Preview: Decision Science Pre-Analysis Framework (WIP)

Below is a draft outline for each section of the Decision Science Pre-Analysis Framework that I will be covering in the coming weeks. All subject to change, but nevertheless a starting point and deeper introduction.

1. Aligning with Business Priorities
   • What It Does: Ensures the analysis is answering an urgent, high-value business question.
   • Why It Matters: If the insight doesn’t align with stakeholder priorities, it won’t be acted upon.
   • Actions Before the Analysis Starts:
     • Engage key stakeholders early – Ask: What decisions are you struggling with right now?
     • Ensure the insight solves a business need – Link it to revenue growth, cost reduction, risk mitigation, or strategic advantage.
     • Predefine what success looks like – Ask stakeholders: If this analysis is successful, what action would you take?
   • Example:
     • Instead of: let’s analyze customer churn trends.
     • Reframe as: let’s find the top 3 leading indicators of churn so we can act before customers leave.
2. Pre-Framing the Decision Context
   • What It Does: Establishes how insights will be used before data is presented.
   • Why It Matters: Stakeholders are more likely to act on insights if they already expect to make a decision.
   • Actions Before the Analysis Starts:
     • Define the “Last Mile” First – What decision will be made based on this insight?
     • Use Default Framing for Action – Instead of “Should we act?” ask “What would stop us from acting on this insight?”
     • Set the Expectation That Change Will Happen — If the data shows X, we will do Y.
   • Example:
     • Instead of: We’ll analyze marketing ROI and report back.
     • Say: We’ll identify underperforming channels and recommend reallocation strategies. Are you open to shifting budget based on our findings?
3. Designing Stakeholder Buy-In & Ownership
   • What It Does: Makes stakeholders feel invested in the insights before they are delivered.
   • Why It Matters: People are more likely to act on ideas they feel they co-created.
   • Actions Before the Analysis Starts:
     • Involve stakeholders in defining the metrics & methodology. Ask: What would make this data more credible for you?
     • Assign a stakeholder “sponsor” who is responsible for actioning the insight.
     • Run a “Pre-Mortem” Session: Ask: If this insight is ignored, why would that happen? Solve objections upfront.
   • Example:
     • Instead of: We’ll send the report when it’s ready.
     • Say: We’re building the analysis with your input. Would you like to review early insights to help shape recommendations?
4. Structuring Data For Cognitive Ease
   • What It Does: Reduces decision fatigue by presenting insights in a clear, intuitive way.
   • Why It Matters: If insights are too complex or overwhelming, they won’t be used.
   • Actions Before the Analysis Starts:
     • Use “Less but Better” Data Presentation -> Show only the critical insights that drive decisions.
     • Pre-Test Visualization Preferences -> Ask stakeholders: Would you prefer a trendline, ranking, or heatmap?
     • Align Insights with How Stakeholders Already Think – If they trust anecdotes, pair insights with real examples.
   • Example:
     • Instead of: A 20-slide deck full of complex charts.
     • Deliver: A 1-page visual with “Here’s What’s happening -> Here’s what to do next.”
5. Embedding Insights into Decision Workflows
   • What It Does: Ensures insights are used in real business processes, not just in reports.
   • Why It Matters: If the insights don’t naturally fit into how stakeholders work, they’ll be ignored.
   • Actions Before the Analysis Starts:
     • Tie insights to existing decision points — When do they decide? Ensure the insight is ready before then.
     • Use Automation & Nudges — If an insight is critical, set up alerts or reminders to reinforce it.
     • Require an Action Plan in Advance — Before the analysis starts, ask: “If the data shows X, what changes are we willing to make?
   • Example:
     • Instead of: Delivering a static dashboard for sales forecasts.
     • Integrate: Sales probability scores directly into CRM workflows so reps see insights as they work.
6. Setting-up Accountability & Reinforcement Mechanisms
   • What It Does: Encourages follow-through by making stakeholders accountable for acting on insights.
   • Why It Matters: Without accountability, even the best insights can be ignored.
   • Actions Before the Analysis Starts:
     • Tie insights to performance KPIs — If adoption of insights is measured, it’s more likely to happen.
     • Follow-up Meetings for Decisions — Schedule a post-analysis meeting with the agenda: What actions will we take?
     • Create Visibility Around Implementation — Recognize teams that successfully act on insights.
   • Example:
     • Instead of: Here’s your churn analysis.
     • Say: We’ll meet in two weeks to review how churn risk scores are being used in retention efforts.

1. Choice Architecture ↩︎
2. From Meaningful Data Science to Impactful Decisions: The Importance of Being Causally Prescriptive ↩︎

Pages: 1 2