Aarki is a 100% real-time bidding (RTB) demand-side platform (DSP). We help companies grow and re-engage their mobile users using machine learning (AI), big data, and engaging creatives. Using our integrations with all the major global exchanges for RTB, we are able to access nearly all of the global inventory from our five data centers. To drive performance, we activate our data assets through proprietary machine learning algorithms and engage users in real-time with personalized creatives.
Our programmatic advertising platform, Aarki Encore, informs our real-time bidding decisions and targets users valuable to the advertiser to re-engage lapsed users. It does this by predicting post-install in-app behavior using our proprietary machine learning algorithms.
Using our creative expertise and Aarki Studio, our proprietary creative suite, we’re able to produce highly personalized HTML5 ads that dynamically optimize messages to the individual user at ad serving time. These strategies are integrated with AI-powered media optimization to ensure advertising is not only personalized but also optimized for efficiency at scale.
Creative platforms have traditionally been disconnected from the performance optimization loop, but we at Aarki have challenged this. We help advertisers deliver superior app marketing performance through the unification of dynamic creative optimization and programmatic media buying. With Aarki Encore and Aarki Studio, app developers can run ROI-positive marketing campaigns showing the most relevant ad creative for a specific user at a specific impression. The intersection of artificial intelligence (AI) and marketing delivers this highly customized experience.
For app developers, acquiring high-quality users is key to long-term success. As a DSP, Aarki is often one of many channels used to acquire and retain users. With real-time bidding, it is crucial for us to predict the expected revenue from a particular ad impression so as to make the correct bid. All marketing campaigns are measured against specific key performance indicators (KPIs) including ROI. We use a combination of Aarki-attributed post-install event data and non-attributed app event data to train our machine learning models for these KPIs.
With the PMI (Pointwise Mutual Information) algorithm, we can effectively model Aarki-specific user conversion funnels while pre-training on non-attributed omni-channel event data. This allows us to better calculate the purchase probability of app users and thus improve ROI prediction. It also allows us to calculate the install probability and decrease the cost per install (CPI) of a campaign.
The advertiser’s non-attributed event data is used to compute pairwise correlations between user profile features and in-app purchase events. These correlations are then used to “warm-up” features in the direct optimization model, which predicts purchase probability at impression time. This feature encoding technique allows features that are strongly associated with the target event to carry more weight in the model.
To determine the value of the PMI model, we analyzed a programmatic advertising campaign for a game app. The result was 62.58% higher normalized ROI, 209% more installs, and 3% lower CPI.