- \n
- Unclear Decision Rights:<\/strong> Authority and workflows are distributed across functions, regions, and projects, blurring the line between input and approval authority. This creates governance vacuums and decision churning.<\/li>\n
- Stakeholder Bloat:<\/strong> Excessive involvement leads to watered-down outcomes and slow progress. Decision velocity and quality drops as group size exceeds about seven participants.<\/li>\n
- Chronic Revisiting:<\/strong> Decisions are reopened repeatedly, eroding trust and momentum.<\/li>\n
- Information Overload:<\/strong> Managers spend more time gathering data than making decisions, leading to analysis paralysis.<\/li>\n
- Cognitive Biases:<\/strong> Confirmation bias, sunk cost fallacy, anchoring, and overconfidence distort judgment.<\/li>\n
- Data Trust Gap:<\/strong> Reluctance to trust algorithmic recommendations slows adoption of decision intelligence tools.<\/li>\n<\/ul>\n
See more about the challenges of matrix management in our comprehensive guide to matrix management<\/a> or in our book Making the Matrix work<\/a>.<\/p>\n
How can leaders clarify decision rights in a matrix?<\/h2>\n
Why is clarity of decision rights so critical?<\/h3>\n
In our matrix management workshops<\/a>, a recurring issue is the lack of clarity about who should be involved in matrix decisions, the process to be followed and who gets to make the final decision.<\/p>\n
Because of the interconnected nature of the matrix potentially many people can be involved in some way in decisions, and everyone has an opinion. Without clarity on decision rights, too many people get involved, and the process becomes slow and bureaucratic.<\/p>\n
Organisations use processes like ARCI (or RACI) to help clarify decision rights, and adding a \u201cD\u201d for decision maker can further sharpen accountability. Unfortunately conducting a full racy analysis is extremely time consuming and bureaucratic and so it is rarely possible.<\/p>\n
What are the four main ways decisions are made in a matrix?<\/h3>\n
![\"\"](\"https:\/\/www.global-integration.com\/wp-content\/uploads\/2026\/02\/How-decisions-are-made-in-a-matrix-300x164.webp\")
<\/p>\n- \n
- In a healthy matrix, decision making should flow to those with the expertise, not just positional authority.<\/li>\n
- Without clarity, organizations often default to collective decision making, which is slower and more complex. This is particularly prevalent in consensus oriented corporate and national cultures or where people feel unable to have the tough conversations that tell people, \u201cYou are not the decision maker.\u201d<\/li>\n
- Sometimes consensus is needed, but often empowered individuals can act faster and better. Expert quote Kevan Hall, \u201cIt cannot be right that the average of the views of the one person who does know and the six who do not is the way the decisions are taken!\u201d<\/li>\n<\/ul>\n
Why does stakeholder bloat slow down decision making in a matrix?<\/h2>\n
What is stakeholder bloat and why does it matter?<\/h3>\n
As organizations emphasize inclusion, teamwork, democracy and risk mitigation, the number of people in decision meetings balloons. Beyond 5\u20137 participants, each additional person adds complexity but little insight. The result is \u201cdeath by committee,\u201d where decisions are watered down and progress stalls.<\/p>\n
Most research and practitioner advice converges on the idea that decision-making groups should be kept lean\u2014ideally under seven people\u2014to maximize quality, speed, and engagement.<\/p>\n
How can you reduce stakeholder bloat?<\/h3>\n
- \n
- Apply the Vroom-Yetton<\/a> model to determine who really needs to be involved.<\/li>\n
- Use Amazon\u2019s \u201cTwo-Pizza Rule\u201d: Teams should be small enough to be fed by two pizzas.<\/li>\n
- Limit decision teams to under seven participants and avoid unnecessary pre-meeting meetings.<\/li>\n<\/ul>\n
How can you prevent chronic revisiting of decisions in a matrix?<\/h2>\n
Why do decisions get revisited so often?<\/h3>\n
Chronic revisiting occurs when decisions are officially made but not psychologically or politically committed to. This erodes trust and creates \u201czombie decisions\u201d\u2014initiatives that shuffle forward without real support.<\/p>\n
It may also happen when people make a decision, they did not have the right to make, or where they have failed to consult with the right stakeholders in advance.<\/p>\n
What are practical steps to improve decision velocity?<\/h2>\n
- \n
- Establish a \u201cdisagree-and-commit\u201d culture: Debate before the decision, but unity after.<\/li>\n
- Use Amazon\u2019s One-Way vs. Two-Way Door framework to distinguish reversible from irreversible decisions.<\/li>\n<\/ul>\n
![\"\"](\"https:\/\/www.global-integration.com\/wp-content\/uploads\/2025\/12\/Amazon-decision-doors-300x164.webp\")
<\/p>\n- \n
- Keep a decision log and set \u201ctripwires\u201d for revisiting only when significant new information emerges. Review this decision log for any decisions that are revisited to see if there is underlying root cause.<\/li>\n<\/ul>\n
How Do You Cut Through Information Overload?<\/h2>\n
Why is information overload such a problem?<\/p>\n
In 2026, the challenge for middle managers has shifted from obtaining data to curating it. We can generate a 15-page report in seconds from AI on any topic we choose. The challenge now is to ask the right questions, and to make sense of the mass of information we get as an answer which may be of very valuable quality.<\/p>\n
Managers also often use data as a \u201cshield\u201d to justify a pre-decided path, rather than a \u201clamp\u201d to find a new one.<\/p>\n
What are the solutions to information overload?<\/h3>\n
- \n
- Design dashboards for clarity, not just volume.<\/li>\n
- Use the Pareto principle: Focus on the 20% of data that drives 80% of value.<\/li>\n
- Train managers in query design\u2014asking the right question before looking at the dashboard.<\/li>\n
- Train managers in dealing with too much collaboration in meetings<\/a> and decisions as part of our matrix management course<\/a>.<\/li>\n<\/ul>\n
How can teams mitigate cognitive biases in decision making?<\/h2>\n
Cognitive bias is a mental shortcut or pattern that causes people to think or make decisions in ways that are not always logical or accurate. These biases can lead us to misinterpret information, overlook facts, or make poor choices without realizing it.<\/p>\n
What are the most common cognitive biases?<\/h3>\n
- \n
- Confirmation Bias:<\/strong> Seeking data that supports pre-existing beliefs.<\/li>\n
- Sunk Cost Fallacy:<\/strong> Continuing to invest in a losing proposition because you have already spent a lot on it, even though future prospects do not justify further investment.<\/li>\n
- Anchoring:<\/strong> Relying too heavily on the first piece of information you see to set the baseline or direction of your thinking.<\/li>\n
- Overconfidence:<\/strong> Assuming timelines and synergies will be met without evidence.<\/li>\n<\/ul>\n
How can you address these?<\/h3>\n
- \n
- Use Pre-Mortem Analysis: Imagine failure before it happens to expose risks.<\/li>\n
- Encourage Red Teaming: Assign a group to challenge the prevailing consensus.<\/li>\n
- Train teams to recognize biases and encourage devil\u2019s advocacy in meetings.<\/li>\n<\/ul>\n
An additional challenge in cross-functional teams is that different functional cultures<\/a> drive different mindsets, language and ways of working.<\/p>\n
What is the human-data trust gap in decision making in the matrix and how do you bridge it?<\/h2>\n
Why are managers reluctant to trust AI in decision making?<\/h3>\n
Despite the availability of advanced AI tools, leaders often hesitate to cede agency to algorithms. \u201cBlack box\u201d anxiety means that even when AI generates a solution in seconds, teams spend weeks validating it, negating the speed advantage.<\/p>\n
How can you build trust in AI?<\/h3>\n
- \n
- Start with AI support tools for routine decisions.<\/li>\n
- Build trust through transparency\u2014explain how recommendations are generated.<\/li>\n
- Gradually delegate routine decisions to AI while maintaining human oversight.<\/li>\n<\/ul>\n
What are the most common barriers to high-velocity decision making in a matrix?<\/h2>\n
- Sunk Cost Fallacy:<\/strong> Continuing to invest in a losing proposition because you have already spent a lot on it, even though future prospects do not justify further investment.<\/li>\n
- Confirmation Bias:<\/strong> Seeking data that supports pre-existing beliefs.<\/li>\n
- Keep a decision log and set \u201ctripwires\u201d for revisiting only when significant new information emerges. Review this decision log for any decisions that are revisited to see if there is underlying root cause.<\/li>\n<\/ul>\n
- Apply the Vroom-Yetton<\/a> model to determine who really needs to be involved.<\/li>\n
- Stakeholder Bloat:<\/strong> Excessive involvement leads to watered-down outcomes and slow progress. Decision velocity and quality drops as group size exceeds about seven participants.<\/li>\n
![\"\"](\"https:\/\/www.global-integration.com\/wp-content\/uploads\/2026\/02\/High-velocity-decision-culture-300x164.webp\")
<\/p>\n