Share
Explore

Getting Started with Decision Intelligence: Practical Exercises for Leaders

Simple steps to begin your journey from data noise to decision intelligence
by , CEO of
megaphone
View or download the presentation here:

1. Decision-Centric Exercise: The 3-3-3 Method ​
image.png

Time needed: 30 minutes with your leadership team
Exercise:
List the 3 most frequent decisions your team makes weekly
List the 3 highest-impact decisions that affect your business outcomes
For each decision, rate the current decision quality (1-10)
Self-Assessment Questions:
Which decisions appear on both the "frequent" and "high-impact" lists?
For which decisions do you currently lack sufficient information?
Which decision, if improved by 20%, would create the most value?
Next Step: Choose ONE decision to focus your AI efforts on. This becomes your "Decision Anchor Point."

2. Data Connection Exercise: The Signal Mapping Workshop

image.png
Time needed: 45 minutes with cross-functional team members
Exercise:
Take your Decision Anchor Point from the previous exercise
Draw a simple table with these columns:
Data signals needed for this decision
Where this data currently lives
Who currently accesses this data
How difficult it is to access (1-5 scale)
Identify the top 3 data connection gaps
Self-Assessment Questions:
Which critical data signals are currently siloed or inaccessible?
Are there any data signals completely missing?
Which data connections would most improve this specific decision?
Next Step: Identify ONE data connection to establish first. Create a simple plan to connect these signals.

3. Operational AI Exercise: The Day 2 Reality Check

image.png
Time needed: 30 minutes with technical and business stakeholders
Exercise:
For your Decision Anchor Point, answer:
Who will be responsible for data quality?
Who will validate that outputs make sense?
Who will collect feedback on effectiveness?
How will we measure success?
Place each responsibility in one of three categories:
Build (internal capability)
Partner (external help needed)
Not Yet (future consideration)
Self-Assessment Questions:
Do we have clear ownership for each operational component?
Which operational aspects pose the biggest risks?
Are we being realistic about our internal capabilities?
Next Step: Create a simple RACI chart (Responsible, Accountable, Consulted, Informed) for your first decision intelligence initiative.

4. Context Building Exercise: Asset & Knowledge Mapping Sprint

image.png
Time needed: 90 minutes with domain experts, functional leaders, product/service owners

Asset Criticality Mapping:
List your top 5 business assets (systems, data, capabilities)
Rate each on a 1-5 scale for business criticality
Identify direct connections to your products/services
Map which teams/processes depend on each asset
Decision Knowledge Capture:
For your Decision Anchor Point, document:
3 key business rules that experts apply
3 signals that indicate "this needs attention"
3 factors that determine priority or severity
Create a simple relationship diagram showing how assets connect to business outcomes
Self-Assessment Questions:
Do we have a shared understanding of which assets matter most?
How do our critical assets relate to our value proposition?
What tribal knowledge about assets exists only in people's heads?
How consistently is this knowledge applied across teams?
Next Step: Create a simple one-page "Asset to Value" map that shows how critical assets connect to business outcomes.

Getting Started Today

Schedule these exercises over the next two weeks
Start small with one specific decision
Involve the right people from different functions
Document everything simply but clearly
Focus on progress, not perfection
Remember: The path to decision intelligence doesn't start with algorithms—it starts with clarity about which decisions matter most to your business.
From Data Noise to Decision Intelligence: A framework developed by Karen Kim, CEO of Human Managed
Want to print your doc?
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.