← Back to Knowledge Graph

Constraint-Based Testing Protocol: One Test Per Week at the Biggest Drop-Off

The Framework

The Constraint-Based Testing Protocol from Alex Hormozi's $100M Leads provides the disciplined methodology for the "Better" step in the More Better New amplification sequence. Instead of testing random improvements across the entire system, you identify the single biggest drop-off in your funnel, run one test per week against that specific constraint, allow four attempts, and if none succeed, move to the next constraint. The protocol prevents the two most common optimization mistakes: testing too many things simultaneously (which makes results uninterpretable) and giving up on a constraint too quickly (which abandons solvable problems).

How It Works

Step 1: Map your funnel and identify the biggest drop-off. Every advertising system is a funnel: impressions → clicks → leads → engaged leads → sales conversations → customers. Measure the conversion rate at each stage. The stage with the largest percentage drop is your constraint — the bottleneck that limits the entire system's output.

For example: if 10,000 impressions produce 500 clicks (5% click rate) but 500 clicks produce only 10 leads (2% conversion), the click-to-lead stage is your constraint. Improving the impression-to-click rate (which is already reasonable) would produce less impact than improving the click-to-lead rate (which is the bottleneck).

Step 2: Design one test against the constraint. Not three tests. Not five. One. A single variable change that you hypothesize will improve the constraint. Change the landing page headline. Modify the lead magnet offer. Rewrite the CTA. Adjust the form length. One change, isolated, measurable.

Single-variable testing matters because multi-variable testing (changing three things simultaneously) makes it impossible to determine which change produced the result. If conversion improves after changing the headline, the image, and the CTA, you don't know which change worked — and you can't replicate the improvement elsewhere.

Step 3: Run the test for one week. Seven days provides enough volume for statistical significance in most advertising contexts (assuming the Rule of 100 is being followed). Shorter tests produce noisy data. Longer tests delay optimization. One week is the sweet spot.

Step 4: Evaluate. Did the test improve the constraint? If yes, implement the change permanently and run the next test against the same constraint (there may be more improvement available). If no, design a different test against the same constraint.

Step 5: After four failures, move to the next constraint. If four consecutive tests fail to improve the bottleneck, the constraint may be structural (not solvable through optimization) or the bottleneck may have shifted to a different stage. Identify the next-biggest drop-off and repeat the protocol there.

Why Four Attempts Before Moving On

Four is calibrated for two reasons. First, one or two failed tests don't provide enough evidence that the constraint is unsolvable — the first attempts might simply be bad hypotheses. Four attempts, each based on a different hypothesis, constitutes a reasonable exploration of the solution space. Second, more than four attempts on a single constraint produces diminishing returns — if none of four genuinely different approaches worked, the remaining hypotheses are increasingly unlikely to succeed.

The four-attempt limit also prevents the perfectionism trap: spending months optimizing a 2% conversion rate to 2.3% when a different constraint's improvement from 15% to 25% would produce 10x the impact. The protocol forces you to move your optimization energy to wherever the highest leverage exists.

Cross-Library Connections

Dib's continuous improvement cycle from Lean Marketing (measure → learn → improve → repeat) is the philosophical foundation. Hormozi adds structural discipline: which improvement to target (biggest drop-off), how many attempts per target (four), and when to move on (after four failures).

Hormozi's More Better New sequence positions this protocol within the broader scaling strategy: you've already exhausted More (volume is maximized). Now you're in Better (optimizing the constraint). The protocol tells you exactly how to execute the Better step before moving to New (adding channels).

The protocol parallels Fisher's interest exploration from Getting to Yes: explore one interest deeply before moving to the next. Fisher tests whether a creative option addresses the core interest; Hormozi tests whether a change addresses the core constraint. Both use focused, sequential exploration rather than scattered, parallel experimentation.

Implementation

  • Map your funnel stages and conversion rates this week. Identify the single biggest drop-off.
  • Design your first test against that constraint. One variable, one hypothesis, one week.
  • Run it and measure. Did the drop-off improve? If yes, implement and test again. If no, design a different test.
  • Track all tests in a simple log: Date, constraint targeted, hypothesis, result (improved/no change/worse). This log becomes your optimization history.
  • After four fails on one constraint, move to the next. Don't get emotionally attached to fixing a specific bottleneck.

    • 📚 From $100M Leads by Alex Hormozi — Get the book