91视频

The challenges of making decisions in the face of globalization, Big Data, growing stakeholder diversity, and blurred boundaries created by multiple goal complexity is compelling forward-looking organizations to emphasize consensus when making important decisions.

John Keith, co-founder of Portland-based Lucid, which offers consulting and technical services to help organizations optimize meeting effectiveness, contrasts consensus to two other major types of decision-making models. One is command and control, intuitive to most people as a decision making where leaders make top-down decisions without consulting their teams. The other is collaborative decision making, where designated leaders still make the important calls, albeit based upon interlocution with others team members possessing relevant information.

Consensus is a more democratic and dynamic form of decision making. It aims not simply to foster trade-offs and compromise within a group but upon making important decisions within a structure that a whole team can buy into and take ownership of and, optimally, be accountable for throughout implementation.

There鈥檚 no shortage of decision support software on the market with features useful to facilitate command and control and collaborative management, including applications that help integrate Big Data and produce predictive analytics, business intelligence and competitive intelligence, mapping, and visualization. But according to research and advisory firm Gartner, software platforms incorporating algorithms dedicated to generating consensus are still in their infancy.

Ideally, an application dedicated to support consensus would integrate the following functions:

• Labelling and visualization for easy understanding by end users not well versed in technology;
• Uniform dissemination of critical information to all group stakeholders;
• Sharing of all stakeholder ideas, perspectives, and priorities;
• Algorithms to establish areas of stakeholder agreement and disagreement
• Predictive modeling of competing and complimentary scenarios;
• Iterative voting, weighing and ranking of proposals among stakeholders;
• Near consensus alternatives when full consensus is not possible;
• Clear and concise reports mapping out decisions, the grounds upon which they are based, and the steps needed to move forward ; and
• Capability for stakeholders to revisit and revise their agreements as necessary during implementation.

This looks like a big ask, but in the age of artificial intelligence, maybe not so much. AI is likely to continue its inexorable march toward replicating the cognitive performance of people. As machines learn more about how humans think and express themselves, they will better be able to parse stakeholder ideas and perspectives and mold these into 鈥渟hared thought鈥� embodying common interests suited to consensus.

Under any circumstances, there will be limits upon what AI-driven consensus building apps can achieve.听 Most prominently, these applications will only be as reliable as the commitment of involved stakeholders. Optimization will thus depend upon stakeholders who bring the following attributes to the table:

• An ability to embrace a common goal and commitment to collectively achieving a desirable outcome by recognizing that the overall success of the group is preponderant;
• The willingness to engage by sharing opinions, listening to those of others, and remaining open to new ideas or directions; and
• A willingness to follow and adhere to well-defined and transparent processes while avoiding actions that could be interpreted as self-seeking manipulation.

Other factors are likely to pave a path forward for machine-assisted consensus. In addition to business, government 鈥� which often looks for consensus when making contentious decisions 鈥� looks like a prime market for adoption. For example, earlier this spring the Washington State Legislature let out a request for proposals for a contractor to facilitate consensus making among competing parties for development of recommendations for statewide regulation of car sharing services. Building consensus will be a tough road to hoe, as the stakeholders include Uber, known for vehemently resisting regulation. Maybe automation would help.

Another factor is increasing adoption of 鈥渨e work鈥� groups 鈥� loosely federated groups of people, pulled together in an ad hoc fashion as needed for specific endeavors 鈥� in place of static organizational team structures.听 This will attach increased importance to speed and nimbleness in interpreting and articulating the preferences and proclivities of shifting stakeholders as they populate the 鈥渨e-work鈥� ecosystem, an ideal role for machine learning.

Optimizing software apps for consensus leadership may require reassessment of organizational cultures. To many, the word consensus conjures a time-consuming recipe for 鈥渒icking the can down the road.鈥� But this should become less of a concern as automation again promises to speed up the consensus-making process.

In general, command and collaborative leadership is best suited for situations requiring reliability and certainty and where the variables are known, whereas consensus best suits scenarios defined by ill-defined problems with that require creativity and departure from norms. Consensus is unlikely to ever supplant more hierarchal approaches to decision making. But boosted by new apps and AI, it may see much more widespread use.

听