Better Decisions in Complex Organizations: Three Alternatives to Consensus and Majority Vote.

At a client engagement, a discussion once arose about how AI-driven work and scaled agile practices could be reconciled. Could the still highly experimental AI work be organized on a structured Kanban board or would the structure slow down progress? Arguments became more nuanced, perspectives more diverse yet the decision didn’t get any easier. In the end, you either agree on a solution that no one is truly happy with, or the loudest voice in the room prevails. This is exactly the point where it is worthwhile to take a closer look at the way we make decisions.

Decisions are the heart of every organization. Yet in practice, we consistently encounter the same three patterns: the endless struggle for consensus, the narrow majority vote, and the top-down power call. All three have their place, but all three have blind spots that can become costly, fast, in complex environments. This article introduces three alternatives that enable additional viable resolutions: consent, systemic consensus, and the advice process.

The Human Factor: Why Are Decisions So Difficult?

Before diving into the tools, it’s worth pulling back the curtain. Decisions in organizations are not purely technical processes. They happen between people with their experiences, interests, identities, hopes, fears, and the stories they tell themselves about a given situation.

Anthro-Complexity: We Are Not Machines

When we talk about decisions in organizations, we are operating in what Dave Snowden calls Anthro-Complexity: a world where human systems do not function like machines. That may sound trivial, but the practical implications are significant. Many decision architectures operate as if social systems can be controlled like technical installations with clear cause-and-effect logic, linear dependencies, and predictable controllability. In human systems, that is only true to a very limited extent.

Anthro-Complexity reveals that humans are never purely rational decision-makers. We are simultaneously Homo faber the tool-building, shaping being; Homo ludens who is playfully experimenting, improvising, and applying rules situationally; and Homo narrans interpreting our environment through stories. On top of that come three formative forces that Dave Snowden calls the 3Is: Intelligence, Intentionality, and Identity.

Intelligence here does not simply mean cleverness, but the human capacity for reflection, abstraction, and higher cognitive processes far beyond mere stimulus-response reactions. Intentionality describes that people not only react to stimuli but set goals, weigh priorities, and make conscious decisions. Identity introduces the question of how we show up differently in different contexts and how tightly or loosely we are bound to certain self-images – a dimension that is often underestimated in social systems and manifests in questions of recognition, respect, and dignity.

A core reality of complex systems is particularly relevant: cause and effect can only be identified in hindsight, not in advance. This is precisely why the longing for that one objectively correct decision is usually deceptive. In many situations, there is no perfect decision, but only more or less viable ways of dealing with uncertainty.

“[...] once a snowflake has formed, I can tell you how it formed… but it doesn’t mean I can predict what the next snowflake will look like because the pattern doesn’t form until it forms.” ⁠

*Dave Snowden - Coping with complexity - management.co.nz

The Illusion of Certainty

In her excellent book Thinking in Bets, Annie Duke describes very precisely why decisions so often lead us astray: we confuse the quality of a decision with the quality of its outcome. This is understandable, outcomes are visible, processes usually are not. But especially in complex contexts, this confusion is dangerous.

Decisions have two fundamental sources of uncertainty. First, before a decision, we always operate with incomplete information about a complex landscape. Second, after the decision, chance, luck, or bad fortune influence what happens next.  A good decision can therefore lead to a bad outcome, while a poor decision can look successful with a bit of luck.

This gives rise to the Outcome Bias, which Annie Duke also describes as “Resulting”: past decisions are judged in hindsight solely by their outcomes. Someone who runs a red light and doesn’t get hit hasn’t made a good decision, they just got lucky. Conversely, a carefully prepared strategic decision can fail, even though the decision-making process was reasonable, reflective, and sound.

This is where a brief excursion into psychology helps. Hindsight Bias refers to the phenomenon where an event that has already occurred suddenly appears to have been predictable all along. From the Cynefin context, the related concept of retrospective coherence is particularly useful: once something has happened, we construct a coherent narrative around it. This coherence in retrospect feels plausible but says little about whether the development was actually predictable beforehand. This is precisely why retrospectives should not only ask about results but also about decision premises, available information, and plausible alternatives.

“The interesting thing about complex systems is no matter how often you look at the past, you can’t predict the future.”

*⁠Dave Snowden - Coping with complexity - management.co.nz

What Makes a Good Decision?

If outcomes alone are not a reliable measure of quality, the decision process itself becomes interesting. A good decision is then not necessarily the one with the best outcome, but the decision that was made under the given circumstances with the most appropriate procedure possible. It is less about being right than about dealing wisely with uncertainty. This can considerably relieve the decision-maker.

Annie Duke describes six steps for better decisions in How to Decide. Covering them in full here would take us too far from the core topic; but the underlying idea is valuable: treat decisions as hypotheses. A decision is not a certainty, but a bet on a possible future, made with the best available knowledge at the current point in time.

There are more possible futures than the one that actually happens ⁠*

*Annie Duke - "How to Decide" (p 7)

Decision Pressure in the Age of AI

The pressure on decision-making processes is further amplified by artificial intelligence. AI accelerates research, analysis, communication, and implementation, and thus also the rate at which decisions arise. What used to take days or weeks is now prepared, simulated, or directly implemented in hours. And this happens repeatedly at increasingly shorter intervals.

However, this also increases the risk. Decisions can be operationalized faster, intervene more strongly in systems, and be harder to roll back. The faster an organization can act, the more important it becomes not only to decide quickly but also sustainably. Another option may be to conduct multiple safe-to-fail probes in parallel and even in contradicting directions, but that is a topic for an additional post.

The Common Methods and Their Weaknesses

In many organizations, three decision patterns dominate. They are familiar, well-established, and often culturally deeply embedded. That’s precisely why it’s worth taking a sober look at their strengths and weaknesses.

Consensus: The Search for the Best Solution Since Sliced Bread

In consensus, all participants jointly search for a solution that everyone can agree to. On paper, this sounds like the ideal form of collective reason. No one is overlooked, everyone is heard, and the outcome is a jointly owned decision.

In practice, however, significant hurdles emerge. Consensus processes have a high potential for gridlock, since any individual can block the entire process. They tend to converge toward the lowest common denominator, an option that bothers nobody but genuinely convinces nobody either. Additionally, there are stalemate situations and false harmony: people agree not because they are convinced, but because it seems strategically appropriate given their current identity or intentions.

Majority Vote: Winners and Losers

The majority vote is the familiar alternative to consensus. Several options are put to a vote, the majority wins, and the decision is made. The process is fast, clear, and seemingly democratic.

But here, too, there are obvious weaknesses. Minorities are overruled and may even experience this as powerlessness, especially if the same patterns repeat. The procedure favors faction formation and polarization, as it becomes strategically sensible to organize votes rather than integrate perspectives. Narrow results also create questionable legitimacy: a 51:49 rdecision is formally resolved but culturally often highly fragile.

Power Call: The Lonely Path at the Top

In a power call (or hierarchical decision), a single person makes the decision – based on their role, expertise, or formal authority. In acute situations, with clear responsibility, or under high time pressure, this is viable and congruent. The method is fast, unequivocal, and leaves no doubt about who decides and who bears responsibility.

The power call becomes problematic when it becomes the default. Then the breadth of information suffers, because no single person can oversee all relevant perspectives alone. Team commitment can erode due to lack of participation. And there’s a growing risk that formal authority gets confused with subject matter expertise with all the well-known risks of power abuse, dependency, and cultural narrowing.

Deciding Differently: Three Alternatives for Robust Decisions

The following three methods address different weaknesses of the classic procedures. None of them is a silver bullet, as all have their own limitations. But all three shift the focus away from winner-take-all logic and toward viable, context-sensitive decisions.

1. Consent: From Consensus to the Viable Objection

Consent originates from Sociocracy and fundamentally changes the guiding question of decision-making. It is no longer about whether everyone agrees, but whether there is a serious objection to a proposal. A resolution is considered adopted as long as no one can reasonably argue that the proposal seriously harms the common purpose or the system.

This may sound like splitting hairs, but the effect is significant. Consent relieves groups of the notion that they must find an ideal solution for everyone. Instead, the group searches for a solution that is good enough for now and safe to fail. Objections are not treated as a nuisance but as a resource for improving decisions.

A typical consent process runs through clear steps: introduce the proposal, clarify questions of understanding, collect initial reactions, examine objections, and integrate them if necessary. The quality of the objection is crucial. Not every personal preference constitutes a valid objection. An objection must demonstrate why the proposal would substantially jeopardize the shared work or shared goal.

A real-world example: A client had a very positive debate culture where participants were attentive to each other and genuinely tried to integrate different viewpoints. This frequently led to discussions going in circles, as finding true consensus proved difficult. The result was endless debate without resolution. An initial Consent session felt slow due to its five structured steps, and the structured suppression of open dialogue was a hard shift for participants. But the outcome was convincing: it was viable for everyone, immediately actionable, and reached within a single meeting. The team went on to use Consent regularly for its decisions.

Benefits

• High participation without requiring everyone to be enthusiastic

• Objections improve the quality of the proposal instead of just blocking it.

• Good balance between participation and capacity for action.

Limitations

• Requires practice in distinguishing opinion, preference, and serious objection.

• Can reach its limits in deep value conflicts

• Requires facilitation or at least strong process discipline

2. Systemic Consensus: Maximizing Acceptance Instead of Agreement

Systemic Consensus changes not just the question but the underlying logic of how options are compared. Instead of maximizing approval, it minimizes resistance. Participants don’t rate which option they like best, they rate how strong their resistance is to each option, typically on a scale of 0 to 10. This can be done with or without a “zero option”, which represents the status quo.

The option with the lowest total resistance is considered the most viable. This shifts attention from enthusiasm to acceptance and reduces the probability of strongly polarizing solutions being pushed through by a narrow majority. In heterogeneous teams, that’s a major advantage. If the zero option proves most viable, the situation simply remains unchanged.

A simple example from IT: Three technical options are on the table. Solution A excites part of the team but triggers massive resistance from others. Solution B excites almost no one but is considered acceptable by all. In a majority vote, A might win. In Systemic Consensus, B will likely come out ahead, because its collective viability is higher.

Benefits

• Makes resistance visible and evaluable rather than ignoring it.

• Protects minorities better than classic majority procedures.

• Well-suited for selection decisions among multiple options.

• Works well in larger groups and digital settings.

• Fast to run

Limitations

• Can tend toward moderate, less ambitious solutions.

• The quality depends heavily on honest ratings.

• Not ideal when options still need to be developed first (it represents the final step of decision-finding, not the exploratory phase).

3. Advice Process: Fast, Informed, and Binding

The Advice Process combines clear accountability with a broader information base. One person makes the decision. However, they are obliged to consult those people who are affected by the decision or can contribute relevant expertise beforehand.

The key point: the consulting person does not have to follow the advice. But they must seek it out, consider it seriously, and transparently explain their decision. This maintains commitment without falling back into the information poverty of the classic power call.

An example from everyday IT life: A developer wants to introduce a new monitoring tool. Before deciding, he consults the operations team, the architecture owners, and the leads of adjacent teams. The feedback does not necessarily change the decision, but it changes its quality. License costs, integration effort, maintainability, and training requirements become visible before facts are created.

Benefits

• High speed combined with a significantly better information base.

• Clear accountability instead of diffused group responsibility.

• Good fit for operational or domain-specific decisions.

• Promotes participation without collectively slowing down decisions.

Limitations

• Can degrade into a rubber-stamp process if consultation is purely formal

• Requires psychological safety so that advice is given honestly.

• Works less well for decisions that are meant to be collectively legitimized

Auxiliary Tool: The OwnershipSpicer by Dov Tsal

A common stumbling block in the Advice Process is the question: Who actually gets to decide here and who should be consulted? A helpful tool here is the OwnershipSpicer, a card set by Dov Tsal, which my colleague Steffen Oehme has translated into German as a Miro template.

Ownership Spicer Cards - translated by Steffen Oehme from the original developed by Dov Tsal

The OwnershipSpicer works similarly to Delegation Poker, but starts from a different angle. Delegation Poker asks from the perspective of the delegating person: How much decision-making power am I giving up? The OwnershipSpicer reverses this logic. Instead of a top-down framework, it creates a peer-level conversation: each person on the team describes the degree of ownership they feel (or can take on) for a specific topic from “I’ll fully delegate” to “I will decide alone”.

This may make the tool useful for the Advice Process. It helps teams clarify who actually has or should have the decision-making authority, who should be consulted as a knowledgeable advisor, and whether the Advice Process is even the right method. If this clarity is lacking, the Advice Process risks becoming an alibi process – formally consulted, substantively ignored.

Methods Comparison: Which Approach Fits When?

“What makes a decision great is not that it has a great outcome. A great decision is the result of a good process, and that process must include an attempt to accurately represent our own state of knowledge. That state of knowledge, in turn, is some variation of ‘I’m not sure.’”


Annie Duke, Thinking in Bets

The actual question is rarely: Which method is the best? More relevant is: which method fits this situation? Depending on time pressure, conflict potential, decision significance, and required participation, the requirements shift significantly.

A rough heuristic for selection: Consensus suits small groups and foundational questions with high legitimacy requirements. Majority voting works when options are clearly defined and consequences are manageable. Power calls have their place under time pressure and clear expertise. Consent is well-suited for ongoing team decisions. Systemic Consensus fits selection decisions among multiple options. The advice process suits operational questions with clear accountability.

Conclusion: The Question Behind the Question

The real question is rarely: which method is best? It is: which method fits here and now?

There is no universal process that resolves uncertainty, tames complexity, or forces human systems into orderly patterns. Consent, systemic consensus, and the advice process are no guarantees for correct decisions. They are invitations to take the decision-making process more seriously than the outcome and in doing so, to open the space where good decisions can actually emerge.

Those who choose tools in a situation-appropriate manner, understand objections as a resource instead of an obstacle, and clearly name responsibility without surrendering participation may not always decide correctly. But they decide in a way that makes learning from mistakes possible, brings stakeholders along, and holds up to scrutiny even in hindsight.

In a world where certainty usually only emerges in retrospect, that is no small thing.