$10 vs $12.50 vs $14.95 monthly price: +33% revenue at $10
Test run byMarketingExperimentsResearch publisher
Commentary byLiftArchiveOur analysis, not theirsCutting a subscription site's monthly price to $10 brought in 33% more revenue than the next best price, $12.50, and more than double the orders of $14.95. MarketingExperiments (MECLABS) ran this four-day A/B/C price test for a paid-subscription website and published the numbers in a research brief.
The only thing that changed between the three landing pages was the price.
- $10/month: 156 orders and $1,560 in revenue, the best of the three on both counts.
- $12.50/month: 94 orders and $1,175 in revenue. The $10 price earned about 33% more revenue than this.
- $14.95/month: 74 orders and $1,106.30 in revenue. The $10 price had about 2.1 times the orders and about 41% more revenue.
- With equal ad spend of $808.02 per page, the reported return on ad spend was 93% at $10, 45% at $12.50 and 37% at $14.95.
- These are first-month figures. The brief's longer-term revenue numbers are projections, not measured results.
The change
Same page, three monthly prices: $14.95, $12.50 and $10.00.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The landing page offered the subscription at $14.95 per month, the highest of the three prices tested.
Variant
Two identical landing pages showed $12.50 and $10.00 per month. Paid search traffic was split evenly across all three pages.
Why it likely won
Our reading of the result, not the practitioner's claim.
- The extra orders at $10 more than made up for the lower price per order, so volume won the revenue race.
- Paid search visitors often compare options quickly, so a lower monthly number may have cleared their mental price limit.
- For a subscription, each extra subscriber can pay again next month, which makes order volume worth more than it looks on day one.
Caveats: This was a four-day micro-test with 324 orders in total and no significance figures shared. The data dates from 2005, and the brief's lifetime revenue figures are modelled rather than measured.
Who ran it
MarketingExperiments ran this test for a paid-subscription website and published the results (MarketingExperiments). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on MarketingExperiments →
How to run this test on your site
- Hypothesis
- If we lower our monthly subscription price, then total first-month revenue will rise because the gain in sign-ups will outweigh the lower price per subscriber.
- Primary metric
- Revenue per visitor (orders x price, divided by visitors) for each price point
- Guardrails
- Churn or cancellation rate by price cohort
- Refund rate
- Customer acquisition cost per paid subscriber
- Who should test it
- Subscription and SaaS businesses with steady paid traffic and enough volume to read results quickly. It fits less well when you sell to a few large accounts or when price is set by sales negotiation.
Questions
Does a lower subscription price always earn more?
No. In the same brief, a separate book price test found the highest price earned the most revenue.
You need to test your own price points.
How should I judge a price test for a subscription?
Look at revenue per visitor first, then follow each price cohort for a few months.
Retention can differ by price, and that changes lifetime value.
How many price points should I test at once?
Three is a common choice because it shows the shape of demand.
Each extra arm needs more traffic to reach a clear answer.