How Many Pricing Plans Should You Show? Test Idea

Support questions about which plan to buy, session recordings of visitors moving back and forth between plans, or a rarely chosen plan whose features largely overlap with another are good reasons to try a slimmer plan set. When simplification helps, visitors who see the smaller set settle on a plan more easily, and revenue per visitor reflects it. If the hesitation comes from prices or unclear plan contents, fix those first, and pick the plan to remove for its overlap, not to leave the plan you most want to sell as the only path.

Revenue per visitor (RPV)

Once both plan sets are defined, compare revenue per eligible assigned visitor over the same period.

Also track

• Average plan valueUse plan value to explain revenue; a lower average does not necessarily mean lower revenue per visitor.
• Plan selection rateCompare the share of visitors who select any plan.
• Cancellation or downgrade rateCompare cancellations and downgrades after the same follow-up period.
• Plan-selection support requestsTrack questions about choosing a plan using the same visitor denominator.

• Plan setsCompare the current set with a version that removes one redundant tier, selected before launch based on overlapping features and the prior 90 days of plan selections.
• Keep constantKeep prices, plan features and card design unchanged.
• RevenueDecide using revenue per visitor, not plan selection alone.
• Selection mixReview how plan choices shift to explain the result.
• AudienceReview customer groups separately without changing assignment rules.

1. 1A higher plan selection rate is not enough to ship the simplified set; revenue per visitor makes the call. A lower average plan value can still come with steady revenue, and more selections can still come with less of it, so read them together and compare cancellation or downgrade rates once both arms reach the end of the follow-up period.
2. 2The result describes these two plan sets, not plan count in general. Keep existing subscribers of the removed plan out of the test, review customer groups separately, and treat removing a different plan as a new test.
3. 3Revenue per visitor swings more than a selection rate, so a small gap may be noise at your traffic level. Before a rerun, get the sample per variant for the current plan selection rate and your target lift from the sample size calculator, then count the days of traffic per arm it takes; revenue, being noisier, needs even more. Also note how plan-selection support requests moved.