Michal Parizek
Senior Growth Product Manager at Mojo
Where to hear and read about Michal Parizek1 episode or test
Michal Parizek is Senior Growth Product Manager at Mojo, a mobile-first app for animated social media content, where paywall and pricing tests lifted ARPU 60% in five months. The most useful lesson is to run many fast tests judged on 7-day revenue per user and retest every win in each market.
There's more shots you take, there's essentially higher chance you win. There's a shorter sort of feedback loop, better learning cycle.
How Mojo Increased ARPU 60% In Just Five Months – Michal Parizek, Mojo
Michal Parizek: the story and the lessons
1Michal Parizek and Mojo's 60% ARPU gain
Parizek helped raise Mojo's ARPU 60% in five months. The gain came from a handful of paywall and pricing experiments, not one big change. Three stood out: a yearly default, monthly price anchoring and lower prices in some regions.
Most lifts are quoted from memory as rough figures, with no sample sizes or confidence levels, and the 60% is not broken down by test.
He treats new-user revenue as the lever that feeds acquisition. Quick iteration on new-user revenue compounds, and it supports higher ad spend. New revenue was easier to scale than renewals.
2The paywall and pricing changes Parizek ran at Mojo
Parizek made yearly the default and hid the monthly plan. Yearly and monthly first sat side by side on the first screen. Moving monthly behind a "view all plans" button raised yearly adoption by roughly 15 to 20 percentage points.
The extra yearly buyers raised new revenue directly.
A small monthly-equivalent line beside the yearly price lifted revenue. Mojo showed the yearly plan as about $10 a month next to a monthly plan of about $25. The lift was about 10% in the US and 30 to 40% in Latin America, notably Brazil and Mexico.
Parizek's explanation for the bigger Latin America lift is lower purchasing power, but he calls it "just my hypothesis", not a tested result.
He tested lower prices instead of using exchange-rate conversions. Mojo first used App Store style prices converted from US prices. Lower prices in Brazil and Mexico, its top two Latin American geos, performed better.
"We obviously tested a bit lower prices and it turned out it was actually better."
Parizek found that a winning paywall design in one market can fail in another. A long scrolling paywall with reviews and a free versus pro comparison gave about a 20% lift in Japan but failed in the US. In the US, a slider, background videos and punchy copy won.
Every winner was retested in each key region before rollout. After a test won with statistical significance, for example in Europe, Mojo reran it in other key markets. It shipped only where the change added incremental revenue, so some changes went to only some markets.
This regional retesting is how the wins stacked into the 60% gain.
3How Parizek measures tests and protects long-term revenue
Mojo judged every monetization test on 7-day ARPU. Parizek chose revenue per user in the first seven days as the umbrella metric. It covers the 3-day trial, and 7-day was the first ARPU view RevenueCat showed.
Early ARPU shortens payback, which lets acquisition spend grow. New-cohort ARPU 7D was tracked in RevenueCat Charts to confirm lifts.
He used 7-day cancellation rate as a stand-in for renewals. On price tests, higher prices usually brought higher cancellation and lower conversion. Mojo modeled a year of revenue using the 7-day cancellation proxy.
The proxy only estimates what renewal rate could look like. It is not real renewal data.
Parizek kept the original price when a higher one won early. In a couple of tests a higher price won on new revenue, but the year-long model showed lower renewals. Mojo took less new revenue now for more over time.
"We would actually sacrifice in a long term."
He reads a lower day-one payer share as healthy. About half of Mojo's payers convert on day one, against roughly 80% across the industry in RevenueCat data. Parizek says it signals strong retention and many existing users who can pay later.
It was his first mobile-first job, so he was surprised by day-one revenue and learned only later that 80% is normal.
4Parizek on existing users and experiment velocity
Mojo shows free users a paywall on app open once a week. The campaign drove about 15% of new revenue from existing users. Support and a thorough review analysis showed no complaints.
The paywall can also trigger after a key action such as sharing. Parizek calls it "kind of a no-brainer to have that" and says few apps run it.
He kept the free tier generous on purpose. Free users get many features and lots of content, less restricted than some competitors. Parizek links this to why users tolerate the paywall and why revenue keeps arriving after day one.
A weak premium tier keeps few users, so few ever see an app-open paywall.
Parizek says to A/B test aggressive ideas rather than trust your fears. He expects users to react differently from what teams fear. The app-open paywall drew no noticeable backlash despite expectations of it.
"You will be surprised that the users will not react the way how you kind of offer fear they react."
Mojo ran parallel test streams across three geo buckets. New users were split into US and English-speaking markets, Europe and Latin America. Each bucket ran tests weekly or biweekly, lasting one to two weeks, and reached significance in that time.
A third-party paywall platform cut Parizek's test cycle to days. It let the growth PM launch tests without waiting on engineering, shrinking idea to live test from months or weeks to days. Parizek says he could not have run so many experiments otherwise.
He calls velocity the number one growth asset and says it may matter even more than good research and prioritization.
How Mojo Increased ARPU 60% In Just Five Months – Michal Parizek, Mojo
Mojo raised average revenue per user 60% in five months through paywall and pricing experiments, tested by region and measured on 7-day revenue per user. The episode shows how a weekly paywall for free users, regional retesting and fast experiment cycles each added revenue.
Parizek helped raise Mojo's ARPU 60% in five months. The gain came from a handful of paywall and pricing experiments, not one big change. Three stood out: a yearly default, monthly price anchoring and lower prices in some regions.
Most lifts are quoted from memory as rough figures, with no sample sizes or confidence levels, and the 60% is not broken down by test.
He treats new-user revenue as the lever that feeds acquisition. Quick iteration on new-user revenue compounds, and it supports higher ad spend. New revenue was easier to scale than renewals.
Parizek made yearly the default and hid the monthly plan. Yearly and monthly first sat side by side on the first screen. Moving monthly behind a "view all plans" button raised yearly adoption by roughly 15 to 20 percentage points.
The extra yearly buyers raised new revenue directly.
A small monthly-equivalent line beside the yearly price lifted revenue. Mojo showed the yearly plan as about $10 a month next to a monthly plan of about $25. The lift was about 10% in the US and 30 to 40% in Latin America, notably Brazil and Mexico.
Parizek's explanation for the bigger Latin America lift is lower purchasing power, but he calls it "just my hypothesis", not a tested result.
He tested lower prices instead of using exchange-rate conversions. Mojo first used App Store style prices converted from US prices. Lower prices in Brazil and Mexico, its top two Latin American geos, performed better.
"We obviously tested a bit lower prices and it turned out it was actually better."
Parizek found that a winning paywall design in one market can fail in another. A long scrolling paywall with reviews and a free versus pro comparison gave about a 20% lift in Japan but failed in the US. In the US, a slider, background videos and punchy copy won.
Every winner was retested in each key region before rollout. After a test won with statistical significance, for example in Europe, Mojo reran it in other key markets. It shipped only where the change added incremental revenue, so some changes went to only some markets.
This regional retesting is how the wins stacked into the 60% gain.
Mojo judged every monetization test on 7-day ARPU. Parizek chose revenue per user in the first seven days as the umbrella metric. It covers the 3-day trial, and 7-day was the first ARPU view RevenueCat showed.
Early ARPU shortens payback, which lets acquisition spend grow. New-cohort ARPU 7D was tracked in RevenueCat Charts to confirm lifts.
He used 7-day cancellation rate as a stand-in for renewals. On price tests, higher prices usually brought higher cancellation and lower conversion. Mojo modeled a year of revenue using the 7-day cancellation proxy.
The proxy only estimates what renewal rate could look like. It is not real renewal data.
Parizek kept the original price when a higher one won early. In a couple of tests a higher price won on new revenue, but the year-long model showed lower renewals. Mojo took less new revenue now for more over time.
"We would actually sacrifice in a long term."
He reads a lower day-one payer share as healthy. About half of Mojo's payers convert on day one, against roughly 80% across the industry in RevenueCat data. Parizek says it signals strong retention and many existing users who can pay later.
It was his first mobile-first job, so he was surprised by day-one revenue and learned only later that 80% is normal.
Mojo shows free users a paywall on app open once a week. The campaign drove about 15% of new revenue from existing users. Support and a thorough review analysis showed no complaints.
The paywall can also trigger after a key action such as sharing. Parizek calls it "kind of a no-brainer to have that" and says few apps run it.
He kept the free tier generous on purpose. Free users get many features and lots of content, less restricted than some competitors. Parizek links this to why users tolerate the paywall and why revenue keeps arriving after day one.
A weak premium tier keeps few users, so few ever see an app-open paywall.
Parizek says to A/B test aggressive ideas rather than trust your fears. He expects users to react differently from what teams fear. The app-open paywall drew no noticeable backlash despite expectations of it.
"You will be surprised that the users will not react the way how you kind of offer fear they react."
Mojo ran parallel test streams across three geo buckets. New users were split into US and English-speaking markets, Europe and Latin America. Each bucket ran tests weekly or biweekly, lasting one to two weeks, and reached significance in that time.
A third-party paywall platform cut Parizek's test cycle to days. It let the growth PM launch tests without waiting on engineering, shrinking idea to live test from months or weeks to days. Parizek says he could not have run so many experiments otherwise.
He calls velocity the number one growth asset and says it may matter even more than good research and prioritization.