---
id: "action-ab-test-defaults"
type: "action-item"
source_timestamps: ["§ What Executives Should Do"]
tags: ["experimentation", "long-term-planning"]
related: ["claim-auto-cancel-yields-more-subs"]
action: "Randomize offers between auto-renew and auto-cancel, tracking cohort performance for a minimum of 12 months."
outcome: "Accurate measurement of the true long-term impact of renewal defaults on total paid subscriber volume."
sources: ["commercial"]
sourceVaultSlug: "hbr-seg-commercial"
originDay: 5
articleStem: "hbr-tier2-08-subscription-auto-renew"
sourceUrl: "https://hbr.org/2026/05/should-your-subscription-business-use-auto-renew"
sourceTitle: "Should Your Subscription Business Use Auto-Renew?"
---
# A/B test renewal defaults over a 12+ month horizon

**Action:** Randomize subscription offers between **auto-renewal and auto-cancel**, and track the cohorts for **at least 12 months**.

Short-term tests are misleading because auto-renewal's early retention advantage (**20–38%**) masks long-term acquisition losses that only materialize over longer windows — the crossover that produces [[claim-auto-cancel-yields-more-subs|23% more paid subscribers under auto-cancel]] appeared only over a 20+ month horizon. Requires [[prereq-cohort-analysis|cohort analysis / LTV]] competence.

**Outcome:** Accurate measurement of the true long-term impact of the [[concept-renewal-default]] on total paid subscriber volume.