ÃÛ¶¹ÊÓƵ

Transcript

A standard customer workflow for Mix Modeler looks like the following. First, ingest. Customers ingest their relevant data into Mix Modeler. This includes event-level data, aggregate-level data, and exogenous variables. We offer various data harmonization and QA tools to streamline the data ingestion process. Second, configure. Customers can easily configure their models to suit their business needs within Mix Modeler’s self-serve UI. Customers can specify things like which conversion KPIs and channels to measure, length of forecast windows and lookback windows, model constraints, and more. Third, score and train. Mix Modeler will use the ingested data and user configurations to train AI-ML measurement models. Mix Modeler will learn the most accurate model and use it to calculate various incremental impacts and ROIs. This AI-as-a-service capability is a key differentiator of our product. Lastly, action. Customers can easily access their measurement results and use these results to create optimizations in ways that increase their ROI. They can do this in multiple ways. Customers can review their results and gain insights. Summary-level results can be viewed in Mix Modeler’s Insights dashboard. Note that at GA, Mix Modeler’s UI does not have the capability to analyze or visualize event-level MTA results. For event-level results, customers can use our direct integration with customer journey analytics for analysis and visualization or egress the event-level scores into a third-party BI tool. Customers can use the budget optimization tool in Mix Modeler to automatically get the optimal spend level for each channel, based on their business goals and the budget constraints that they specify. Customers can also use the scenario planning tool in Mix Modeler to develop multiple budget scenarios and compare their impact on the business. This helps them make more informed spend decisions and explore what-if questions about budget.