This past week a McKinsey article titled "Bias Busters, A better way to brainstorm" gave supporting stories for what promises to be a way to get us out of our silos and take a fresh approach to tackle our most challenging problems. In it, authors Renaldo, Koller, and Schatz, outline the power of structuring a conversation so that ideas are separated from a person's identity and position, unlocking new possibilities for transparency and organizational learning. Could this insight signal a new, more broadly adopted approach to corporate decision-making?
The article suggests that for management to get out of its own rut, it needs to stretch its organizational imagination by broadly tapping into ideas unfettered by the filtering and marginalization triggered consciously or unconsciously by identity and position. Properly managed, such an approach is not a threat to solid leadership but leads to a new type of leadership that operates from a broader perspective, unleashed by powerful organizational learning.
All this sounds good on paper, but how do you manage such a process? In my early experience as a Silicon Valley CEO, we discovered a new direction forward that required a company name change. I decided to open it up for suggestions, thinking a collaborative process creates ownership. A fire was started that was hard to manage. For that reason, many leaders resist, and even some management firms recommend control mechanisms for decision processes. A potential response to the McKinsey article could be a "good idea, but how do you manage such a process with the everyday pressures on operating a business under pressure?"
Resolution of the management problem requires technology that enables listening broadly and deeply while at the same time creating real-time analysis for productive management engagement. It turns out that work on technology to solve this problem, while in progress for years, is now available for broad adoption. The key is managing the complexity of alignment of individual preferences into an organizational model that reveals strategic pathways where leadership can move with confidence.
How do we deal with another leadership concern: "a leadership style based on trying to please everyone, consensus management, fails." As a young CEO, I wanted to listen broadly. I came into the CEO role from scientific research and academic background. Open processes based on peer review heavily influenced my leadership style. The answer to the fear that opening up to input leads to groupthink is to realize: alignment is complex. Resolution of such complexity requires a deep technical solution.
Work in collective intelligence, as popularized by Tetlock's "Superforecasters" sparked interest in tapping into the power of the collective mind. "The Difference" by Scott Page demonstrated mathematically that a cognitively diverse group of individuals are more accurate in their predictions and decisions than any individual alone. Together we have greater intelligence. Rather than focusing on the power of "crowd intelligence," collective intelligence focuses on how diverse areas of intelligence expand our perspectives, enabling a greater ability to forecast with confidence.
The subtitle of this piece is "collective reasoning at work." I have a background in knowledge-based expert systems, and some years ago worked in a company applying adaptive intelligence from anonymous inputs to product innovation¹. That journey led to the concept of collective reasoning. Collective reasoning is a way to peer into participants' minds in collective intelligence predictions and decision making. At the beginning of this piece, the image displays the "idea space" of a group involved in collective reasoning. The concept of "collective reasoning" is discussed in more detail in a Toward Data Science article titled: "Transforming Organizational Decision-Making with Collective Reasoning."
Collective intelligence points to a more accurate way to predict and decide by embracing cognitive diversity. Collective reasoning pulls back the curtain on the individual's reasons for a prediction or decision and opens the door to a process of generative brainstorming. Your reason may spark a new idea in me that may spark an individual who anonymously sees my idea. In collective reasoning, a junior engineer can suggest something that may spark an idea in the CMO's thinking that sparks a strategic idea in the CEO. All along, collective reasoning technology is learning a comprehensive map of the alignment structure of all contributors. The value of sharing and rating ideas anonymously is highlighted in the McKinsey article as a way to reduce bias in considering strategic pathways forward. Collective reasoning takes that observation to the next level for a new kind of scalable, generative, predictive system for agility in management.
The lead title of the McKinsey article is "Bias Busters." Six years ago, our team at CrowdSmart applied collective intelligence and collective reasoning to early-stage innovation. Could a group of diverse angel investors and experts predict which startup teams with innovations in products or services create the traction and support required to attract sustained investment?²
To "make it real," we raised a small fund so that we could participate in the stream of thinking associated with the identification and funding of seed-level companies, typically the territory of angel investors. To the point raised in the McKinsey article, structure creates focus. In our case, we used best practice research in angel investing funded by the Angel Capital Association. You can find details in the footnoted article.
The highlights of the project support and expand the "bias buster" concept. Over 50% of the high-scoring companies funded were led by female or minority founding teams (>40% female). This result is in stark contrast to venture capital and other innovation investment processes because it was unintentional. The participating evaluating teams were selected for their expertise. There was no attempt to reduce bias in any other way except to use a single-blind process while assessing the strengths and weaknesses of a particular innovation.
Bias reduction was not the only benefit. Accuracy was the leading benefit. Over 80% of high-scoring companies raised follow-on funding and success. Over $700 million of value was created(so far) by companies that scored in the "invest" zone. We calculated a quantitative diversity score (characterized by variance in quantitative scores) and a qualitative (reasoning diversity) score in the analysis process. There was a direct correlation of accuracy with increased diversity. Increased cognitive diversity leads to greater accuracy while reducing bias.
The McKinsey paper calls attention to an approach of anonymous sharing as a way to reduce bias. The CrowdSmart project created significant support for this approach to bias reduction, proving that separating the "idea" from the "identity" leads to predictive accuracy and reduced bias. CrowdSmart learned from this initial validation and turned it into a set of easily configurable cloud service applications targeted at product and service innovation.
In summary, the McKinsey paper points to an observation — separating identities from ideas improves decision-making, a process is worth scaling.
 For more background on the journey, see: https://towardsdatascience.com/getting-out-of-the-current-ai-rut-e5f10faa1983