Unleash the power of adaptive machine learning and true impression level decisioning with RTB Optimizer. Any company operating a real-time bidding system can use RTB Optimizer to improve performance across display, video, mobile; and all other programmatic channels.
What does it do?
RTB Optimizer is a real-time bid management system that computes a unique probability score for each impression, then uses that score to calculate an optimal bid price based on campaign goals. It can be integrated with any front-end RTB stack and is powered by Datacratic’s Real-Time Machine Learning Platform.
Customers who have integrated RTB Optimizer with their real-time bidding system have realized a reduction in CPM of up to 60%, and up to 65% reduction in CPA when compared to manually optimizing bid prices. These performance benefits combined with the operational efficiency of being able to scale across hundreds or even thousands of campaigns makes RTB Optimizer the logical choice for real-time bid management and optimization.
Who is it for?
RTB Optimizer is applicable to any organization operating, or considering operating their own real-time bidder; including small to mid size DSP’s, Ad Networks and other enterprises who operate large-scale programmatic buying systems. The system can be deployed by DSP’s and Agency Trading Desks who run RTB campaigns on both open and private exchanges, or by Ad Networks who run campaigns on curated 3rd party inventory.
DSP's and ATD's
Ad Networks who wish to add RTB as an inventory supply source, or who want to better optimize their own supply, can integrate RTB Optimizer.
SEM platforms, e-commerce sites, social ad buyers and others can integrate real-time machine-learning and optimization into their platform.
Features and Benefits
Real Time Learning and Optimization
Algorithms Learn from all Available Attributes
Any and all data available in the request at bid time is used in the decision. This allows the front-end bidder to augment any available source of 1st and 3rd party data. Things like regression and probability scores, location data, weather data, audience segment membership are all in scope. The system will ensure all of the available data is used to compute the final bid response.
RTB Optimizer uses a hybrid online/offline learning model: the online portion learns from linear data-sources in true real-time, (i.e. information about outcomes such as clicks are instantly incorporated into the very next bidding decision) and the models are retrained to learn the non-linear sources every 2 to 3 million impressions.
For each optimized campaign the system collects a random sample of bid requests. Newly trained models are then validated on this random sample to ensure training has taken place in a non-biased way.
Network Level / Cross Campaign Optimization
Clients have the option of ensuring data and learnings are never shared between campaigns, or they can choose network level data and cross-campaign learning. This allows models to be trained using data from all campaigns, or any combination of campaigns. By combining the performance data from multiple campaigns, RTB Optimizer receives a significant boost in learning data that yields greater performance.
At each stage of a campaign, RTB Optimizer intelligently chooses the optimal modeling techniques and algorithms from Datacratic’s Machine Learning Platform. As a result, performance improvements happen faster with fewer click or conversion events. As more data becomes available, RTB Optimizer generates additional features and becomes more efficient at finding new and more efficient sources of conversions or clicks.
Frequency / Recency Optimization
RTB Optimizer automatically incorporates frequency and recency calculations to predict success events. Campaigns are optimized based on these important attributes so there is no requirement for campaign managers to manually input frequency caps. Of course the system will respect frequency caps when necessary.
Budget Management and Pacing
Pace on Budget or Impressions
RTB Optimizer modulates bidding behavior intelligently throughout the campaign, responding to changes in the available inventory. This enables the system to maximize ROI on an impression-by-impression basis while pacing evenly and meeting budget or impression-delivery objectives.
Elastic Price CPM provides campaign managers with flexible options to meet campaign budget objectives and maximize revenue (e.g. profit margin) by finding the lowest possible bid price. This enables the campaign to stay within budget while adjusting to the best possible conditions including site quality, time of day and other filters.
Flexible Campaign Goals and Strategies
Place RTB Optimizer into ProfitMax mode to evaluate bid requests based on the potential margin between campaign cost and revenue. ProfitMax will maximize margin while staying within CPC or CPA targets.
Optimize to CPC
The Optimized CPC bidding strategy scores each impression against the probability of generating a click and makes bid decisions accordingly. CPC campaigns are paced to ensure they run evenly within flight dates.
Optimize to CPA
RTB Optimizer improves CPA by training models and scoring real time bid requests based on user actions such as conversions, download, launch, lead generation, and more.
Small Campaign Optimization
With features like Dynamic Optimization, Network Level Data and Cross Campaign training even campaigns with very small budgets and extremely tight flight dates can benefit from optimized performance. Campaigns running locally with extremely small budgets can be optimized at a category level.
Dashboards and Reports
A comprehensive series of dashboards and reports ensure operations managers and developers have visibility into what is happening at an aggregate level while the system is in production. The Campaigns Dashboard provides users with an overall view of campaigns including status and metrics such as CPA, CPC, CPM, CTR and CVR. A collection of operational reports are available on demand to enable developers and campaign managers to monitor bid request volume, spending, bids & wins and service levels. Dynamic pivot tables enable clients to analyze campaign's metrics as well as slice and dice metrics across multiple dimensions.
RTB Optimizer is compatible with the OpenRTB bid request format. A sandbox environment is provided for integration testing. There is a fully documented REST API for pushing campaign configuration information from your UI and the system supports JSON over HTTP for win, click and conversion messages.
Get Started with RTB Optimizer
Contact Datacratic to get started and experience improved campaign performance, increased ROI with RTB Optimizer.