Orchestrate and optimize all customer engagements for maximum revenue


MatchMaker AI can predict conversion probabilities and potential revenue for all products and services potentially offered to customers to understand which currently has the highest revenue boosting potential. The optimization also takes churn and revenue loss risk into account when deciding the optimal offer to present each customer.




By implementing MatchMaker AI into your data and marketing execution your marketing and data science departments will gain a significant edge in presenting customers what they want while ensuring it is also the best revenue driving activity to do.

Minimizing spam! As MatchMaker AI predicts users interests across all major potential offers it is also clear when NOT to interact with a specific customer to minimize risk of churn and retain optimal customer satisfaction. It is actually often the case that users with high churn probability might be triggered to churn with a bad interaction which MatchMaker AI can predict and hence avoid.



Monetize the value of millions of customers and strong first party data


MatchMaker AI can enable operators to boost or create new revenue from ad buyers. This is achived by understanding when and for which individual customer more revenue can be achieved by selling ad space in in-house channels vs internal advertising

MatchMaker AI gives operators a unique insight in the revenue potential of the internal product portfolio as well as risk of churn. By adding external advertisers and the revenue opportunity per display or click as well as learning each customers preference to the type of product advertised MatchMaker AI can dynamically optimize between promoting internal portfolio or churn prevention offers and external advertising revenue in real-time for each customer.


Optimize revenue from specific predictions on plan upgrades, churn prevention and digital service engagement


MatchMaker AI has specific predictions and benchmarks for optimizing sales of operators core products: Plans, Subscriptions and digital services as well as churn prevention


Plans & Subscriptions: MatchMaker AI matches each customer with the combination of most likely and highest revenue boosting plan or subscription in relation to the one currently used. The predictive model not only considers the plan/subscription price but also the long term effects of a potential upgrade.

Churn Prevention: MatchMaker AI understands the current churn risk as well as customer value and potential revenue loss for each customer at any given time. It also predicts the probability of retention given available anti-churn campaign options available in order to optimize if churn prevention offer should be provided as well as which type. Certain retention offers or interactions can have the opposite effect and instead increase churn, MatchMaker AI understands these effects and takes it into account when providing its recommendations.

Digital Services: MatchMaker AI matches each customer with the most likely digital service to start engaging with to boost revenue and loyalty. The predictive model takes revenue and loyalty (churn risk) impacts into account when proposing a new service to a specific customer.


Data Collection


MatchMaker AI securely collects data from two main sources: An in-house developed data collection SDK (“app plug-in”) and customer data from operators data warehouse.

The app SDK: Quickly installs in any Android or iOS app capturing unique device and user behavior events. The impact on device resource use and battery life is minimal and is unnoticeable for end users. The data capture process is fully GDPR compliant and all data is encrypted using a standard 2048 bit RSA key.

Device Insights SDK captures approximately 50 unique data attributes per user/collection, contributing to an extensive data lake where millions of data points are analysed using machine learning and AI. The device data attributes needed for our predictions are all non-personal.

Operators Data Warehouse: Data is synchronized securely using the latest encryption technologies and all customer identifiers are hashed meaning Marakanda is not able to identify any real customers. Normally processing is done in Marakanda’s secured cloud environment but local deployment in country or in-house IT environment is also available to ensure compliance with the most strict consumer and privacy regulations.


MatchMaker AI - How it works



Prediction Processing and Audience Output


MatchMaker AI processes all available data to produce the in-scope predictions and pushes the resulting optimized audience files to marketing teams and execution systems

MatchMaker AI combines the data collected, processing multiple layers of pre-processing and data quality checks and time-series data for customer behavior. All in-scope predictions are processed and conversion probabilities are generated for all in-scope products, services and user actions. Depending on implementation scope the top level optimization is applied to cross optimize revenue potential to identify each customers optimal area of engagement and specific offer.

The resulting audience file is pushed using secure integration to market execution team(s) and most commonly used campaign execution systems.

After campaigns has been executed conversion and funnel feedback is sent to MatchMaker AI to analyze and continuously optimize the predictions.