Pricing page tier presentation test
Problem
As potential GitLab user, I do not understand the value provided at each tier. The pricing page references features of GitLab that I know nothing about and as a result, it is not clear what value GitLab provides with the features at each tier.
We discovered this issue when conducting customer interviews for the new SaaS paid signup flow. It was apparent that users could not quickly parse the value that GitLab provides in the paid tiers. As a result, Most users were either visiting the features page and being linked to documentation pages where they would get lost in technicality or simply signing up for free. Our conversion rate from the pricing page to starting the paid signup process is only 2.4% - 3.9% (SaaS, Self-Managed), we believe this conversion rate should be consistently 4.5% - 5% for both products (Several studies have shown best in class conversion rates for our industry and for freemium products to be around 5%. E.g. Dropbox).
Proposal
Given user feedback from 11 pricing page interviews (issue 1, issue 2, summary), we are proposing the changes attached in the following design. The major items changing are as follows:
- We have moved all of the bulleted items into a easy to comprehend "Jobs To Be Done" style framework, where we speak to the value that gitlab provides with its feature set rather than just a reference to the feature itself. For example:
| Old | New |
|---|---|
| Merge approvals | Reduce risk by requiring team leaders to approve merge requests |
| Multiple Group Issue Boards | Plan and organize parallel development with multiple issue boards |
-
Aligned with GitLab Value Framework, we pulled aspects of the feature comparison page directly into /pricing. The majority of users were moving off of the pricing page to the features page to get more detail. We believe that if they have the ability to quickly grok this feature list without leaving the page they will be significantly more likely to convert.
-
We moved the hosting choice to after the tier decision. If you look at our conversion rate on this page, self-hosted is ~33% lower. We believe this is because the user has to click into a different tab in order to see self-hosted plans. By splitting this off and adding it later in the process the user can make the best informed decision about their tier and then make the best informed decision about their hosting without having to discover how to view the different features for each hosting choice.
Experiment design
This page receives ~70k views per month with a drop off rate of 36%. The experiment will start at 25% of traffic and will need to run for ~ 1 month - 2 months to react significance.
| Old | New |
|---|---|
 |
 |
Specs Preview
Risks
- Less conversions to Free trial from /pricing
- 1.03% of traffic from /pricing converts to free trial signup
- Confusion from Self Hosted Re: Runner Minutes
- In testing, users who were interested in self hosting understood that they could configure their own runners, but a broader audience may need more clarification.
Expectations
At 5% conversion we can expect an additional 98 direct paid signups per quarter (at our current paid signup completion rate which is subject to change given this experiment). Will update with revenue impact when this issue is resolved.
What is next
Given the outcome of this research, we will be following this up with extensive research to discover the features that compel users to signup or upgrade to each tier and how we can best express those during the users introduction to GitLab.