Planning & ROI Prediction | Overview – ChannelMix

marketing-sales-pipeline-predictive-modeling-recommendation__computer_.png

ChannelMix AI enables marketers to make decisions about how and where to spend marketing dollars to maximize ROI smarter and faster than they could do on their own.

Plan Performance

Create a single source of truth for marketing performance through the Pipeline scorecards

Using historical information, calculate the budget you need to hit your goals.

Once you've established your goals, predict ROI for different spend scenarios and media mixes to determine the how to best distribute your budget across marketing channels.

Quantify marketing’s influence on new leads, transactions, and revenue. Pick your scenario to find the most cost effective way to generate your target metric.

Model Insights Tab

Dig deeper into the mode insights by understanding how trends change over time. The “Model Insights” are just that - insights. They are not there to make business decisions. This tab is here to help you build trust in the model by understanding the recommendations it makes.

The model compares spend over time in multiple channels to your target variable counts. Looking for trends we cannot necessarily see with our naked eye, the model is able to define the amount of spend needed to gain additional target variables by feature (e.g., Channel or Platform).

However, the story doesn't stop there.

The Diminishing Return on Ad Spend shows market saturation for each feature. So in this example, you can see Paid Search has a high cost per and is also a pretty saturated market, so it's not likely the model will recommend spending a lot there.

Next, easily see trends or variability in your data using the ROAS vs Cost Trend to see changes over time by each feature. See how cost compares to the acquisition of your target variable over time.

Lastly, see what other factors might be contributing to your model performance, including organic media, seasonality, and data defined using ChannelMix Control Datasets.

Model Summary Tab

Use the Model Summary tab to "check under the hood" of your model. See how well your model is performing when it was last trained, what dates were included in the training, and any errors or warnings you should review.

Use the Model to look for variations in your data that you might need to look at more closely to better refine your model. Learn more in AI Explained | How do I interpret the Model Performance score for MIM?

In the Model Errors and Warnings section, learn reasons your model is not performing well or ways to make it perform even better.

In this Article

Planning & ROI Prediction | Overview

Plan Performance

Model Insights Tab

Model Summary Tab

Additional Resources