04 — Design & Run Experiments
Purpose: Use Swift CNS to design and execute experiments to test your hypotheses
Outcome: Have experiments designed and running in Swift CNS to validate hypotheses
Audience: PM / Dev / Both
Time: 1-2 weeks per experiment (ongoing)
Prerequisites: 03 — Form Testable Hypotheses - Hypotheses created in Swift CNS
Learning Outcomes
By the end of this chapter, you will be able to:
Use Swift CNS to design experiments that test specific hypotheses
Create experiments manually or from chat conversations
Track experiments and their status in Swift CNS
Run experiments efficiently with minimal resources
Monitor experiment progress in Swift CNS
Jobs-to-Be-Done
When: I have testable hypotheses ready in Swift CNS
I want: To design and run experiments to test them
So that: I can validate or invalidate assumptions with data
Inputs
Testable hypotheses from 03 — Form Testable Hypotheses in Swift CNS
Success criteria for each hypothesis
Access to users or test environment
Time and resources for experimentation
Activities
1. Design Experiments in Swift CNS
Two Ways to Create Experiments:
Option A: From Chat Conversation (Recommended)
Continue your chat conversation in Swift CNS
The AI will guide you through experiment design
Follow the AI's prompts to create experiment plans
The AI can help you create experiments directly from chat
Option B: Manual Creation
Navigate to your project in Swift CNS
Click "Create Experiment" button
Fill out the experiment form manually

💡 Tip: Let the AI guide you first. It helps ensure experiments are well-designed.
2. Create Experiment in Swift CNS
Using the Create Experiment Form:
Click "Create Experiment" on your project page
Fill out the experiment form:
Basic Information:
Experiment Name: Descriptive name (e.g., "Landing Page Interest Test")
Theme: Desirability, Feasibility, Viability (choose one)
Hypothesis: Paste your hypothesis statement
Description: Detailed description of how the experiment will be conducted
Experiment Type: e.g., Landing Page Test, Survey, Concierge Test
Success Criteria:
Primary Success Criteria: Main measurable outcome
Key Metrics: Add metrics (name, target, purpose)
Example: Conversion Rate, 30%, Measures user interest
Operational Details:
Setup Complexity (1-5): How complex is setup?
Run Time (1-5): How long does it take?
Cost (1-5): How expensive is it?
Data Reliability (1-5): How reliable is the data?
Method Details:
Test Method: e.g., Survey, Landing Page
Sample Size: e.g., 100 users
Duration: e.g., 2 weeks
Resources:
Time Required: e.g., 2 hours per day
Budget: e.g., $500
Tools: e.g., Google Forms, Typeform
Team Members: e.g., Product Manager, Designer
Risks & Mitigation:
Risk: Describe the risk
Mitigation: How to mitigate this risk
3. Set Up and Run Experiments
After Creating Experiment:
Review your experiment plan in Swift CNS
Set up the experiment (create landing page, survey, etc.)
Launch the experiment
Drive traffic (if needed)
Collect data
In Swift CNS:
Experiments are tracked on the Experiments tab
You can see status: Draft, Approved, Running, Completed, Archive
Monitor progress and results
4. Track Experiments in Swift CNS
View All Experiments:
Navigate to Experiments tab in your project
Or go to global Experiments page (from main nav)
Experiment Status:
Draft: Not yet started
Approved: Ready to run
Running: Currently in progress
Completed: Finished, results analyzed
Archive: Archived for reference

5. Monitor Experiment Progress
In Swift CNS:
View experiment details
Track metrics and results
Update status as you progress
Document observations
Key Metrics to Track:
Conversion rates
Signup rates
Usage metrics
Time spent
Completion rates
Apply It Now
Task: Design and create your first experiment in Swift CNS
Continue your chat conversation OR click "Create Experiment"
Design your experiment (use AI guidance or manual form)
Fill out all experiment details (hypothesis, success criteria, method, resources)
Create the experiment in Swift CNS
Set up the experiment (create landing page, survey, etc.)
Launch and run the experiment
Track progress in Swift CNS
Artifact: An experiment in Swift CNS with:
Experiment design
Success criteria
Method details
Resources identified
Status: Running
Artifacts
You'll create in Swift CNS:
Experiment plans
Experiment records
Status tracking
Results documentation
Worked Example
Situation: Creating experiment for retrospective tool in Swift CNS
Steps in Swift CNS:
Click "Create Experiment" on project page
Fill Out Form:
Experiment Name: "Landing Page Interest Test"
Theme: Desirability
Hypothesis: "We believe teams will use an AI-powered retrospective tool. If we create a landing page with mockups, then at least 30% of visitors will sign up for early access. We'll know this is true when we see 30%+ conversion rate after 100 visitors."
Experiment Type: Landing Page Test
Test Method: Landing page with mockups
Sample Size: 100 visitors
Duration: 1 week
Primary Success Criteria: 30% conversion rate
Key Metrics: Conversion Rate, 30%, Measures user interest
Click "Create Experiment"
Set Up Experiment: Create landing page, add tracking
Launch: Drive traffic, collect data
Track in Swift CNS: Monitor results, update status
Result in Swift CNS:
Experiment created and visible in Experiments tab
Status: Running
Metrics tracked
Results documented
Checklist
Before proceeding to the next chapter, verify:
Self-Assessment
How can you create experiments in Swift CNS? (Select all)
What should you include in experiment design? (Select all)
What can you track in Swift CNS? (Select all)
Exit Criteria
You're ready to proceed when:
Dependencies & Next Steps
Prerequisites Completed
03 — Form Testable Hypotheses - Testable hypotheses in Swift CNS
Next Steps
Continue running experiments and collecting data
Proceed to 05 — Extract Key Learnings to analyze results in Swift CNS
Use Swift CNS Learning Cards to document learnings
What This Enables
Running experiments in Swift CNS enables:
Data-driven validation
Tracked progress and results
Systematic learning
Clear go/no-go decisions
💡 Tip: Use the AI guidance first. It helps ensure experiments are well-designed. 📝 Note: You can create multiple experiments. Start with the highest priority hypothesis.
Last updated