On Consensus

In the early 1990’s, the Internet was in its civilian (non-university) infancy and the influx of ideas, people and technology created roadblocks to the growth of the Internet. There are many parallels today between Business Intelligence/ Analytics (BI/A) now and the Internet then. In the early 1990’s, some countries were well advanced in their implementation of the Internet, yet other modern countries were still woefully behind. The same is true today with BI/A, some industries are running point, and others are still using tech from 2008. The inability to effectively and efficiently address issues and to move forward was a problem then…and it is a problem now.

In 1992, the International Engineering Task Force (IETF) held its 24th Proceedings at MIT and Dave Clark, a senior research engineer delivered a presentation that gave birth to what Pete Resnick claims is the credo of the IETF;

We Reject: Kings, Presidents, and Voting

We Believe In: Rough Consensus and Running Code

How does this apply to Business Intelligence / Analytics (BI/A)?

We all have our circle of competence and we all have our stakeholders and our chain of command. These factors produce biases…its not personal, it just is. The elimination of a king or president removes the ability to have one person approve or deny something because of these biases.

Additionally, voting can be problematic. Some people wish to avoid conflict so they will refuse to vote…or they will vote for their friend…or maybe vote the way they think their manager wants them to vote.

“Lack of disagreement is more important than agreement”

The goal is not unanimous, uncontestable agreement…the goal is not to “make everyone happy” . The goal is to get to a point where we can meet the customers needs in the most efficient and effective way. Its to do what is right and what is needed, not what the perfect outcome that we can post about on LinkedIn.

Resnick points out that coming to a consensus is different than determining consensus. I know it is subtle but as Resnick states “engineering…involves …tradeoffs”. It is important to determine and understand what these tradeoffs are? There must be honest conversation about the tradeoffs. When we discuss the trade offs the group must not be allowed to simply “not like the idea” or hide behind fictious “best practices” or use the catch all phrase “that’s not the way we do it here”.

This isn’t about “gotcha’s”, the process is to uncover the trade-offs that our biases hid from us…its about wrestling with the ideas to see if they really work for the system. This is not asking “Who doesn’t like A?”, this is trying to understand what is it about A that is a problem for the system? What is it about A that you can’t live with? What are the downstream impacts of A?

Rough consensus is not unanimous consensus

As we wrestle with the ideas we are not trying to negotiate a solution that makes everyone happy. We are trying to do what is best for the system, that is the consumer and the organization. Resnik writes that “rough consensus is achieved when all issues are addressed, but not necessarily accommodated“.

Addressing is not solving…addressing is investigating the tradeoff…its using due diligence to ensure that the negative impact being mentioned is either mitigated or determined of minimal risk. Addressing the issue is not supposed to leave the originator with a happy, warm feeling…addressing the issue is supposed to be done in an intellectually honest way. If, however, we find there are indeed issues that must be accommodated, then it is our responsibility to accommodate them.

Ego’s Control Outcomes

Gang, listen…the bottom line is that culture must change. In the 1990’s if we let votes win, we would all be using America Online today. In the 1980’s if we let one person decide our tech, we would not have the mouse, or a graphical user interface.


We have to remove ego’s from the decision making process.

We must remove “closed loop” thinking from the decision-making process.

We are a team of teams, we must want to do what is best for the organization and the consumer.

We must be willing to do what is right, not what we want to do.