The win came from re-defining activation. The CTA move was the consequence, not the cause.
Email integration was hidden behind two clicks. 0.3% of new users connected. The easy fix was a more discoverable CTA. We didn't take it.

Before · email connect lived in-product: a popup two clicks deep, reached from the go-to-market checklist rail.
Team lead. Owned the activation definition across the surface, the call to put a new step into onboarding, and the staged rollout that proved it was safe.
The bar for "activated" was too low. Move it to "email sent" instead of "email connected." Senders were +294% more likely to visit pricing.
"Email sent" became the milestone, not "connected." Optimizing for the action that earned revenue, not a vanity step.
PLG default says keep onboarding short. We made it longer on purpose — because the right activation moment was downstream of connection.
Tested at 50% to prove no hit to completion rate. Then to 100%. Discipline over hope.
Email integration was the gate to phase 2. We restructured what onboarding was for.

After · the new onboarding step: "Set up your email in a minute." The value spelled out in five lines, one-click connect, and a skip for later. This step drove 96% of new connections.
Aug at 50%: 0.7% → 12.6%, no hit to onboarding completion. Sep at 100%: 11× growth (0.3% → 3.3%), 96% of new connections driven by onboarding.

The tie-in · contact profile before engagement: the activity feed sits empty.

After engagement · the same profile with email metrics living in the product: scheduled, sent, opened, replied.
Optimized for "email connected" for too long before realizing it was a vanity step. Before you set an activation milestone, make sure it actually correlates with the downstream metric you care about.