3 Steps to Train More Citizen Data Scientists

In Asia and elsewhere, citizen data scientists are increasingly coming under the spotlight. Easier to find and more cost-effective than expert data scientists, they complement the work of the latter to give forward-looking enterprises the insights they need for data-driven decisions.

Often armed with years of hard-wrought business and industry domain expertise, citizen data scientists are also well-positioned to integrate data analytics and machine learning output into the heart of the business for a more immediate impact on the bottom line.

Identify citizen data scientists

Despite the growing attention, the fact that hardly anyone advertises for a citizen data scientist means there is a lack of clarity around their value and job responsibility. Obviously, this must be addressed before businesses can identify such employees.

I liked Tibco’s definition of a citizen data scientist the best: “A citizen data scientist is a knowledge worker without formal training in advanced mathematics and statistics that uses applications to extract high-value insights from data.”

In a nutshell, a citizen data scientist is essentially any employee who has the skills of a data scientist but not the qualifications. And we all know people like that: Someone good with numbers and who can draw inferences from data that – once explained – appears obvious in hindsight.

In a blog entry, BMC’s Stephen Watts outlined some characteristics that a citizen data scientist should ideally possess. Some of the most compelling are:

• Divergent thinking: Someone who can think outside the box and make the connection to the data.
• Organizational context: A worker who “understands the vision, mission, and needs” of the organization.
• Ability to access information meaningfully: Have an above-average ability to draw meaningful conclusions from the data in front of them.
• Able to outline business value: The ability to communicate their analysis of business data and its implications to the team.

To get ahead, businesses must first identify and engage these employees as citizen data scientists. Remember, these employees often have other responsibilities to manage, so start off with a lower bar, assigning them tasks such as validating data quality, merging data, or identifying data sources.

Democratize data science in the organization

While it is possible to impose an organization-wide mandate to give citizen data scientists what they need, siloed data might not be worth the effort of extricating. Even valuable data stored within disconnected repositories might be ignored due to how hard it might be to access.

Ultimately, data democratization is the bedrock of any initiative to raise more citizen data scientists. A culture with a strong awareness of data or high data literacy can serve to encourage more data-savvy employees to step up, as well as simplify their jobs when communicating with colleagues.

It helps that there are plenty of good business intelligence tools today, including low-code or no-code tools that allow more employees to leverage machine learning tools or perform relatively sophisticated analyses. Expect the barrier to come down further over time; organizations with a strong pool of citizen data scientists will be well-positioned to take advantage of it.

As I noted in “Building an Organization to Win With Data”, the rewards of democratizing data are worth it. Top performers in machine learning can have more than twice the impact in half the time compared to the average company.

Focus on strategic projects

A recent report in the Harvard Business Review (HBR) suggested that organizations should turn their data science efforts on the problems with long-term strategic importance, instead of focusing on where they have the most data as is commonly done.

Citing the example of a hypothetical media company facing two choices, one being to deepen user experience using data generated by its apps, or to leverage data to inform a once-in-two-year licensing bid, HBR observed that doing the latter poorly can result in greater harm.

A second suggestion is to focus on initiatives with the highest probability of project success. A “stone-cold sober evaluation” of potential data science projects must be made, and in the absence of set answers, to do more small data projects – which falls squarely into the domain where citizen data scientists can make an impact.

Even with larger “moonshot” projects, a strong team of citizen data scientists can also free up an organization’s data scientists to focus on these strategic opportunities. Or as HBR sums it up: “Data science is about people and the more strategically and broadly you bring these people and data together, the better results you’ll see.”

Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].​

Image credit: iStockphoto/cofotoisme