What does it mean to be data driven?

FormulatedBy
4 min readJun 4, 2019

Exploring Data Literacy with Afternoon of Data

If data is only useful to highly-skilled data practitioners, it’s not living up to its original promise of making your organization smarter, faster, and more effective. We sat down with speakers with leading experts in the industry to explore the promise and pitfalls of data, promoting data literacy and ethics, and the role of data leaders in shaping organizational conversations around data.

Becoming Data-Driven

When your company is first making the decision to use data in it business analysis, it’s not always easy to know where to begin. Tom Schenk Jr., Managing Director at KPMG, suggests to “start with some challenges and prove some initial value behind a data-driven approach. Starting with low-hanging fruit is a great way to start and you need to build to tackle bigger challenges over time.” Emily Riederer, Senior Manager Analytics at Capital One, goes further: “In my experience, some of the most impactful “low hanging fruit” comes in figuring out how to effectively share and scale data analysis efforts in an organization. The more data analysts can work reproducibly and extensibly, the more momentum they can build to keep tackling the next big questions.”

Companies that are reticent to using data analysis are definitely missing out. According to Addhyan Pandey, Senior Director Data Scientist at Cars.com, “companies that don’t use data in their decision making process suffer through a lot of uninformed operational inefficiencies. Many data (and data science) applications can help companies operate more efficiently by learning rewards and risks in every investment they make prior to making those investments.”

Reaping the Benefits

There’s no question that becoming more data-driven can payoff in a big way. “For example,” Schenk says, “we’ve seen huge ROIs to government using data. In Chicago, they are more likely to find restaurants with poor sanitary practices, find where West Nile Virus might pop-up, and children who are likely to have lead poisoning.”

But data alone can’t solve everything. Pandey says that “most of the time just using data is not enough. My team and I have brought the practice of using data, math and scientific process together to solve business problems and so far it has paid dividends.”

Avoiding the (Data) Pitfalls

Making the jump to being data-driven isn’t without risk. Like any tool, it must be used correctly to be both useful and safe. “The bias towards equating data with scientific rigor or objective truth is the biggest pitfall,” says Riederer. “This can manifest in many ways — from not interrogating data quality or implicit biases at the start of an analysis to being seduced by false confidence or precision of the final results. Data is most powerful when we understand its shortcomings and limitations along with its power and potential.”

Knowing the potential of your data means knowing where your data came from. “Data is collected in the context of administrative processes, surveys, and users,” says Schenk. “It can be easy to focus on the data you have, but often, you need to be thinking about the data you don’t have and why you don’t have it. That will help you understand the data you do have and what you can learn from it.”

Even so, it’s up to data scientists to help guide the data end-user. “With data, common pitfalls originate from common assumptions,” says Pandey. “We assume that the consumer of the data understands the data but that is not always true, and therefore it’s our job to communicate better, to make sure that definitions of certain entities are the same and the underlying assumptions around data collection or sampling are shared. Not having a good understanding of these assumptions can result in biases in the decision making processes.”

Broaden the Conversation

In the end, data is a resource for the entire organization and should be open and democratized whenever possible. “Reports and dashboards are a great way to bring non-practitioners to the table, but make sure that these efforts are truly empowering and not gate-keeping,” says Riederer. “Enabling a broader user-base to independently access and analyze the information they need is critical to working at scale, avoiding misunderstanding, and creating a healthy data culture.” But that doesn’t mean that the work of a data scientist leader stops after data is open. “The best leaders in data are clear communicators,” says Schenk. “Ask non-practitioners what is challenging them. Explain how data might be a piece in the puzzle that can help them. Then execute so you can show value.”

Check out our virtual events→ Data Science Virtual Salons in 2022:

- DSS Virtual | AI & Machine Learning in the Enterprise (April 20, 2022 // Online) #DSSVirtual

- DSS Virtual Mini Salon| Retail & Commerce (May 18, 2022 // Online) #DSSVirtual

- DSS Miami |AI & Machine Learning in the Enterprise (September 22, 2022) #DSSHybrid

- DSS Virtual | Media & Advertising (November 16, 2022) #DSSVirtual

- DSS NYC | Finance & Technology (December 7, 2022) #DSSNYC

--

--

FormulatedBy

Specializing in conversation-centric strategy in demand generation. Serving over 100+ B2B companies from machine learning to DevOps. EST 2015 #demandgen 📈