Keeping “11 AM ET” in the challenge headline produced the higher observed opt-in rate: 35.36% versus 32.33% for the date-only version across 857 users. The reported 82.15% win probability is directional, not conclusive, and the prerecorded format makes accurate expectation-setting more important than manufacturing urgency.
What the experiment compared
The page promoted a five-day challenge. Control included a calendar date and “11 AM ET”; treatment kept the same date but removed the time. A feature flag changed the headline, and a welcome-page view served as the opt-in proxy.
The article historically shows different example calendar dates in its introduction and test-design section. That inconsistency prevents this revision from claiming which date appeared in production. The defensible comparison is the presence versus absence of the time, assuming the original experiment configuration confirms the rest of the headline was identical.
Observed result and limits
| Measure | Exact-time control | Date-only treatment |
|---|---|---|
| Users | Part of 857 total | Part of 857 total |
| Reported conversion rate | 35.36% | 32.33% |
| Reported win probability | 82.15% | 17.85% |
The rounded rates differ by 3.03 percentage points. The article reports the treatment as 8.59% lower relative to control. Per-arm assignments and conversions are not available in the current page, so this is a transcription of the historical result rather than a reconstructed analysis.
A welcome-page view supports a conclusion about reaching that page, not necessarily a valid registration, attendance, completion, qualified lead or purchase. A replication should verify that reloads and direct visits cannot inflate the proxy and should connect assignment to the intended downstream outcomes.
Do not turn a directional result into a psychological claim
The observed result is consistent with the time making the offer feel more concrete, but the test did not measure clarity, commitment, uncertainty or decision fatigue. Those are possible explanations, not findings. With only an 82.15% reported probability and unrecovered arm counts, the result should not be generalized into a rule that exact times always convert better.
The expectation problem
Participants could watch the prerecorded webinar at other times. If the headline implies that attendance is required at 11 AM ET, the page can create a false constraint even if that phrasing increases initial opt-ins. Conversion rate does not override the obligation to describe the experience accurately.
A clearer follow-up would compare truthful options such as “Starts at 11 AM ET—replay available” against a date-only or on-demand message. Keep the underlying availability identical, label the timezone, and measure registration, actual viewing time, completion and downstream conversion.
Decision for this page
Treat the exact-time result as a promising historical observation, not a final prescription. Before reusing it, recover the exact two headlines, arm counts, exposure rule and conversion definition. For a new campaign, choose wording that matches the real delivery model, then test the honest alternatives.
For measurement design, see the A/B testing metrics guide. For implementation, use PostHog’s current experiment documentation and verify the resulting exposure and conversion events rather than assuming the UI alone proves the setup.
