Experiment Backlog & Tracker

warningWhy You Need This

Building features based on gut feel is a gamble. Most features fail because you learn they are wrong after weeks or months of work.

This playbook lets you test ideas in days, measure results objectively, and double down only on what works.

checklistHow to Use This System

1. counter_1

   Capture Ideas in the Backlog: Every idea starts with a hypothesis.

2. counter_2

   Prioritize with ICS: Rank by Impact, Confidence, and Speed.

3. counter_3

   Design Experiments: Define metrics and the minimum viable test before building.

4. counter_4

   Run Experiments Weekly: Ship small tests every week, not big bets every quarter.

5. counter_5

   Log Results: Capture learnings and update your backlog based on evidence.

timelinePlaybook Steps (In Order)

1. counter_1

   Step 1: Experiment Types (Guide) - Choose the right test before building.

2. counter_2

   Step 2: Experiment Design (Template) - Define hypothesis, metrics, MVT, and decision criteria.

3. counter_3

   Step 3: Experiment Backlog & ICS Prioritization (Template) - Capture and score every idea, including the ICS guide.

4. counter_4

   Step 4: Experiment Results Tracker (Template) - Log outcomes and learnings.

5. counter_5

   Step 5: Weekly Experiment Rhythm (Guide) - Plan Monday, review Friday, repeat.

categoryExperiment Types & Examples

Choose the right test before you commit engineering time.

lab_profileExperiment Design Template

Define hypothesis, success metrics, MVT, and decision criteria before you build.

dashboardExperiment Backlog & ICS Prioritization

Use this backlog template to capture every idea, then score it with ICS to prioritize experiments fast.

insightsICS Scoring Guide

Impact (1-10): How much will this move your key metric?

• chevron_right

  10 = Could 10x growth

• chevron_right

  7 = Meaningful improvement (20-50% gain)

• chevron_right

  5 = Modest improvement (10-20% gain)

• chevron_right

  3 = Minor improvement (< 10% gain)

• chevron_right

  1 = Negligible impact

Confidence (1-10): How sure are you this will work?

• chevron_right

  10 = Strong evidence (similar tests, competitor data, clear user demand)

• chevron_right

  7 = Moderate evidence (some data, educated guess)

• chevron_right

  5 = Weak evidence (theory-based, anecdotal feedback)

• chevron_right

  3 = Very uncertain (complete guess)

• chevron_right

  1 = Just a hunch

Speed (1-10): How quickly can you test this?

• chevron_right

  10 = < 1 day (copy change, simple A/B test)

• chevron_right

  7 = 2-3 days (mockup, fake door test)

• chevron_right

  5 = 1 week (basic functional version)

• chevron_right

  3 = 2-3 weeks (some dev work)

• chevron_right

  1 = 4+ weeks (major undertaking)

receipt_longExperiment Results Tracker

Log every experiment to build institutional knowledge.

scheduleWeekly Experiment Rhythm

Build a habit of shipping experiments every week.

history_eduExample: How StartupFlow Used This System

Company: StartupFlow (5-person project management tool for startups)

Problem: Roadmap bloat and months wasted on low-value features.

Old Approach: Gantt charts (4% adoption), Slack integration (12%), mobile app (8%). Total wasted effort: 6+ months.

New Approach:

• chevron_right

  Week 1: Fake door test for Templates, 42% click rate, build it.

• chevron_right

  Week 2: Prototype to 10% of users, 67% usage, ship.

• chevron_right

  Week 3: Messaging test, "Start Projects Faster" lifted usage by 28%.

• chevron_right

  Week 4: Pricing experiment, $220 MRR from 4 hours of work.

Results After 12 Weeks: 8 experiments shipped, 3 major wins, 5 killed ideas, +$1,200 MRR.

Top Learning: Stop asking "What should we build?" and start asking "What is the fastest way to test this?"