Packaged pricing tiers instead of per-server: +114% revenue
Commentary byLiftArchiveOur analysis, not theirsSwitching from per-server pricing to packaged tiers starting at US$99 a month raised revenue by 114%, from $394 to $833, even though free trial sign-ups fell by about 25%.
Server Density, a server monitoring SaaS, ran this test, and it was published as a VWO case study.
The revenue gain rests on only 20 vs 15 paid conversions.
- Revenue: up 114%, from US$394 to US$833.
- Average order value: up from US$19.70 to US$55.53.
- Free trial sign-ups: down 24.96%, from 6.92% to 5.19% (135 vs 100 sign-ups). This is a real cost.
- Paid upgrades: 20 on control vs 15 on the variant, from 1,950 and 1,925 visitors. Fewer people paid, but each paid more.
The change
The variant swaps per-unit pricing for fixed packages starting at $99.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
Configurable pricing where customers paid per server and per website monitored. It was built to attract first-time, single-server users, though most customers had 7 servers.
Variant
Packaged pricing tiers, with the lowest package starting at US$99 a month. Monitoring 10 servers and 10 websites dropped from US$130 to US$99 a month.
Why it likely won
Our reading of the result, not the practitioner's claim.
- The packages set a higher floor, so small single-server buyers spent more or left.
- Packages are easier to compare than a per-unit calculator.
- Most customers had several servers, so a bundle matched how they really used the product.
Caveats: Only 20 vs 15 paid conversions, so the revenue result could easily be noise. The revenue win comes from higher order value, not more buyers, and no duration or confidence level was shared.
Who ran it
Server Density ran this test and published the results (VWO case study). The numbers are theirs; the summary and commentary on this page are ours.
Read the original case study on Server Density →
How to run this test on your site
- Hypothesis
- If we replace per-unit pricing with packaged tiers, then revenue per visitor will rise because buyers choose a bundle that matches real usage at a higher floor price.
- Primary metric
- Revenue per visitor (or new MRR per visitor)
- Guardrails
- Free trial sign-up rate
- Paid conversion count
- Churn and downgrades in the first months
- Who should test it
- Usage-based SaaS where most customers sit well above the smallest unit. Run it long enough to collect many paid conversions. It may hurt if your growth depends on many small, low-cost customers.
Questions
Is packaged pricing better than usage-based pricing for SaaS?
In this test, packages doubled revenue but cut free sign-ups by a quarter.
It depends on whether your customers mostly sit above the smallest unit.
Why did revenue go up when fewer people signed up?
Each paying customer spent much more, with order value rising from about $20 to $56. That outweighed the drop in buyers.
How many conversions do I need for a pricing test?
Far more than 20 vs 15. With so few paid conversions, a couple of extra sales on either side would change the result.