Kochava Intelligence: Business Services Driven by Predictive Analytics and Machine Learning
Services Include Campaign Insights, Churn Modeling and Ad-Buying Recommendations
Kochava Inc, the industry leader providing holistic measurement solutions for mobile and connected devices, announced an enhancement to their Kochava Intelligence suite of business intelligence tools with campaign analysis on attribution, Insight Reports on ad-inventory buy recommendations, and churn modeling to reduce attrition.
“Marketers are overwhelmed with data and being able to extract the most meaning from it. They must consistently parse and export data to various providers to fully understand their campaign results,” said Grant Simmons, director of client analytics at Kochava. “We address those problems by developing comprehensive tools and maximizing their use on behalf of marketers. Our team of experts is adept at using these predictive tools and surfacing data relationships marketers might not realize on their own.”
Also Read: Kochava Named LINE Ads Measurement Partner
Kochava reveals the true value delivered by media partners and provides recommendations on how to “clean” the signals, ultimately boosting return on investment. They have performed business value assessments (BVA) for some of the biggest companies in the ecosystem. Using machine learning tools, they perform campaign insights and explore the correlations among impressions and clicks and the resulting installs.
“Advertisers don’t realize that the signal — the data stream — they send us may be corrupt. Whether intentional or not, the data from media partners may not be ‘clean.’ Signals for impressions and clicks may be mixed, skewing how marketers use attribution to predict revenue or make future marketing decisions,” Simmons said.
A recent BVA for a gaming company showed that their click-to-install relationship was irrational. Under the premise that clicks drive installs, the analytics team looked for a statistical relationship between the two. If true, more clicks would mean more installs. Their analysis showed the opposite; more clicks led to fewer installs. The reason was an attribution signal full of fraud and impressions sent as clicks rendering the underlying data useless.
“Attribution relies on a good signal. Once they sent us impression, click and install data separately, we saw a drastic increase in the trend line between the company’s clicks and installs,” Simmons said. “Now they can better predict how many clicks are required to satisfy an install goal to boost revenue. They are better able to make marketing decisions with the cleaner data.”
Kochava experts are also available to create custom churn models in an effort to reduce attrition. Marketers can receive data of users most likely to churn and reactivate them in the platform. They can further segment the data for precise retargeting or push campaigns through Engagement.
In line with predictive modeling is the new Insights Report. The report uses machine learning technology to make “buy” and “do not buy” recommendations on certain ad inventory. The report is created using their Optimization Beacon technology for marketers to base their future ad-buying decisions on.
“As a measurement provider, we are the nexus between publishers and advertisers and have the ability to analyze data relationships marketers may overlook,” Simmons said. “We make ourselves available to assist in interpreting the data our tools yield to their fullest extent.”
Recommended Read: Kochava Announces Engagement Push Notifications