How to Organize the Citizen Data Scientists

What is a Citizen Data Scientist (CDS)?

Gartner defines a citizen data scientist as a person who can perform both simple and moderately sophisticated analytical tasks, but whose primary job function is outside the field of statistics and analytics. Citizen data scientists are super users of various BI and statistical tools at an organization.

These power users have developed or are keen to develop a skill-set that is in very high demand. Additionally, they bring deep domain knowledge and organizational context which an expert data scientist may lack. Therefore, organizations must recognize them and find ways to leverage them to the fullest. 

These days the quest for the next competitive advantage has to be constant and consistent. Unprecedented amounts of data and technological innovations can give such an edge, but only to those who are seeking it while maintaining daily operations. 

Empowering citizen data scientists can build such agility into the culture of the organization. 


As the pace of the market and the complexity of data projects increase, citizen data scientists can help capitalize on opportunities and dodge threats that were previously unforeseen. But in doing so they often run into challenges like:

  • Silos that make information less accessible and more disconnected
  • Policies and procedures that clog new ideas
  • Pressures to meet benchmarks that create tunnel vision
  • Complacency or insufficient buy-in that blocks innovation
  • A limited number of leaders to inspire and sustain change

John Kotter lists these 5 factors as limitations of a hierarchical organization. He further notes that when organizations are structured as hierarchies, they are optimized for efficiency rather than speed. While a pyramid structure meets the daily demands of running a company, it is not equipped to meet the bullet-like pace of the business world. It often gets in the way of identifying the most critical elements early enough to execute innovative responses.

Citizen data scientists run into these issues because their projects often touch multiple levels of the hierarchy, horizontally and/or vertically. The hierarchical reporting structure can unwittingly get in the way of their ability to respond to market challenges creatively and with speed.

A common way to deal with such inherent limits is to bring together interdepartmental teams to tackle problems or utilize resources that span across multiple departments. Such efforts can make silo walls thinner but the walls still exist. The policies and procedures can sometimes multiply because of the numerous departments involved. New inter-department benchmarks can impede just as much. The teams trade silos for speed bumps and potholes.

Solution = Dual Operating System

To harness the power of CDSs within the hierarchy of large organization, Kotter’s dual operating system might be the key. 

The existing structures and processes that together form an organization’s operating system need an additional element to address the challenges produced by mounting complexity and rapid change. The solution is a second operating system, devoted to the design and implementation of strategy, that uses an agile, network like structure and a very different set of processes. The new operating system continually assesses the business, the industry, and the organization, and reacts with greater agility, speed, and creativity than the existing one.

John Kotter's Dual Operating System for Citizen Data Scientists
John Kotter’s Dual Operating System

A dual operating system would be such that the network of CDSs complements the existing hierarchy. The CDS network would be the more dynamic counterpart of the organization that can take on and prioritize projects as needed. While the hierarchy may not change quarterly or even annually, the network can evolve as needed, as often as needed. Being outside the bureaucratic layers, allows individuality, creativity and innovation to proceed in the network at much greater pace.

Populated with employees from all across the organization and up and down its ranks, the network liberates information from silos and hierarchical layers and enables it to flow with far greater freedom and accelerated speed.


The Eight Accelerators

Kotter lists the following 8 processes to enable the CDS network to function:

  1. Urgency on big opportunity
    • Most companies, regardless of the industry, recognize that becoming data-informed will give them a competitive advantage. The opportunity to become more data literate and analytically mature is the foundation upon which the agile CDS network can be built.
  2. Guiding coalition (GC) of volunteers
    • Kotter recommends selecting the GC to represent each of the hierarchy’s departments and levels, with a broad range of skills to ensure that the information is gathered and processed in a way that the hierarchy couldn’t. It also allows transfer of information between the hierarchy and the network. It keeps them from becoming two giant silos. The GC for the CDS would help identify which data and analytics projects should be launched and help line up resources to accomplish that. 
  3. Change vision and strategic initiatives
    • The vision describes what success looks like. It is what keeps all the projects pointing to the strategic true north. For the CDS network this could describe where the organization wants to be in terms of analytical maturity. E.g. as an organization we want to evolve from a Level 2 (opportunistic) to Level 3 (systematic) by aligning the D&A strategy with the business strategy. The GC would then help identify projects that are critical for achieving this vision.
  4. More and more volunteers
    • The network invites people to bring their ambition, creativity, and commitment. These are employees waiting to break the boxes and bring new ideas to life. Being a part of the CDS network is an opportunity for them to hone analytical/statistical skills, learn new programs, make new professional connections, and get a peek into other career paths.
  5. Barriers knocked down
    • “Design and implementation occur in the network and are instituted within the hierarchy.” CDS network gets to experiment and set the path of least resistance for others in the organization. All the hierarchal silos can put their heads together to act quickly and efficiently.
  6. Wins celebrated
    • To increase the buy-in, the network must celebrate every small win. By making wins visible to others in the organization, the skeptics can be won over, the volunteer army grows and the excitement around the initiatives rockets. The CDS should celebrate the launch of a new dashboard, adoption of a new metric, buy-in from a new department, access to a previously inaccessible dataset, establishment of a new master dataset. Success breeds success.
  7. Relentless action
    • “When an organization takes its foot off the gas, cultural and political resistance arise.” It is important to keep cultivating new ideas and exploring new opportunities. For the CDS network, this could mean exposing them to others in the industry, via guest speakers and conferences. 
  8. Changes institutionalized
    • “A new direction or method must sink into the very culture of the enterprise—and it will do so if the initiative produces visible results and sends your organization into a strategically better future.” Supporting and trickling the CDS network’s initiatives into the organization will ultimately permeate the much sought after data culture.

A CDS network can help fill the analytical skills gap that organizations are facing right now by expanding in scale, scope and power the smaller informal networks that accomplish tasks faster and cheaper. Networks tend to evolve more organically based on the skillset, interests and needs of its members. As such a dual system will leverage the citizen data scientists in a way that the hierarchy couldn’t by making them citizen data accelerators.