// 01 — Overview
Trial Performance Summary
966 organisations trialled the platform Jan–Mar 2024. After removing 67,631 duplicate events, 102,895 clean events remain across 28 activity types.
Organisations
966
Jan–Mar 2024 cohort
Conversion Rate
21.3%
206 converted
Median Days to Convert
30.0
Mean: 30.4 days
Clean Events
102K
67,631 dupes removed
Dormant Orgs
81%
785 orgs ≤5 events
⚠ Critical Finding
Conversion is a day-30 pricing decision, not a behaviour signal. Converters and non-converters have identical median engagement. No feature predicts conversion above random chance (AUC ≈ 0.50–0.54).
Conversion Split
206 converted · 760 did not
Module Adoption Rates
% of orgs using each module
Cohort Conversion by Month
Jan–Mar 2024
Weekly Activity Retention
Orgs active each trial week
Top Activity Adoption
% of orgs performing each activity at least once
// 02 — Activation Goals
Trial Activation Framework
An organisation is Activated when it completes all 5 goals. These are evidence-informed hypotheses, not proven conversion levers.
Activation Rate
9.3%
90 of 966 orgs
Activated Conv. Rate
22.2%
vs 21.3% baseline
Activation Lift
1.04x
Marginal signal
Best Single Goal
22.7%
Goals 2 & 4 tied
Zero Goals, Converted
21.3%
357 orgs — same as base
Goal Definitions & Performance
Completion rates and conversion lift per goal
| # | Goal | Condition | Completion | Conv% if Done | Lift |
|---|
Goal Completion Funnel
Orgs completing each goal
Goals Completed Distribution
Conversion rate overlay per tier
📌 Framework Note
The 1.04x lift is statistically weak. The framework provides a structured baseline for A/B testing. One quarter of treatment-group data will validate or invalidate it.
// 03 — Driver Analysis
What Drives Conversion?
Three independent methods all converge: in-app behaviour alone cannot predict conversion in this dataset.
Logistic Regression
0.478
±0.039 · 5-fold CV
Below random baseline. No linear signal.
Random Forest
0.537
±0.058 · 5-fold CV
Marginal improvement. Shift.Created weakly predictive.
Gradient Boosting
0.538
±0.057 · 5-fold CV
Identical to RF. Signal is genuinely absent.
Permutation Feature Importance
Mean AUC decrease when shuffled (RF)
Conversion Lift by Activity
Chi-square — all p > 0.05
Weekly Event Pattern
Median events per org per trial week — use Conversion filter to isolate groups
⚠ Null Result — Honest Finding
Zero activities are statistically significant predictors of conversion. All models AUC ≈ 0.50–0.54. Median conversion day = trial expiry (day 30). 37% of orgs complete zero goals yet convert at exactly the baseline rate.
Next steps: Collect sales touchpoints, NPS, team size, and industry data.
Next steps: Collect sales touchpoints, NPS, team size, and industry data.
// 04 — Engagement Segments
4-Segment KMeans Clustering
Even High Engagement orgs convert at only 25% — barely above the 21.3% baseline. Use the Tier filter above to highlight a segment.
High Engagement
Orgs16
Conv Rate25.0%
Median Events1,564
Active Days30.5
Week 1 Events385
Medium Engagement
Orgs79
Conv Rate17.7%
Median Events442
Active Days23.0
Week 1 Events94
Low Engagement
Orgs86
Conv Rate25.6%
Median Events134
Active Days8.0
Week 1 Events44
Dormant
Orgs785
Conv Rate21.1%
Median Events5
Active Days1.0
Week 1 Events4
Segment Conversion Rates
vs 21.3% baseline
Segment Size Distribution
785 of 966 orgs are dormant (81%)
Monthly Cohort Breakdown
Use Cohort filter to highlight a month
Month
Orgs
Conv
Rate
%
Jan 2024
305
70
22.9%
Feb 2024
347
79
22.8%
Mar 2024
314
57
18.2%
// 05 — Recommendations
Product Team Action Plan
Six prioritised recommendations. Ordered by expected impact.
01
Collect Missing Signals
AUC ≈ 0.50 means behavioural data alone is blind. Instrument NPS at day 7 and track CRM touchpoints alongside events to find the real conversion levers.
02
Attack Week 1→2 Drop-off
924 orgs active in week 1, only 206 (22%) in week 2. Automated nudge at day 5–6 for orgs with <10 events is the highest-leverage onboarding intervention.
03
A/B Test the Activation Goals
Current lift is 1.04x — statistically weak. Assign 50% of new trialists to goal-completion prompts. One quarter validates or kills the framework.
04
Investigate the Dormant Majority
785 orgs (81%) with ≤5 events convert at baseline (21.1%). A day-3 in-app survey reveals whether they are evaluating, stuck, or never serious.
05
Monitor the March Cohort
March 2024 converted at 18.2% vs ~23% for Jan/Feb — a 4.6pp gap. Monitor next quarter to distinguish structural shift from noise.
06
Prioritise PunchClock & Templates
Goals 2 & 4 show 22.7% conversion among completers — the highest of any goal. Focus onboarding flows on shift approval and first punch-in.