of-maslows-hierarchy-of-needs.