Optimization Experiment Protocol.

Improve the system through structured tests: form a clear hypothesis, change one thing at a time, measure the impact, and keep what works, so the engine compounds toward reliability instead of changing at random.

The experiment card

Run one experiment at a time. Fill a card for each.

Hypothesis (changing ___ should improve ___)

The one change (exactly what you will change, and nothing else)

How measured (which scoreboard metric, and the baseline it starts from)

Test window (how long before you decide)

Decision rule (the result that means keep, versus the result that means discard)

The rules

One change at a time, or you will not know what worked. Measure against your scoreboard, over the defined window. Keep what beats the baseline; discard what does not. Where an experiment points upstream, fix it in the guide that owns it: a weak channel to Build Your Lead Generation Engine, a message that is not converting to Sharpen Your Value Message, leads going cold to Convert Interest into Pipeline, thin referrals to Activate Referrals.

The experiment log

Keep a running record, so improvement compounds and you do not repeat dead ends.

Date

Hypothesis

The one change

Result vs baseline

Keep or discard

Your answers stay in your browser on this page and are never sent anywhere.

Acts on decisions from the Lead Generation Review Cadence; measures against the Lead Generation Scoreboard; routes upstream fixes to the owning guides. Back to the guide, or start over.