Conversion Logic was started in 2014, from its offices in Los Angeles, California. The company, which has a revenue is split 85% from software and 15% from services is operated by 24 employees and led by its CEO and co-founder, Brian Baumgart. Its target market is retail and the educational industries, with clients like Pearson, Pizza Hut, and Proactiv.
The company’s XC Logic Platform employs an ensemble method to measure conversions across channels. In addition to attribution, the platform provides tools for reporting and simulation. Conversion Logic uses a different model to incorporate offline and no media factors such as seasonality. It combines this aggregate view with its Multitouch Attribution (MTA) model to provide unified media measurement. Its goal is to provide a measurement tool that is accessible for the business user without heavy consulting or IT support.
Conversion Logic delivers attribution, evolved: a cross-channel measurement platform that combines cloud analytics and machine learning for enterprise marketers. Built from the ground up for agility, innovation, and speed, Conversion Logic translates the most sophisticated data science on the market into clear, actionable insights across the customer journey. With media agnostic, real-time analytics and optimization, clients reduce friction, adapt, and realize value more quickly than ever before.
The methodology that Conversion Logic employs consists of the models of Follow the regularized leader (FTRL), Logistic regression, Random forest, and Ensemble models. Follow the regularized leader (FTRL) is a form of standard online gradient descent algorithm (also known as stochastic gradient descent), which is a class of algorithms that are able to adapt predictive models rapidly based on streaming data. Logistic regression is, albeit a regression model, this model is used as a classifier. It maps the dependent variable onto an interval between 0 and 1 and so can be translated into a probability that the variable is in a particular class. Random forest is to ensemble method that combines many dozens (or hundreds) of decision trees, which adds randomness into the test conditions at each node to reduce overfitting. Ensemble models can be defined as a method to reduce variance in individual models by combining a number of them and averaging predictions. Many dozens (or hundreds) of decision trees can be combined into a random forest, which adds randomness into the test conditions at each node to reduce overfitting. Together these models form the algorithms that works together as an uber-model to improve accuracy.
The edge that Conversion Logic offers that that this platform’s ensemble approach makes it more adaptable to different client scenarios, while its relatively recent launch means it is built on the latest infrastructure. Plus, this is a solution to be considered if you have an in-house data science team or agency that appreciates a flexible method that can adapt to your own requirements, and if you’ve had some false starts with other attribution providers and would like to try a new and different approach.
Something to watch out for it’s the expansion that Conversion Logic will be making. In March of this year, the company got a $9 million injection from Pelion Venture Partners, with participation from Rincon Venture Partners, Crosscut Ventures, Lerer Hippeau Ventures, Founder Collective, Revel Partners and TenOneTen. That meant that Conversion Logic has raised $14.1 million which will be used for its expansion of its sales and marketing, as well as data science and engineering efforts. The company already works with some of the globe’s biggest companies like ADT and Microsoft, providing them with advanced enterprise analytics solutions. The result is a more sophisticated approach which improves cross-channel marketing efficiency by 30-50%. More so, a view across all offline and online marketing channels, through to conversion, is imperative to understand true return on investment and to inform how and where to optimize and scale. The company’s Ensemble machine learning framework and proprietary methodology resonates with CMOs who are no longer willing to accept a “one size fits all” single algorithm approach. Conversion Logic’s data science driven platform provides the solution to the market’s most pressing challenges – analytics and applications for cross-channel attribution. The platform offers marketers clear visibility into what’s happening across all channels, and provides a portfolio view into synergy between stages of the customer path and relevant channels, as well as tactical insights to increase scale and cut waste.
The case of Netmarble Games
As part of release of the latest movie in the Marvel franchise, Captain America: Civil War in 2016, Netmarble Games launched its Marvel Future Fight campaign. One issue was that it was Netmarble Games’ first TV ad experience. The company, which is based in Seoul, Korea, and is known for its successful mobile games such as Marvel Future Fight, Seven Knights, Monster Taming, Raven, Everyone’s Marble, and the Magu Magu Series, wanted to get a comprehensive breakdown of how the campaign performed across channels, shows and time frames. The application of Conversion Logic’s analysis provided them with insights, but allows them to repeat and optimize their campaign success in the future. What is important is that TV advertising is one of the most expensive platforms. It is thus vital to know how each investment performed in relation to actual game installs and we were able to deliver the data necessary for them to make confident, informed decisions for future TV campaigns. Built with the media practitioner in mind, Conversion Logic is the only unified marketing analytics provider that utilizes the Ensemble Method, combining the traits of multiple established algorithms, always resulting in higher accuracy than any one model. The newly updated XC Logic™ platform provides a comprehensive unified view across all marketing channels.
