The Sub-Model Insights module enables users to explore detailed performance insights for specific sub-models, such as brand equity, behavioural interactions, or non-paid media effects. This view supports granular analysis of intermediate KPIs and complements the other modules.

TABLE OF CONTENTS

• 1. Global Controls and Page Structure
  • 1.1 Page layout
  • 1.2. Time Period Selector
  • 1.3. Driver Filter (Media/Business Drivers)
  • 1.4. Sub-Model Selector
  • 1.5. KPI Selector
• 2. Media Tab
  • 2.1. Media Performance Summary 
  • 2.2. Media Performance Data Table
• 3. Media page controls
  • 3.1 Data View widget:
  • 3.2 Data Table buttons
• 4. Business Tab
  • 4.1. Tree map chart
  • 4.2. Waterfall Chart
  • 4.3. Evolution Over Time Chart
  • 4.4 Controls
• 5. Use Cases

1. Global Controls and Page Structure

1.1 Page layout

The Sub-Model Insights page consists of two key tabs:

• Media Tab – shows the aggregated impact of media drivers.
• Business Tab – shows observed business outcomes and driver contributions.

Each tab is fully interactive and includes filter controls, tooltips, drilldowns, and configurable dropdowns.

1.2. Time Period Selector

Shows the period for which insights are being displayed. 

E.g. “27 Feb 2022 – 26 Feb 2023”.

1.3. Driver Filter (Media/Business Drivers)

Filter panel showing only relevant drivers based on the active tab:

• In Media tab, only media drivers are shown (e.g., Paid Search, VoD).
• In Business tab, only business-related drivers appear (e.g., Distribution, Competitor Activities).

1.4. Sub-Model Selector

Dropdown to choose which sub-model you want to analyse:

• Brand
• Web Visits
• App Engagement

Changing the sub-model updates adjusts available KPIs and drivers accordingly.

1.5. KPI Selector

Dropdown to choose the intermediate KPI tracked within the selected sub-model:

• For Brand, common KPIs may include: Awareness, Consideration, Usage.
• For Digital sub-models, you may see KPIs like Site Visits, App Opens, or Session Length.

Changing the KPI updates the data shown in both Media and Business tabs.

Example: Selecting “Brand” as the sub-model and “Consideration” as the KPI allows you to assess how media impacted brand consideration over time.

2. Media Tab

The Media tab presents how media investments have impacted using the aggregated effect method. In this method, the full impact of media - including delayed or decayed effects - is reported on the day of investment.

Key Components

2.1. Media Performance Summary 

Located at the top of the screen, these panels display four core metrics as defined by your organisation:

• Media Investment: Total spend attributed to each media channel.
• Media Generated Effect: Total KPI uplift generated by each channel. E.g. the effect can be usage, awareness or consideration.
• Media Efficiency: The amount of effect per unit of investment.
• Cost per Action: The cost required to generate one unit of effect.

Each Column includes:

• A channel-wise breakdown (e.g., TV, Paid Search, VoD). 
• Percent change compared to the previous period.
• A trend line to illustrate historical progression.
• A tooltip icon (ⓘ) that explains the metric in more detail, including scale (e.g., millions).

You can click on any tile, to drill down further in the media hierarchy. 

2.2. Media Performance Data Table

Beneath the summary cards is a detailed table showing the same metrics in tabular format, grouped by media dimensions.

Functionalities include:

• Expandable rows to drill down into hierarchy (e.g., campaign, publisher).
• Columns for each metric: Media Investment, Generated Effect, Efficiency, and CPA.
• Controls in the top-right corner for filtering or adjusting the table view (e.g., “Filter” and “Focus”).

3. Media page controls

3.1 Data View widget:

It is present at the top right end of the screen offering customisation and deeper insight views.

3.1.1 Hierarchy

• Defines how data is structured in tables.
• Allows reordering dimensions such as:
  • Media Grouping → Channel → Publisher
• You can deselect or reorder levels to control the table depth.

3.2 Data Table buttons

Customise and extract value from tabular data using the following tools:

3.2.1 Download/Export Options

• Export current table view or full dataset as Excel (.xlsx).
• Options typically include:
  • Displayed Values Only
  • Full Dataset Grouped
  • Full Dataset Granular

3.2.2 Show/Hide or Reorder Columns

• Add/remove columns as needed.
• Drag-and-drop to change column sequence.
• Useful for comparing specific KPIs side-by-side (e.g., ROI vs CPA).

3.2.3 Table Filter Option

• Hide rows with zero values.
• Filter out irrelevant or low-impact rows for cleaner views.

3.2.4. Expand/Collapse Buttons

• Located beside hierarchical rows.
• Expand to reveal sub-levels (e.g., Publisher within Channel).
• Collapse to simplify the view.

4. Business Tab

The Business tab displays results using the observed effect method. This means the platform reports driver impacts at the time they are actually observed, not when the investment occurred. This view is ideal for tracking real-world attribution and lagged effects.

Key Components

4.1. Tree map chart

You can switch between a Tree-Map and waterfall view using the “View As” dropdown. This visual groups the contributionof drivers (e.g. publishers or channels) as proportional blocks.

Features:

• Each coloured tile represents a driver (e.g. PMIPUBLISHER8).
• The size of the tile reflects the magnitude of its  contribution to the KPI.
• A legend appears alongside, listing each driver and its absolute contribution and share of total positives.
• You can click on the tile to drill down the hierarchy. And also hover over to view a detailed description.
• Grouping can be customised using the “Group By” dropdown (e.g. Publisher, Product, Format, Media grouping, channel etc).

4.2. Waterfall Chart

A horizontal waterfall chart shows how each driver has contributed to the total value of the selected KPI (e.g., uplift in sales or brand index score).

• Each bar represents a different driver (e.g., Base, Media, External Factors, Competitor Activity).
• Bars may be positive or negative depending on whether they drove or reduced performance.
• Hovering on a bar reveals detailed tooltips, including:
  • Numeric value
  • Percent of total impact
  • Contextual notes where applicable

The waterfall chart includes a “Show Legend” button, located in the bottom-right area of the chart. When enabled, this legend helps you identify colour-coded drivers and their contribution types (e.g. media, base, distribution).

• The legend enhances interpretability of the waterfall, especially when many drivers are visualised.
• You can click “Hide Legend” to reduce visual clutter and maximise chart space.
• The legend is dynamic and updates based on the active filters, time selection, and driver grouping.

4.3. Evolution Over Time Chart

Located below the treemap/waterfall chart, this section displays monthly or weekly performance over time:

• A stacked bar chart showing how various drivers evolved across time.
• An overlaid line chart tracks total KPI movement.
• A dotted line allows you to compare as well.
• A legend allows you to toggle driver visibility for clarity.
• Hover tooltips display specific driver values and annotations per time point.

4.4 Controls

1. Time Period Selector: Allows you to define the time granularity (e.g., Quarter, Week or Year).

2. View As Selector: Choose how to visualise contributions, you can choose between Graphical table, data table and timeline view.

a. Graphical Table

• Combines a standard data table layout with mini bar charts for visual comparison.
• Each row corresponds to a time period (e.g. months), and each column represents a driver (e.g. publisher, media channel).
• Horizontal bars within cells help highlight relative contribution size.
• Ideal for users who prefer visual summaries with numeric precision.

b. Data Table

• A pure numeric format without visual bars.
• Offers a clean, text-based view of contributions.
• Suitable for users preparing reports or exporting data for further analysis.

3. Show Context

• Activate this mode to scale graph bars based on maximum value in each column or across the whole table.
• Useful in the graphical table view for interpreting relative contributions.

5. Use Cases

 1: Evaluating Media Efficiency for Brand Awareness

• Select sub-model: Brand
• Select KPI: Awareness
• Use Media tab to identify which channel had the best efficiency (e.g., Paid Search may have 10.4M effect with low investment).
• Use this insight to optimise future media plans.

2: Investigating a Drop in Consideration

• Sub-model: Brand
• KPI: Consideration
• In Business tab, use Waterfall Chart to find that “Competitor Activity” drove negative effect in Q4.
• Use this to plan counteractive brand activity.

3: Tracking Digital Engagement via App Opens

• Sub-model: App Engagement
• KPI: App Opens
• Use Media tab to see which digital channels (e.g., Paid Social or Display) delivered highest engagement.
• Use Business tab to see impact of user experience changes or external factors.