In my previous post I explained that even if your organization does not have anyone with data steward as their official job title, data stewardship plays a crucial role in data governance and data quality.
Let’s assume that this has inspired you to formally make data steward an official job title. How should you go about finding good candidates for such an important role? You could take inspiration from some of the examples noted in the bestselling book Rework by Jason Fried and David Heinemeier Hansson.
One example is an airplane manufacturer that determines the right people to promote internally to management positions using a role-playing exercise simulating the typical day of an executive. Candidates interpret reports, conduct meetings, arbitrate a dispute between employees and respond to complaints from irate customers. However, candidates aren’t told until after the process is over that the reports, meetings, disputes, and complaints were all staged. The point of the simulation is to observe how candidates perform actual work tasks as opposed to relying on résumés and standard interviews to pick someone who can theoretically do the job.
Another example is a major international airline that uses a role-playing exercise to interview flight attendants. Three job candidates are brought into a room to be interviewed as a group by a panel of senior flight attendants. During the interview, two of the candidates get into a heated disagreement over the best answer to one of the open questions posed to the group. Unbeknownst to the other candidate, the two arguing candidates are actually also senior flight attendants. The only real candidate is being closely observed to see how they react, whether they attempt to defuse the argument and, if so, if they can do so successfully.
This might seem like an unusual technique until you realize that an international flight attendant’s job is performed while trapped for hours in an enclosed metal cylinder traveling hundreds of miles an hour, thousands of feet in the air. To do this job well, you must be able to quickly defuse a heated disagreement or settle down a disgruntled passenger.
My point is that finding good data steward candidates is often more challenging than you think.
The characteristics of a good data steward are varied. They won’t always show up on a résumé, be revealed in a standard interview, or come from the part of organization you’d expect. Data steward is a tough role to play. A role-playing exercise could help. Try letting candidates play data steward on upcoming projects. Observe how they perform tasks such as interpreting data profiling reports, conducting meetings about data quality requirements, and providing two-way translations of the business and technology aspects of data governance policies.
