Model Insights

Insights turns your model results into clear visual stories, so you can see which channels are driving your outcomes, how that has changed over time, and how much of your results comes from factors beyond media.

It organizes model output into Media Analyses and Baseline Analyses, and is hidden by default. You can enable it from Customize Tabs.

See what your media delivers

Media Analyses show how your marketing channels contribute to your outcome, how efficiently they do it, and how their roles shift over time. Where available, switch between channel and tactic (the way a channel is used, such as prospecting or retargeting) views to see results at the level you need.

Contribution waterfall

The waterfall shows how your media variables (the channels, campaigns, or tactics your model measures) build up to the selected outcome. Use it to identify your largest contributors and understand the relative scale of each one. For example, you can quickly see whether a single channel accounts for most of your media-driven revenue.

Efficiency grid

The grid compares contribution with efficiency, helping you tell apart channels that drive volume from channels that drive value.

• High contribution, strong efficiency: these channels may be your most important current drivers.
• Low contribution, strong efficiency: these channels may warrant a closer look. Check their scale, uncertainty, and saturation (the point at which extra spend starts returning less) before deciding whether they can grow.

Ranking chart

The ranking view shows how channel or tactic positions change across the selected periods. Focus on sustained movement rather than one-off jumps, then check whether a change in spend or efficiency explains it.

Contribution heatmap

The heatmap shows how contribution changes over time. Use it to spot periods where contribution is concentrated, recurring patterns, and changes worth comparing against your campaign activity. For example, a spike that lines up with a product launch is easy to explain; one that doesn't is worth investigating.

Understand what drives your baseline

Baseline Analyses explain the non-media portion of your modeled outcome (the results you would expect even without paid media, driven by factors like brand strength, pricing, seasonality, and market trends). Understanding your baseline helps you see how much of your performance media is truly responsible for.

• Baseline factors: lists the variables that contribute to baseline demand.
• Relative baseline: compares the scale of the available baseline components.
• Trend lift: shows longer-term movement associated with the modeled trend.
• Decomposition: shows how baseline components change over time.
• Seasonality and holiday analysis: highlights recurring calendar patterns and configured events, such as holiday peaks or annual sales periods.

Confirm before you act

Insights are analysis views, not standalone recommendations. Before acting on a finding, back it up with supporting evidence from the rest of your model.

For media findings

• Validate the result in Contribution.
• Review its causal paths (how the model connects each channel to your outcome) in Graph.
• Check model health in Diagnostics.

For baseline findings

Compare the timing with known pricing, distribution, promotional, seasonal, and market events. For example, a rise in baseline demand that matches a price cut or a new retail partnership is likely explained by that event.

---

Frequently Asked Questions

Why can't I see Insights?
Insights is hidden by default. Enable it from Customize Tabs.

What is the difference between Media Analyses and Baseline Analyses?
Media Analyses show how your marketing channels contribute to your outcome. Baseline Analyses explain the non-media portion, the results driven by factors like seasonality, pricing, and longer-term trends.

Can I view results by tactic instead of channel?
Yes, where available. Media Analyses let you switch between channel and tactic views.

What should I do with a channel that has low contribution but strong efficiency?
Take a closer look at its scale, uncertainty, and saturation. Strong efficiency at low volume may point to room to grow, but only if the channel isn't close to saturation and the estimate is reliable.

How should I read changes in the ranking chart?
Focus on sustained movement across periods rather than one-off shifts, then check whether changes in spend or efficiency explain it.

What does the baseline represent?
The baseline is the non-media portion of your modeled outcome: results you would expect even without paid media.

Are Insights recommendations I can act on directly?
No. Insights are analysis views. Validate media findings in Contribution, Graph, and Diagnostics, and compare baseline findings with known business and market events before acting.

Why does a baseline change not match any media activity?
Baseline changes are driven by non-media factors. Compare their timing with pricing, distribution, promotional, seasonal, and market events to find the likely cause.

---

Related Articles