Core Concept

Theory of Constraints

A methodology for identifying the most limiting factor (constraint) in a system and systematically improving it. Your system is only as strong as its weakest link.

⛓️
The Chain Analogy

A chain is only as strong as its weakest link. Strengthening any other link doesn't increase the chain's strength. The same applies to product workflows—improving non-constraints won't improve overall throughput.

What is Theory of Constraints?

Theory of Constraints (TOC) was developed by Dr. Eliyahu Goldratt in the 1980s. The core insight: every system has at least one constraint that limits its performance. To improve the system, you must find and fix that constraint.

Key TOC Principles

1
Systems thinking

The system's performance is limited by a small number of constraints

2
Focus

Improve the constraint, not everything else. Local optimizations often harm global throughput.

3
Continuous improvement

Once you fix one constraint, another appears. This is progress!

For Product Teams

Your constraint might be:

  • Engineering capacity: More ideas than engineers to build them
  • Design reviews: Everything waits for the senior designer
  • Research validation: Can't ship without user testing, but tests book out weeks
  • Deployment process: Manual QA takes 3 days before each release
  • Decision-making: Everything needs executive approval (meeting bottleneck)

The Five Focusing Steps

TOC provides a systematic process for managing constraints. Follow these steps in order:

1

IDENTIFY the Constraint

Find the bottleneck—the step where work piles up or waits the longest.

How to identify your constraint:
  • • Look for WIP (work in progress) piling up before a step
  • • Track where work waits the longest (use Value Stream Mapping)
  • • Ask: "What do we always complain is too slow?"
  • • Measure cycle time for each phase of your workflow
  • • Find the step with the lowest capacity relative to demand
Example: Research to Backlog Flow

Looking at the Research to Backlog flow, you notice:

  • • Interviews complete quickly (10 hours)
  • • PRD creation waits 7 days for PM availability ← CONSTRAINT
  • • Task breakdown happens fast once PRD is ready

Constraint identified: PM capacity to write PRDs

2

EXPLOIT the Constraint

Get maximum output from the constraint with existing resources. Don't add resources yet—just use what you have better.

Ways to exploit:
  • • Eliminate waste in the constraint step
  • • Ensure constraint never sits idle
  • • Remove low-value work from the constraint
  • • Give constraint the best tools and support
  • • Automate parts of the constrained process
Example: Exploiting PM PRD bottleneck
  • ✓ Use AI to draft PRD from research synthesis (saves PM 60% time)
  • ✓ Create PRD template so PM only fills in gaps
  • ✓ Have researcher pre-populate "User Insights" section
  • ✓ Block PM calendar for PRD-writing time (no meetings)
  • ✓ PM reviews AI draft instead of writing from scratch

Result: PRD writing time reduced from 6 hours → 2 hours. Wait time from 7 days → 2 days.

3

SUBORDINATE Everything Else

Align all other processes to support the constraint. The constraint sets the pace.

Counter-intuitive insight:

Don't maximize efficiency everywhere. That creates inventory (WIP) before the constraint.

  • • Non-constraints should run at lower than maximum capacity
  • • Feed the constraint just enough work to keep it busy
  • • Don't create work faster than constraint can consume it
Example: Subordinating to PM constraint
  • ✓ Don't conduct interviews faster than PM can write PRDs
  • ✓ Limit WIP: Max 2 research syntheses waiting for PRD
  • ✓ Engineering doesn't start breaking down tasks until PRD is ready
  • ✓ Research team helps PM by pre-drafting sections
  • ✓ Schedule work to arrive at constraint when constraint is available

Key insight: It's OK if researchers have idle time. Their "efficiency" doesn't matter—only system throughput matters.

4

ELEVATE the Constraint

If exploiting isn't enough, add capacity to the constraint. This usually costs money or time.

Ways to elevate:
  • • Hire more people for the constrained role
  • • Outsource work to increase capacity
  • • Buy tools that dramatically speed up the constraint
  • • Redesign the process to skip the constraint entirely
  • • Train others to handle constraint work
Example: Elevating PM constraint
  • 💰 Hire another PM (increases capacity 2x)
  • 🤖 Train AI to write 80% of PRD (PM only edits, doesn't write)
  • 📚 Train senior designer to write PRDs (distribute constraint work)
  • 🔄 Change process: Skip PRD for small features, go straight to tasks

After elevation: PRD creation is no longer the constraint. A new constraint emerges (maybe engineering capacity).

5

REPEAT: Don't Let Inertia Set In

Once you break one constraint, another appears. Go back to step 1. This is continuous improvement.

Warning: Policy constraints

After elevating, watch out for policy constraints—rules and habits from when the old constraint existed.

  • • "We batch PRDs because PM was the bottleneck" ← No longer true!
  • • "We can only do 2 research projects per quarter" ← Old constraint thinking
  • • Outdated processes designed around old constraints become new constraints

Applying TOC to Your Workflows

TOC + Research to Backlog

Apply the Five Focusing Steps to the Research to Backlog flow.

1.Identify: Where does work wait the longest?
2.Exploit: Use AI prompts to speed up constraint
3.Subordinate: Limit WIP before constraint
4.Elevate: Add capacity or change process
5.Repeat: Find the next constraint

Common Product Constraints

Discovery Phase

User research scheduling, participant recruitment

Planning Phase

PM capacity, stakeholder alignment, PRD reviews

Design Phase

Senior designer reviews, design system updates

Development Phase

Engineering capacity, code review, testing

Deployment Phase

QA cycles, approval processes, release windows

How saaslete Tools Help

Identify Constraints

Exploit Constraints

  • AI prompts in flows: Automate constraint work
  • Automation tools: Remove manual work from constraint
  • Templates & checklists: Standardize constraint processes

TOC vs Other Improvement Methods

MethodFocusBest For
Theory of ConstraintsFind & fix the one biggest bottleneckMaximizing throughput quickly
Value Stream MappingVisualize entire flow, eliminate all wasteUnderstanding current state, identifying multiple improvements
Agile/ScrumIterative delivery, team collaborationOrganizing team workflow and ceremonies
KanbanLimit WIP, visualize flowManaging steady flow of work

💡 Pro Tip: Combine Them

Use Value Stream Mapping to visualize your workflow, then apply TOC to focus on the biggest constraint. Use Kanban/Agile for day-to-day execution.