Back to all articles

How We Outperformed Re-subscription Cost KPI 4.5x for Wavve Content Streaming Platform

Download the PDF version of the case study

Persona.ly worked with Korean mobile marketing agency Mobidays, to re-engage users of the Wavve app, one of the most popular video streaming services in Korea enjoyed by millions of users.

Campaign Goal

The campaign’s ultimate goal was to re-engage users who used to be paying subscribers but stopped their active subscription within a certain period with Purchase Cost (re-subscription cost) being a KPI.

Results at a Glance:

>100

Reached and outperformed the Re-subscription cost KPI by week 2

4.5x

Outperforming cost per re-subscription KPI by week 9

Programmatic Re-engagement Process

When we started running the campaign for Wavve, after receiving the audience for re-engagement the exploration phase of the campaign has started.

Persona.ly’s programmatic DSP is integrated directly with tier 1 ad exchanges and is processing over 3 million ad auctions per second. For each auction we access over 60 data-points to define the most valuable user cohorts. This approach allows our machine-learning algorithm to build a unique classification model for each app. 

Within the first few days of running the campaign, after receiving enough positive signals (re-installs and re-attributions, for our algorithm to create a basic targeting model. By week two, our bidder was able to distinguish between audience who had a higher chance to convert (re-subscribe to Wavve), which later allowed us to focus on the audiences who had a higher probability of becoming paying subscribers again.

Application of Contextual bandits for creative optimization, where certain creatives were displayed to corresponding user cohorts based on their affinity to the specific promoted content genre, amplified the campaign performance significantly. The close collaboration with Wavve and Mobidays ensured full alignment on the strategy and receiving critical deliverables and feedback to improve the outcomes.

Re-engagement Campaign Results

Cost per Re-subscription and Conversion Rate

The initial signals gathered within the first week of running the campaign allowed our ML algorithm to create a basic targeting model.

Targeting cohorts of users who would more likely to re-install or re-open the app, we reached the KPI and brought audience who would re-subscribe to the app at a cost twice lower than the KPI

Further campaign optimization focusing on audiences that would re-open/reinstall the app and re-subscribe to the service allowed us to almost double the re-subscription conversion rate and significantly reduce the cost per re-subscription, outperforming the KPI 4.5 times.

Contextual Bandits for Creative Optimization

The utilization of contextual bandits for creative optimization became a game-changer for the campaign performance. 

Deriving from MAB (multi-armed bandit) creative optimization where the most prominent creatives get the highest share while the rest of the creatives are being tested to explore new creative champions, contextual bandits allow to display different creative champions to different audiences based on multiple contextual signals.

In the case of Wavve, where 4 types of VOD content (4 genres: historical drama, comedy, romantic comedy, and action) were promoted, each leading to the specific page within the app using deep links, Persona.ly’s bidder matched the creative champions with the audience based on their affinity to the genre. The affinity was calculated using the combination of factors defined by Persona.ly’s proprietary ML algorithm, and in case of Wavve the main factors were: predicted gender, predicted age and placement context.

This allowed us to display the most relevant creatives for each audience type leading to a significant performance improvement.

Yoonju Song, Manager at Mobidays

“Our team-up with Persona.ly flowed smoothly, thanks to their quick integrations, clear communication and outstanding outcomes leveraging their proprietary data to surpass our KPIs. Their custom optimization tactics gave us amazing insights every step of the way”

Yoonju Song, Manager at Mobidays

About Wavve

Wavve stands out with its diverse content library, featuring everything from popular Korean dramas to exclusive originals, appealing to a wide range of viewers. By forging strategic partnership with leading Korean broadcasters, wavve ensure a steady flow of high-quality content, enhancing its appeal and market presence, all while protizing technological innovation, user-friendly interfaces, and personalized monetization strategies.

About Mobidays

Mobidays, a leading digital marketing agency, specializes in business-centric solutions, adept at optimizing resources through strategic KPI tracking, ensuring seamless campaign management and exceptional quality throughout the entire process. 

Jieun (Rachel) Seo, Business Development Director - Korea at Persona.ly

We at Persona.ly are thrilled to enter the Korean market and extend our presence globally.  We see a great potential on the market, where the advertisers are extremely savvy and understand the value of high-quality traffic along with ML-driven targeting.

The mutual trust we achieved with Mobidays and Wavve allowed us reach truly great results and we are looking forward to growing our collaboration.

Jieun (Rachel) Seo, Business Development Director – Korea at Persona.ly

About Persona.ly

Persona.ly is a mobile-first DSP operating worldwide. Using our proprietary bidder and machine-learning algorithms, we offer transparent, performance-driven, highly targeted UA and retargeting solutions at scale with access to over 3 million ad auctions per second. We are trusted by Nexon, NextNinja, Papaya Gaming, Rapido, Games24x7, Ubisoft, Tilting Point, and many others.

Persona.ly strives to be more than just a vendor for its partners, but a partner that helps generate actual value, growth, and broad marketing insights that can be used across channels.

Share
Processing...
Thank you! Your subscription has been confirmed. You'll hear from us soon.
ErrorHere