3 Upgrade Triggers from Free to Paid Plans

Dear Readers,

Hitting a limit while using a free account triggers customers to sign up for a paid plan to unblock a product or feature. 

However, too many limits and paywalls alienate users and drive churn. 

I spent the last year as a senior growth designer at Parabol running experiments to discover what was the right limit to trigger the most upgrades while driving the least churn.

After testing three limits—SURPRISE!—we did NOT choose the limit with the most conversion power. Read on to find out why…

Yours without limits,

Alicia Cressall, Senior Growth Designer

Parabol is a Series A startup that’s an agile meeting tool for remote teams.

How the sausage is made

Parabol works in a squad model, so everything discussed here is the result of work spread across a team of developers, a designer, a product manager, and a data scientist.

Our PM introduced us to the cake model of experiments, and throughout these experiments, we worked to refine our process of starting small and learning quickly. 

When the squad took on the goal of improving self-serve conversion we knew very little about the users we wanted to convert. We spent time talking to users and creating user personas for that tier. We took two personas and created user journeys for each, looking especially at pain points and opportunities. We brainstormed “How might we…” statements based on the journey maps and voted on where we wanted to focus.

The Experiments

Once we had our area of focus we brainstormed ideas on how to answer our HMW question and scored them on effort, size, and risk to prioritize. We implemented three experiments toward our goal. 

1. Limit the number of free templates

Paid Team Accounts are slightly more likely to use custom templates. Users who are just starting to use Parabol often don’t use many public templates and might slowly be teased by the Team tier when they see inaccessible premium templates. We will restrict custom templates and limit the number of public templates in the free tier to incentivize users to upgrade.

2. Limit the number of free templates

Paid Team Accounts are slightly more likely to use custom templates. Users who are just starting to use Parabol often don’t use many public templates and might slowly be teased by the Team tier when they see inaccessible premium templates. We will restrict custom templates and limit the number of public templates in the free tier to incentivize users to upgrade.

3. Limit the number of teams

We have limits on the number of teams specified in the plans, but currently the application does not prohibit going beyond the limits in any way, does not notify about it, and does not impose any restrictions afterwards. We will implement an application-level limit on the number of teams, and we believe that this will incentivize users to upgrade themselves.

The Results

Each of the experiments was successful in converting accounts to paid tiers, but an important metric for us to watch was churn. Many users see a limit and decide to leave if the feature isn’t valuable enough. The chart below shows which features triggered an upgrade.

🥉THIRD PLACE: Team Instances

Team limits were a strong lever, though a lagging indicator. Teams needed to expand to the point of adding a third team, at which point they would receive a 30-day warning to upgrade before being locked out. Because it was such a hard limit this was the most susceptible to churn, so we decided to roll it back until the top of funnel matured. 

🥈 SECOND PLACE: TemplatesIt’s very clear that template limits were a strong lever, but as this was being tested we noticed our top of funnel was slowing down. Were template limits preventing stickiness and expansion? We’re retesting a less restricted version now to find out!

🥇THE WINNER: Meeting History

Meeting history was a solid lever and had little-to-no impact on churn, and as such was the most successful experiment.

Conclusion

These experiments were crucial in teaching us that we need to look for balance in growth.

We can cannibalize one metric in service of another and it’s important to keep an eye on unintended impact.

These rate-limit experiments improved self-serve revenue by 2x in 10 months with minimal churn.

About the Author

Alicia has spent nine years designing products, specializing in systems that put users first. The last three years have had a focus on growth, marrying user needs with business value. When not designing, you’ll find her deep in a book, going for walks with her dog Opie, and traveling.