Bridging Mobile Apps with Engaged Users

Programmatic DSP for Mobile User Acquisition and Retargeting.

Using our proprietary Machine Learning algorithms, we offer transparent, performance-driven, highly targeted UA and retargeting solutions at scale.

mock
Trusted by:
Programmatic UA clients
  • Our bidder processes over 2.8M ad auctions per second
  • We analyze 60 data-points before serving each impression
  • Our global reach allows us to launch a campaign in almost any market within hours
Machine learning programmatic model

Utilizing Machine Learning

A proprietary machine learning algorithm is at the core of our programmatic DSP. Processing over 2,800,000 ad auctions a second while precisely targeting the audience, our platform accurately predicts the value of each impression and which users are going to interact with an app. It then bids accordingly, allowing us to achieve deep-funnel event KPIs.
Various classification models and data enrichment enable reliable classification, including age and gender prediction.

Our Data-Driven Process

While similarities do exist between apps within a specific vertical, each app is unique and finding more engaged, loyal users requires data, cooperation, and creativity. In our years of experience running mobile app campaigns, we have gathered knowledge based on proven business logic to lead and expedite the initial targeting model generation process, allowing us to reach KPIs and scale faster.

  • Exploration

    Exploration

    Bidding on impressions based on basic demographic definition to start gathering data.

  • Basic logic

    Basic logic level

    A basic targeting algorithm based on the initial gathered data and proven strategies.

  • Unique classification Model

    Unique classification model

    A tailor-made targeting algorithm, optimized towards installation.

  • Engagement optimization

    Engagement optimization

    We utilize full-funnel data to optimize the algorithm towards improved retention and in-app purchases.

  • Scaling

    Scaling

Utilizing First-Party Data

Our algorithms analyze over 60 data points — including aggregated data from our DMP—allowing us to reliably predict which users are going to click, install, and bring revenue to your app.

First-party data in programmatic

Impression Level Features

  • Placement in App
  • Day & Time
  • Creative Type
  • RTB Exchange
  • Historical Placement Performance

User Centric Features

  • Category Retention
  • Geo Location
  • Installed Apps
  • Usage Hours
  • Perceived Persona
  • Session Depth

Engagement Analysis

  • Click Tendency
  • In-ad Behavior
  • Click Depth
  • Viewability
Programmatic ios campaigns

SKAdNetwork campaigns

For SKAN and post-iOS 14.5 campaigns, we utilize machine learning-driven natural language processing (NLP), enabling us to determine the contextual relationship between our advertisers′ apps and the apps where the ads are shown. This helps us model and bid according to the contextual ′distance′ of each apps couple. Aсcess SKAdNetwork performance data in our dashboard.

Dynamic Audiences Re-Engaged

Using our proprietary Dynamic Audiences segmentation engine, Persona.ly′s DSP can define user segments based on their in-app behavior to drive high ROAS.

Programmatic dynamic audiences
Multi-Armed Bandit (MAB)

Scientific Approach to Creative Selection

Our ML-driven model is adjusted to promote best-performing ads from the first positive event while constantly testing the rest of them based on a Multi-Armed Bandit (MAB) algorithm. As learning progresses, the ML model gradually transitions to the Contextual Bandits approach, where different creative ″champions″ are displayed to different groups of users based on group features.

Engaging Ad Formats

  • Playable & Interactive Ads
  • Dynamic Ads
  • Video
  • Native
  • Banners
Mobile playable ads
Mobile dynamicsAds ads
Mobile video ads
Mobile native ads
Mobile banners ads

Playable ads are short (approx. 30 sec.) interactive ads where users can experience the product or game before downloading it.

Auction Price Prediction

Machine learning algorithm predicts the auction price based on various audience attributes and bids as much as needed to win the impression. Known as bid shading, this technique ensures the optimal bid price without overpaying in the realm of first-price auctions.

Bid shading in programmatic

Adding Transparency to the Black Box

Our dashboard provides our clients with detailed insights into their campaigns and allows them to see where ads were shown to potential users. Additionally, it shows how each creative is performing in real time.

  • Creative Analysis

    Easily view and analyze the performance of your different creatives - from conversion to click and install, as well as in-app events, retention, and ROAS.

  • Performance Reports

    Easily view and analyze the performance of your different creatives - from conversion to click and install, as well as in-app events, retention, and ROAS.

  • Demographic Insights

    Easily view and analyze the performance of your different creatives - from conversion to click and install, as well as in-app events, retention, and ROAS.

  • Placement Insights

    Easily view and analyze the performance of your different creatives - from conversion to click and install, as well as in-app events, retention, and ROAS.

CreativeAnalysis
PerformanceReports
DemographicInsights
PlacementInsights
IncrementalityMeasured

Brand Safety-Centric

Because we strive to be the most reliable service for your agency, we work only with trusted partners to ensure high-quality traffic. We guarantee:

  • Premium placements within trustworthy SSPs only
  • Privacy-centric approach
  • Fraud prevention. Our ML models are focusing on anomaly detection both pre-bid and post-bid to prevent any fraudulent activity
Brand safety
Ua experts

Expert Support

We provide full campaign management service to ensure the best performance. Our global team of programmatic experts is there to provide insights on campaign performance, support you at critical times, and guide your team throughout the entire process.

Analytics and Measurement Partners

  • adjust
    adjust
  • appsFluer
    appsFluer
  • kochava
    kochava
  • branch
    branch
  • singular
    singular
  • firebase
    firebase
  • myTracker
    myTracker
  • tenjin
    tenjin
  • mparticle
    mparticle
  • adbri
    adbri

Ready to Gain the Most from Our Expertise?

Take your mobile app user acquisition and re-engagement strategy to the next level.