Increasing conversion
in Foxford's elementary education segment
Increasing conversion
in Foxford's elementary education segment
INDUSTRY:
EDTECH · KIDS
YEAR:
2022
EXPERIENCE:
PRODUCT DESIGN
Increasing conversion
in Foxford's elementary education segment
INDUSTRY:
EDTECH · KIDS
YEAR:
2022
EXPERIENCE:
PRODUCT DESIGN
01. about
01. about
01. about
Foxford is an online school for preschoolers and students, offering both supplementary education and a full alternative to traditional schooling. It is one of the top 3 largest EdTech companies in Russia’s children’s segment, with over 7.5 million users and more than 2.5 million unique monthly visitors.
Foxford is an online school for preschoolers and students, offering both supplementary education and a full alternative to traditional schooling. It is one of the top 3 largest EdTech companies in Russia’s children’s segment, with over 7.5 million users and more than 2.5 million unique monthly visitors.
[ 01 ]
[ 01 ]
My role
I led product design for Foxford's elementary segment end-to-end: research and journey mapping, redesigning the activation funnel, and building the gamification system that grew into a core part of the product.
[ 02 ]
[ 02 ]
Team & timeline
Partnered with a PM, an analyst, and 5 engineers over two seasons, 2022 and 2023. Results were measured against the previous year's cohort.
02. Goals
02. Goals
02. Goals
The primary education segment had extremely low conversion to purchase – most users registered, then dropped off without ever reaching a paid course.
The main goals of the segment are to increase conversion to activation, disclosure to purchase, increase the marginality of the product line, and save ARPPU users for the year
The primary education segment had extremely low conversion to purchase – most users registered, then dropped off without ever reaching a paid course.
The main goals of the segment are to increase conversion to activation, disclosure to purchase, increase the marginality of the product line, and save ARPPU users for the year
Increase conversion
from registration to paying user
Increase engagement
keep kids coming back
Increase conversion
from registration to paying user
Increase engagement
keep kids coming back
03. Discovery
03. Discovery
03. Discovery
I started by mapping the full learning experience and building a CJM for all three segments, to see where they really differ. I looked at how students move through online learning, from registration and choosing a course to accessing materials, working with tutors, and finishing tasks. From interviews and feedback, I pulled out the issues they run into along the way.
I started by mapping the full learning experience and building a CJM for all three segments, to see where they really differ. I looked at how students move through online learning, from registration and choosing a course to accessing materials, working with tutors, and finishing tasks. From interviews and feedback, I pulled out the issues they run into along the way.
04. Research Insights:
Defining User Needs and Challenges
04. Research
Insights: Defining
User Needs
and Challenges
04. Research Insights:
Defining User Needs
and Challenges
I began by analyzing the available data on the primary education segment. In addition to this, I conducted several interviews to gain a deeper understanding of user pain points, their current solutions, and emotional challenges related to children's education. These efforts provided us with a more complete picture of our audience's needs. Here are a few interesting insights we uncovered:
I began by analyzing the available data on the primary education segment. In addition to this, I conducted several interviews to gain a deeper understanding of user pain points, their current solutions, and emotional challenges related to children's education. These efforts provided us with a more complete picture of our audience's needs. Here are a few interesting insights we uncovered:
[ 01 ]
[ 01 ]
Parents distract children with gadgets but try to compensate for the lack of attention with educational activities
[02]
[ 02]
Children mainly lose interest in studying due to boring and monotonous materials
[03]
[ 03 ]
A parent’s (mostly a mother’s) desire is to have free time without feeling guilty
[04]
[ 03 ]
About 60% of parents sometimes do their children’s homework for them
[05]
[ 03 ]
The main disadvantage of online education is the lack of socialization
05. Where users dropped off
05. Where users
dropped off
05. Where users dropped off
These behaviors showed up directly in the numbers. The catalog has a huge list of courses split by goal, grade, and format, so after the landing activity people land in it, get lost, and quietly leave. And without a clear pain in elementary school, no exam to pass, no grades to fix, they don't understand what they need or why to pay.
Initially, the funnel looked like this:
These behaviors showed up directly in the numbers. The catalog has a huge list of courses split by goal, grade, and format, so after the landing activity people land in it, get lost, and quietly leave. And without a clear pain in elementary school, no exam to pass, no grades to fix, they don't understand what they need or why to pay.
Initially, the funnel looked like this:
06. The value loop
06. The value loop
06. The value loop
In this segment the parent pays, but the decision runs through the child: if the child isn't into it, they won't buy. So a direct sell doesn't work here. We took the other route – create real value for the child, and let them carry it back to the parent. That turns into a loop:
In this segment the parent pays, but the decision runs through the child: if the child isn't into it, they won't buy. So a direct sell doesn't work here. We took the other route – create real value for the child, and let them carry it back to the parent. That turns into a loop:
07. Ideation: free courses to engage users
07. Ideation:
free courses
to engage users
07. Ideation: free courses
to engage users
Hypothesis: lead some users to the home page instead of the catalog, and fill it with free courses from our flagship products for kids.
Before the season, we added that block of free courses. It shows that we already have an educational game and school-prep products available for free, and it lets us read what users are actually interested in. The home page started as just an empty state, so I reworked it and added the free courses. From inside a free course, we then guided users toward a paid one:
Hypothesis: lead some users to the home page instead of the catalog, and fill it with free courses from our flagship products for kids.
Before the season, we added that block of free courses. It shows that we already have an educational game and school-prep products available for free, and it lets us read what users are actually interested in. The home page started as just an empty state, so I reworked it and added the free courses. From inside a free course, we then guided users toward a paid one:
08. Outcomes
08. Outcomes
Tested only in the primary segment, the free-course flow lifted conversion to purchase 7x year over year. In the test, the 0-4 segment reached the highest buyer conversion of all groups (3.68%), clearly ahead of segments still routed to the catalog.
Tested only in the primary segment, the free-course flow lifted conversion to purchase 7x year over year. In the test, the 0-4 segment reached the highest buyer conversion of all groups (3.68%), clearly ahead of segments still routed to the catalog.
0.53 % → 3.68 % (7x)
0.53 % → 3.68 % (7x)
0.53 % → 3.68 % (7x)
09. Next lever: gamification as a hook
09. Next lever:
gamification
as a hook
09. Next lever: gamification
as a hook
Free courses worked, but the core constraint hadn't changed: the parent pays, yet the decision runs through the child. So the real job was to make the child want to come back every day, on their own.
Here I turned to Yu-kai Chou's Actionable Gamification and his hook model. The takeaway: to make someone come back on their own, you have to create real value and an emotional attachment that keeps pulling them in.
Free courses worked, but the core constraint hadn't changed: the parent pays, yet the decision runs through the child. So the real job was to make the child want to come back every day, on their own.
Here I turned to Yu-kai Chou's Actionable Gamification and his hook model. The takeaway: to make someone come back on their own, you have to create real value and an emotional attachment that keeps pulling them in.
10. The currency test
10. The currency
test
The platform already had a character room, where a child creates an avatar and dresses it up. Kids loved it, but it didn't go anywhere: only a handful of costumes, so you'd set up your character once and never open it again. No reason to return, and no link to learning at all.
Hypothesis: if we introduce a currency earned through educational actions, and let kids spend it on new, exclusive costumes for the character, it will lift activation of paid products, deepen engagement in learning, and potentially improve LTV.
The platform already had a character room, where a child creates an avatar and dresses it up. Kids loved it, but it didn't go anywhere: only a handful of costumes, so you'd set up your character once and never open it again. No reason to return, and no link to learning at all.
Hypothesis: if we introduce a currency earned through educational actions, and let kids spend it on new, exclusive costumes for the character, it will lift activation of paid products, deepen engagement in learning, and potentially improve LTV.
Coins, earned by learning.
11. Coins, earned
by learning.
by learning
I introduced coins as a currency, rolled out as an A/B test, and built the task system behind them. Kids earn coins for educational actions: starting a course, watching a lesson recording, solving homework without hints, winning a round in Magical Duel (our educational game). Paid-course actions carry a bonus, so the more a child commits to learning, the faster the coins add up. They then spend everything in the store on costumes for their character.
I introduced coins as a currency, rolled out as an A/B test, and built the task system behind them. Kids earn coins for educational actions: starting a course, watching a lesson recording, solving homework without hints, winning a round in Magical Duel (our educational game). Paid-course actions carry a bonus, so the more a child commits to learning, the faster the coins add up. They then spend everything in the store on costumes for their character.
What's new in the room
12. What's new
in the room
in the room
The room, now with a coin balance up top and a wardrobe of costumes to unlock. The more you learn, the more you earn, and the more your character can wear.
The room, now with a coin balance up top and a wardrobe of costumes to unlock. The more you learn, the more you earn, and the more your character can wear.
Outcomes
Outcomes
Outcomes: kids who saw the coins converted to activation 13 percentage points higher than those who didn't, and half of them activated within 3 hours of seeing the coins. A solid result, so we decided to keep building on it.
Outcomes: kids who saw the coins converted to activation 13 percentage points higher than those who didn't, and half of them activated within 3 hours of seeing the coins. A solid result, so we decided to keep building on it.
Extending the loop: Magical Duel
14. Extending the
loop: Magical Duel
Extending the loop:
Magical Duel
The room worked, so the next step was to pull more of the product into the same loop. The obvious candidate was Magical Duel, our educational game.
In it, kids play 1-on-1 matches and learn through them. They think they're up against another student, but the opponent is actually a bot. We dress that bot in standout costumes on purpose. Facing a cool-looking opponent makes a child want to look the same, and the only way to get there is back in the character room, earning and spending coins.
Until then, the game was free and lived on its own. So I plugged it into the system: winning a duel now earns rewards (coins, costumes, coupons) on the same currency as the room and the courses. Kids earn through learning and duels, spend in the room, and every part points back to studying.
The room worked, so the next step was to pull more of the product into the same loop. The obvious candidate was Magical Duel, our educational game.
In it, kids play 1-on-1 matches and learn through them. They think they're up against another student, but the opponent is actually a bot. We dress that bot in standout costumes on purpose. Facing a cool-looking opponent makes a child want to look the same, and the only way to get there is back in the character room, earning and spending coins.
Until then, the game was free and lived on its own. So I plugged it into the system: winning a duel now earns rewards (coins, costumes, coupons) on the same currency as the room and the courses. Kids earn through learning and duels, spend in the room, and every part points back to studying.
Outcomes
15. Outcomes
The rewards weren't just coins, but costumes and discounts on courses too.
First-month results: 45% of users opened the achievements screen, 86% of them earned at least one reward, and we got 30+ purchases through Magical Duel achievements.
The rewards weren't just coins, but costumes and discounts on courses too.
First-month results: 45% of users opened the achievements screen, 86% of them earned at least one reward, and we got 30+ purchases through Magical Duel achievements.
Clans: turning the game
into a referral engine
16. Clans: turning
the game into
a referral engine
The loop so far deepened engagement and monetization, but only for users we already had, like home-school students who were already paying a lot. It didn't bring new people in. We needed new registrations, and Magical Duel was the most social part of the product, so the question was: how do we make the game itself attract new players?
The answer was a clan system. A child creates a clan, invites friends and relatives by link, and the clan competes for prizes together. Inviting people is the natural thing to do, which turns kids into a referral channel of their own, beyond our marketing.
Before launching, I redesigned the game's main menu to make room for clans and added a "Create a clan" button. The response was immediate: 26% of players who saw the page tapped it, and 70% of them said they were eager for the feature.
The loop so far deepened engagement and monetization, but only for users we already had, like home-school students who were already paying a lot. It didn't bring new people in. We needed new registrations, and Magical Duel was the most social part of the product, so the question was: how do we make the game itself attract new players?
The answer was a clan system. A child creates a clan, invites friends and relatives by link, and the clan competes for prizes together. Inviting people is the natural thing to do, which turns kids into a referral channel of their own, beyond our marketing.
Before launching, I redesigned the game's main menu to make room for clans and added a "Create a clan" button. The response was immediate: 26% of players who saw the page tapped it, and 70% of them said they were eager for the feature.
[ Results ]
Key Results & Impact
+0%
+0%
Activation Rate
+0%
+0%
Purchase Conversion
+0%
+0%
Activation Rate
+0%
+0%
Purchase Conversion