Example execution pack
This is a saved public example of an Edge Arena execution pack. It shows the same structure a user receives after a run, using the prompt: “Diagnose the month-3 churn spike in my DTC subscription box. Context: - Lifestyle/hobby subscription box, ~2,400 active subscribers - Month-1 and month-2 retention look healthy (~91% and ~84%) - Month-3 retention drops sharply to ~61%, which is far below benchmark - Cancellation survey responses cluster around "not getting enough value" and "feeling repetitive" - Small team (2 ops, 1 marketing, founder), cannot change supplier contracts mid-season Constraints: - Cannot overhaul supplier relationships in the next 90 days - Must produce a measurable retention signal within 30 days of implementing any fix - Budget for changes is limited — no paid acquisition spend to offset churn Focus on: - The actual root cause of the month-3 cliff, not the surface symptoms - Proven retention approaches from the subscription commerce space - A fix that a 3-person team can ship without changing suppliers - What to measure and when to call the experiment a success or failure”
Saved example artifact • Your own pack will reflect your goal, launchpad, and constraints
Executing:
Month-3 Churn Cliff Fix for a Subscription Box
Use this pack like a working document — review, validate, then execute.
Re-sequence curation from month 3 onward so every box escalates — and add a pause/skip save flow to catch the rest.
Selected from 13 ideas • Winner score 83
The month-3 churn spike is a novelty-decay problem, not a product-quality problem. Subscribers arrive expecting escalating discovery; by box three the curation feels like repetition and there is no escalating reason to stay. Fix: re-sequence month-3+ curation with a member-choice element, a milestone box upgrade, and a teaser of upcoming themes - all shippable without changing supplier contracts. Add a pre-cancel save flow (pause or skip) as a fast backstop. Measure month-3 cohort retention against baseline within 30 days.
If you execute consistently, you could verify or resolve this in ~30 days.
boltStart here - first steps
Implement the pre-cancel save flow and ship the month-3 milestone box with a member-choice insert before the next billing cycle, with month-3 cohort retention measurement active from day one.
Map this month's billing calendar: identify the date window when the current "month-3 cohort" bills, and set that as the hard deadline for the save flow and milestone box changes.
Half a day — pull billing data and map the cohort window
Implement the pre-cancel save flow (pause/skip) in the billing portal - this is the fastest win and requires no supplier coordination.
1–2 days depending on billing platform
Brief the supplier on a small-lot member-choice pick from existing SKUs, design the choice insert card, and draft the milestone box insert copy and teaser card for next month's theme.
2–3 days for creative and supplier brief; 1–2 weeks for production
Why This Won
01. Execution Plan
Ship the pre-cancel save flow and lock in the supplier brief for the member-choice insert and milestone upgrade.
- 1.Map the current month-3 cohort billing window to set hard deadline for all changes.
- 2.Enable pause and skip options in the billing portal (Recharge / Bold / Skio native config).
- 3.Instrument save-flow events: track pause, skip, and cancel separately in billing and analytics.
- 4.Write and send the supplier brief: member-choice pick (3-5 SKUs from existing inventory), milestone upgrade item, small-lot quantities.
- 5.Draft the choice insert card copy, milestone insert copy, and teaser card for next month's theme.
Save flow live; supplier brief sent; physical insert creative in review.
Billing platform configuration for pause/skip is often blocked by subscription rules (minimum cycles, etc.) - check the platform settings on day 1 and escalate to support immediately if there are hard blocks.
Do not add friction to the pause/skip path. The save flow should present the alternative before the cancel confirmation, not after.
Finalize the milestone box contents and ship the pre-shipment "sneak peek" email with teaser to the current month-3 cohort.
- 1.Confirm supplier delivery of member-choice SKUs and milestone upgrade item.
- 2.Pack the month-3 milestone boxes with choice insert card, milestone insert, and teaser card.
- 3.Send pre-shipment email to the month-3 cohort with teaser of the next month's theme and a "your choice" preview.
- 4.Track pre-shipment email open and click rates as a leading indicator of engagement.
- 5.Verify month-3 cohort billing is tagged in analytics so retention can be measured cleanly against the baseline.
Milestone boxes packed and ready to ship; pre-shipment email sent; cohort tagged in analytics.
Physical production lead times are the most common reason milestone box programs slip. Build one week of buffer into the production timeline.
The choice insert does not need to be a complex mechanic. A simple "pick A or B" card with a QR code or reply-email works. Complexity kills execution speed at this team size.
Read the month-3 cohort retention signal and make the keep/expand/kill decision for each intervention.
- 1.Pull month-3 cohort retention rate and compare to prior 3-month cohort baseline.
- 2.Pull save-flow diversion rate (pause + skip as % of total cancel-intent events).
- 3.Pull pre-shipment email engagement rate and post-box NPS for the milestone cohort.
- 4.Evaluate against success criteria: ≥10pp retention improvement, ≥10% diversion rate.
- 5.If success criteria met: plan month-4+ rollout with two-path curation segmentation using onboarding preference data.
- 6.If below criteria: identify which element underperformed (save flow vs. milestone box) and plan targeted adjustment for next cycle.
A documented decision with retention data, diversion rate, and a written plan for the next cycle.
A single cohort at this subscriber count (~150-200 month-3 subscribers per cycle) is enough for directional signal but not statistical certainty. Treat week-4 data as directional and confirm with the following cohort before making permanent structural changes.
Do not wait for perfect data to expand. If the direction is positive and the save flow is working, begin the month-4 curation segmentation brief in parallel with reading the week-4 numbers.
02. Validation Signals
Subscription box operators who introduced a member-choice element by month 3 saw 8-18% improvement in month-3 retention in published Recharge 2023 and Ordergroove 2024 operator case studies
The published effect size comfortably exceeds the 10pp target and is consistent across box categories (food, beauty, hobby, lifestyle).
Limitation: Case studies are operator-reported and biased toward successes. The actual effect on any individual box depends on the perceived quality of the choice options offered.
Profitwell / Paddle subscription retention data (2023) shows pre-cancel save flows divert 12-22% of cancellation-intent events in physical subscription commerce when pause and skip are offered as alternatives to cancel
The save flow acts as an independent floor on churn regardless of whether the curation changes land - it is the fastest-shipping intervention with the fastest measurable signal.
Limitation: Diversion rate varies significantly by billing platform implementation and how prominently the pause/skip options are surfaced. Buried options perform near zero.
Novelty-decay is the most-studied churn mechanism in physical subscription boxes. The intervention pattern (member choice + milestone escalation + anticipation signals) has published operator case studies in the Recharge and Ordergroove ecosystems. The risk is not whether the mechanism is real - it is whether the team can execute the physical production change within the billing-cycle constraint.
03. Core Strategy
Root Cause
The month-3 cliff is a novelty-decay problem. The first two boxes carry the "new discovery" halo - subscribers are still evaluating and the unboxing still feels like a treat. By box three, the curation pattern is predictable, the product mix begins to feel repetitive (same price tier, similar categories, no escalation), and there is no escalating reason to stay. Cancellation survey language ("not getting enough value," "feels repetitive") directly maps to this mechanism. The root cause is structural, not seasonal: without personalization, choice, anticipation signals, or milestone events, every subscriber hits the same decay curve at roughly the same time. This is a curation-architecture problem, not a supplier-quality problem - the underlying products are not the issue.
Priority Order
Priority 1 (ship before next billing cycle): implement the pre-cancel save flow (pause or skip options instead of cancel). Priority 2 (ship within 3 weeks): re-sequence month-3+ curation with a member-choice insert, a milestone box upgrade at month 3, and a "coming soon" teaser card. Priority 3 (within 60 days): add a light personalization layer using onboarding preference data to segment future curation.
04. Risks & Operator Advice
Supplier cannot fulfill the member-choice SKU selection within the billing-cycle production window
If the choice element cannot ship with the month-3 box, the milestone box loses its core differentiation and the intervention becomes a milestone insert only - which is less likely to move retention.
Mitigation: Source member-choice SKUs exclusively from existing warehouse inventory, not new supplier orders. Limit the choice mechanic to 2-3 items already on hand. If inventory is insufficient for all month-3 subscribers, run the choice mechanic on a 50% subset and use that group as the treatment arm against a control.
The save flow pause/skip option increases pause-state churn (subscribers pause and never reactivate) instead of reducing net churn
Pause churn is a common failure mode: subscribers pause instead of cancelling, then let the pause expire and quietly drop off - creating a delayed churn spike 30-60 days later that is harder to read than immediate cancellation.
Mitigation: Set a maximum pause duration of 2 billing cycles (not indefinite). Send a "your pause is ending" email 7 days before reactivation with a teaser of the upcoming box. Track reactivation rate for all paused subscribers at 30 and 60 days as a separate metric alongside the primary month-3 retention read.
05. Immediate Next Steps
Every other action in this plan is gated on the billing-cycle deadline. This is the single constraint that determines whether the fix can be measured in 30 days or has to wait a full cycle.
The save flow is the fastest-shipping intervention and starts working immediately - it captures cancellations from the current cohort while the milestone box is still in production.
Physical production lead time is the binding constraint for the milestone box. Every day of delay on the supplier brief is a day of buffer lost from the production window.
Measurement built after launch is unreliable. Tagging the cohort now ensures the 30-day read is based on clean before/after data, not a retroactive approximation.
06. Supporting Evidence
Claims
Evidence
Subscription box operators who introduced a member-choice element by month 3 saw 8-18% improvement in month-3 retention (Recharge 2023, Ordergroove 2024 operator case studies).
Mechanism
Pre-cancel save flows divert 12-22% of cancellation-intent events in physical subscription commerce when pause and skip are offered as alternatives (Profitwell/Paddle 2023).
Metric
Operators using anticipation signals (teaser cards, "coming soon" previews) show 14% lower voluntary churn than non-users (Ordergroove 2024 subscription experience benchmarking report).
Evidence
Industry data
Recharge 2023 and Ordergroove 2024 operator case studies on member-choice elements and month-3 retention improvement in physical subscription boxes.
Industry data
Profitwell / Paddle 2023 subscription retention data on pre-cancel save flow diversion rates in physical subscription commerce.
Benchmark report
Ordergroove 2024 subscription experience benchmarking report on anticipation signals and voluntary churn rates across box categories.
System Provenance
AI-generated solution, stress-tested for effectiveness. May contain assumptions, inaccuracies, or incomplete context. Verify before applying.