Ripio is a pioneer fintech and crypto platform operating in Argentina, Brazil, Mexico, Colombia and Uruguay.
They grew fast in terms of user acquisition during bullish years. However with the market in a downturn, active users were declining, and fewer were power users. Ripio needed to kill churn and rebirth its growth.

To revive its growth and make it sustainable, Ripio had to rethink how to engage its users across the entire funnel. This meant retaining users after activation to increase the number of trades and expand their lifetime value (LTV).

Retention needed to be approached in a holistic way. A solid acquisition strategy in Paid Media to attract users should be complemented with personlized touchpoints at lower funnel stages. This meant orchestrating data, integrating the tech stack,  aligning a content strategy, and personalizing customer journeys with Owned Media messages to maximize results.

Ripio’s Growth Team worked together with Winclap’s Paid Media and Retention & Engagement teams to plan, execute, and optimize a full-funnel customer journey to acquire, retain and monetize users.

The project was structured in three stages:

1. A Retention and Engagement Diagnosis (understanding the current scenario)
2. An Owned Media Journeys Audit (recommending improvements)
3. Owned Media Journey Management (executing those improvements)

1. A Retention and Engagement Diagnosis:

The first step was to understand Ripio’s user activity. Winclap Data Scientists used Ripio’s transactional data from the last twelve months to feed an unsupervised machine learning model and extract user behavior patterns. The clustering algorithms in this model segmented users according to their activity level. With this grouping technique, Ripio could identify which were casual, core, or power users. 

Machine learning clustering algorithms were used to segment users and calculate their value to the business.

Retention curves were built for each cohorts of users to understand how its in-app activity evolved over time.

Analyzing metrics like Quick Ratio ([New+Reactivated]/Churn), share of active Users per group, trade volume, and revenue per cluster provided insights into how much value each segment brought to the business.
Additionally, the team calculated each cohort’s retention curves to understand how its in-app activity evolved over time.
The team now had a clearer picture of the different user segments, and how much value each one had for Ripio. With these insights, they could now start planning a Retention strategy.

2. An Owned Media Journey Audit:

Once the Diagnosis was finished, it was time to audit Ripio’s current tech stack and customer journey touchpoints and compare them against industry best practices. This allowed Ripio to identify gaps and opportunities to increase user engagement.

This Audit included:

Diagnosis of the current situation:

• Tech stack analysis: understanding how martech tools were being used to track and interact with users. Including assessing how the Customer Engagement Platform was integrated with Ripio’s app and how key events and parameters were being tracked.
• Current user segmentation analysis for Argentina and Brazil.
• Customer Journey Analysis: reviewing user activity tagging, message content strategy, and syntax at every stage of the journey.

Gap Analysis & Recommendations:

• Best practices and directions for Customer Data Platform and Customer Engagement Platform integrations, including tagging and data enrichment.
• Segmentation: grouping users into clusters and personalizing messages in re-engagement campaigns for every stage of the lifecycle (Onboarding, Engagement, and Resurrection).

Ripio’s tech stack and customer journeys were audited to improve how user data flow between systems and to personalize Owned Media touchpoints and increase user engagement.

3. Owned Media Journey Management:

After understanding the current scenario and recommending improvements, it was time to start executing those suggestions.

More opens, more onboarding:

One of the Audit findings was a low engagement rate at the onboarding stage. To counter this, new onboarding journeys were implemented to increase message open rates and increase user activations after installs. Using One Signal’s journey tools, the Retention & Engagement team laid out a series of A/B tests to find out the optimal touchpoints to increase conversions. The team tested out push notifications varying time of day, personalized vs generic messaging, frequency,  images, and even tone of voice. This optimization increased Push Notification Open Rates by 2.8x.

Multiple customer journeys were tested. Alternative push notifications were sent testing for optimal time of day, personalization, images, and other parameters.

Less friction, more activations:

With clear onboarding messages and straightforward CTAs, Ripio reduced friction at the onboarding stage. The team was able to decrease the mean time to activation by a 33%. By reducing activation time, more registered users became active ones.
On the first month the New Onboarding Journey was implemented, Ripio obtained the largest share of sign-up and KYC over installers.  Sign-up and KYC completion percentages increased by 14% and 8% respectively.

Ripio tested multiple messages, varying images, tone of voice and product offerings to find the optimal messages to increase onboarding completions.

Tailoring journeys:

One journey doesn’t fit all segments. To increase user engagement, Ripio needed to create unique app experiences for each user of segments at every stage of their lifecycle.
With the segmentation provided by Predictive Classification Model, Ripio understood every cluster’s value and could tailor messages and offers for every one of them. They accurately targeted campaigns for re-engagement, churn prevention, and user resurrection where and when it made sense. This translated as a 20% increase in Retention day60 and a  40% increase in the number of users transitioning from a low-value segment to a high-value segment.

Predicting Churn:

A key aspect of this retention campaign was to target users close to churning. Thus, the first step was being able to predict which users were more likely to churn and engage with them.
Winclap’s Retention & Engagement data scientists ran a Machine Learning Predictive Classification Model to predict user churn based on their past behavior.
Using this model, the team could predict 88% of actual churners one month in advance. This became a crucial tool to launch targeted campaigns to prevent users from abandoning the app.

Overall, Ripio saw a 10% reduction in churn rate vs modeled expectation.
After one month of Winclap’s intervention, Ripio started seeing its user base grow again.
User growth was driven by an increase in reactivated users, which accounted for 60%+ of the total active users for the month.