Mojo

Making yearly the default by hiding monthly raised yearly adoption sharply. Mojo moved the monthly plan from beside the yearly plan on the first paywall screen to behind a "view all plans" button. Yearly plan adoption rose by roughly 15 to 20 percentage points.

The extra yearly buyers raised new revenue directly. It was one of three experiments Michal Parizek, Senior Growth Product Manager, singles out from a series of mostly paywall and pricing tests.

Sub Club▶ 2:06

A monthly-equivalent line made the yearly plan look like a bargain. Mojo added a small line beside the yearly price saying it equals about $10 a month, next to a monthly plan at about $25. Revenue rose about 10% in the US and 30 to 40% in Latin America, notably Brazil and Mexico.

Mojo had also considered raising the monthly price to flatter the yearly plan, but the anchoring line was what won. Parizek's explanation for the bigger Latin American lift, that lower purchasing power makes buyers weigh subscription costs more, is in his words "just my hypothesis".

Sub Club▶ 3:11

Local price points beat exchange-rate conversions of US prices. Mojo first used App Store style exchange-rate conversions of US prices, then tested lower prices in Brazil and Mexico, its top two Latin American geos. The lower prices performed better.

Quote: "we obviously tested a bit lower prices and it turned out it was actually better." The lifts across all three changes are quoted as rough figures from memory, and the 60% total is not broken down by test.

Sub Club▶ 4:59

One early-revenue metric, 7-day ARPU, judged every monetization test. Mojo used average revenue per user in the first seven days as the umbrella metric for all paywall and pricing experiments. Early revenue shortens payback, which lets ad spend grow, and new revenue was easier to scale than renewals.

Seven days was long enough to cover Mojo's 3-day trial and was the first ARPU view RevenueCat offered. After each rollout Mojo tracked new-cohort 7-day ARPU in RevenueCat Charts and tried, as a best effort, to separate its changes from user acquisition shifts.

Sub Club▶ 5:54

Higher prices that won early were dropped when modeled renewals fell. For price tests, Mojo used the 7-day cancellation rate as a proxy for renewals and modeled a year of revenue. In a couple of tests a higher price won on new revenue but showed far higher cancellation, so Mojo kept the original price.

Higher prices usually brought higher cancellation and lower conversion. The proxy stands in for real renewal data, which Mojo did not use. Quote: "we would actually sacrifice in a long term".

Sub Club▶ 7:47

Japan preferred a long, information-heavy paywall that failed in the US. A long scrolling paywall full of reviews, social proof and a free versus pro comparison lifted revenue about 20% in Japan. In the US the same design failed, and a clean design with a slider, background videos and punchy copy won.

Quote: "the same design kind of failed in the US".

Sub Club▶ 11:21

Every winner was retested in each key market before rollout. A test that reached statistical significance in one market, such as Europe, was rerun in the other key markets and shipped only where it added incremental revenue. Some changes shipped in some markets and not others.

This regional retesting is how Mojo stacked individual wins into the 60% ARPU increase.

Sub Club▶ 10:05

Mojo's video paywalls became a template other apps copy. Jonathan Parra of Superwall built a winning paywall inspired by Mojo's heavy use of video: a phone-in-hand Rotato video of the app in action, paired with social proof and simple bullets.

He swaps components such as the video, a slider or text, and targets the copy to the demographic. Quote: "you get like a clear value prop by seeing what the app does."

Superwall▶ 44:00

A once-a-week paywall on app open drove 15% of new revenue. Mojo showed existing free users a paywall on app open, capped at once per week. It drove about 15% of new revenue from the existing user base, with no complaints found in support or in a thorough review analysis.

Such campaigns can also trigger after a key action such as sharing. Parizek says few apps run them and calls it "kind of a no-brainer"; his lesson is to A/B test ideas like this rather than trust the fear of backlash, since users "will not react the way how you kind of offer fear they react".

Sub Club▶ 13:47

A generous free tier keeps users around to pay later. Mojo's free tier deliberately offers many features and lots of content, more than some competitors. Only about half of its payers convert on day one, against about 80% across the industry in RevenueCat data.

Parizek reads the lower day-one share as a "very healthy" signal of strong retention and a large base of existing users who can still be monetized. The generosity likely also explains why users tolerated the repeated paywall.

Sub Club▶ 13:43

A third-party paywall platform cut test cycles from weeks to days. Using a third-party paywall tool, the growth PM could build and launch tests without waiting on engineering, shrinking the cycle from months or weeks to days.

Parizek calls velocity the number one growth asset: "There's more shots you take, there's essentially higher chance you win." He says he could not have run that many experiments in five months without it.

Sub Club▶ 18:25

Three geo buckets ran parallel test streams every one to two weeks. New users were split into US, Australia and Canada; Europe; and Latin America, each running its own stream of tests. Each test lasted one or two weeks and reached statistical significance in that time.

Sub Club▶ 19:09