Automating DEI, a benefit or detriment?
As a solid advocate of technology, I’m a fan of leveraging automation when there’s a clear business case. When it comes to DEI Analytics & Reporting, I am easily persuaded that there’s a business case for automating data collection (including visualizations and dashboards) - data collection is a repeatable task and the manual effort is substantially reduced, likely by 95%+, only requiring manpower time to track and chase the ever-present human tendency to disregard auto-generated email follow up to complete a survey - so you need manpower for the personalized email or the instant message to those who have not yet responded to the survey. But once the DEI data is collected, can the analysis be automated? My perspective is that a DEI analysis requires human intervention because if the truth be told, telling the DEI story behind the data is less likely to be fit for repeatability - there are themes to be gleaned as supported by the data insights, a narrative that digs into the details and highlights the positives and negatives (as well as neutrality outside of normal standards), recommendations to target improvements in line with best practices or to conform with policy or regulations, and a conclusion to reinforce key takeaways - perhaps, the analysis phase is where the lawyer in me takes rule because I want to tell the whole truth and nothing but the truth, although in a bold and balanced way.
When it comes to DEI, so much of how the analysis takes shape is driven by human emotion. How one feels about his or her experiences determines whether inclusion and belonging are being fostered in the workplace. Whether policies or practices are fair is initially gauged by feelings and then weighed against outcomes. The fairness analysis requires an in-depth review of policies, processes, patterns and practices, and oftentimes a gathering of additional qualitative data to ensure objectivity prevails. And while a machine can be programmed to do an if/then analysis, the variety of ifs and the less predictable thens may defeat the efficiency benefit of automation. In other words, repeatability could break down when analyzing DEI perceptions against realities and thus an automated assessment may foil when considering the more fluid factors attributed to fairness - that’s where context becomes central to the analysis in establishing the why and circumstances surrounding the data. And this is where human intervention, in my opinion, is superior to automation.
A multi-layered, contextual analysis may require cross-referencing multiple data sources and weighing the credibility of each source. And while it’s true that a formulaic approach could be created to standardize such an analysis and automate the outcome, it seems somewhat like the anti-thesis of DEI, which is about differences not sameness. The underlying premise of a DEI analysis seeks to understand if the uniqueness of people is being valued; to a certain degree, a standardized approach to a DEI decision-analysis diminishes the value of the outcomes. For example, the performance management policy by itself may not shed light on certain manager’s determination that an individual employee (or a group of employees) was not ready for promotion. With an automated analysis, standardized decision-making may not consider that facts and circumstances might vary significantly and configuring a decision tree to address all the possibilities could become cost prohibitive (or worse, require manual re-work, generating even more costs than anticipated for a rationally based and comprehensive analysis).
Still, I trust technology to drive efficiencies, accelerate speed of delivery, reduce processing errors and improve productivity at a minimum; however, the human element in a DEI analysis is a must-have to tell the data story and appropriately target meaningful outcomes. As such, the analysis and relevant reporting of those outcomes must be customized, not standardized. Although facts may be predictable, circumstances might be unique to an organization and its people. Human intervention will likely be required to make the distinction. Therefore, organizations should carefully determine if a DEI automated solution alone will effectively guide their DEI journey based on their intended objectives. It begs the question of whether your DEI analysis ends with the data or if it continues with a data story; if the latter, then your DEI data story should be customized, not standardized - to do so necessitates human intervention in the analysis and subsequent reporting of your DEI assessment. Nonetheless, there’s no question that technology has its place when it comes to DEI - my view is that data collection (including visualizations and dashboards) is a clearly established space for DEI automated technology as part of DEI assessment. However, it’s debatable if a DEI analysis (and the reporting deliverable) should be automated. What’s your perspective on leveraging automation (or other technology) for DEI analytics and reporting?
To learn more about the DEI Analytics & Reporting niche service offering at TULIP Advisory Professionals LLC, visit www.tulipadvisory.com.