TABLE OF CONTENTS

• 1. Introduction
• 2. Feature Overview
• 3. FAQs
  • 1. What is the Portfolio Long-Term Model and how is it different from the regular Portfolio Model?
  • 2. What is the difference between Observed and Consolidated effects?
  • 3. How do I view long-term impact separately from short-term effects?
  • 4. What are Short-Term Direct, Short-Term Indirect, and Long-Term Indirect effects?
  • 5. Can I analyse multiple products or effects at once?
  • 6. What does the “Long-Term Effect Period” preset do?
  • 8. How can I drill down easily into media performance by channel or publisher?
  • 9. How many products can we have in a portfolio long term model?
  • 10. What should I do if the long-term data appears flat or missing?
  • 11. Can I switch between Observed and Consolidated views after selecting filters?
  • 12. How is brand equity calculated in the long-term model?
  • 13. What data is required to run a Portfolio Long-Term Model?
  • 14. What happens if brand data is missing for some products?
  • 15. What’s the difference between Global and Halo media effects?
  • 16. Why can’t I view multiple impact types at once?
  • 17. Can I use this model for products with seasonal media bursts only?
  • 18. Is the Long-Term Effect Period fixed or configurable?

1. Introduction

The Portfolio Long-Term Model is the most advanced model within the Insights Platform, combining the capabilities of both the Portfolio Model and the Long-Term Brand Model. It is specifically designed to evaluate media effectiveness across multiple products, incorporating both short-term sales impact and long-term brand equity effects.

This model enables users to:

• Understand how media investment influences outcomes at both the individual product and overall portfolio levels.
• Evaluate direct, cross-product, and halo effects.
• Differentiate between short-term spikes in performance and enduring brand-driven growth.
• Analyse ROI using two perspectives: Observed (real-world unfolding) and Consolidated (carryover rolled back to date of investment).

2. Feature Overview

The Portfolio Long-Term Model introduces several unique features on top of the existing Portfolio and Long-Term frameworks:

This model is particularly suited for enterprise brands with:

• Multiple investment products
• Shared or overlapping media campaigns
• A need to balance tactical performance with strategic brand growth

3. FAQs

1. What is the Portfolio Long-Term Model and how is it different from the regular Portfolio Model?

The Portfolio Long-Term Model combines the cross-product analysis of the standard Portfolio Model with long-term brand equity modelling. Unlike the standard model, it captures:

• Short-term direct and indirect effects
• Long-term indirect effects driven by brand memory
• Carryover and decay of media impact across time

It enables more realistic forecasting, especially for brands with lasting media influence.

2. What is the difference between Observed and Consolidated effects?

• Observed Effect shows how media impact naturally unfolds over time, including carryover.
• Consolidated Effect rolls all future impact back to the original investment week, making it easier to evaluate ROI for each media period.

Both views represent the same total impact but distribute it differently over time.

3. How do I view long-term impact separately from short-term effects?

To isolate long-term impact:

1. Change Data Grouping in the widget to Impact
2. Use the Impact Toggle to select Long-Term Indirect
3. The dashboard will update to show only delayed effects associated with brand equity

4. What are Short-Term Direct, Short-Term Indirect, and Long-Term Indirect effects?

Here’s a clear explanation of Short-Term Direct, Short-Term Indirect, and Long-Term Indirect effects, with examples:

Short-Term Direct Effect

Definition: The immediate and measurable result of a marketing action.

Example:
 You launch a paid Instagram ad campaign. Within 24 hours, you see a 30% spike in website visits and a 10% increase in product purchases.

Short-Term Indirect Effect

Definition: A secondary effect that happens shortly after the direct effect, often as a consequence of it.

Example:
 The spike in website visits leads to a surge in customer support queries and newsletter sign-ups, even though those weren’t the campaign’s primary goals.

Long-Term Indirect Effect

Definition: A delayed impact that emerges over time, often influenced by both direct and indirect short-term effects.

Example:
 Months after the campaign, brand awareness improves significantly, leading to increased organic traffic, higher customer retention, and more word-of-mouth referrals.

Use the Impact Toggle to view these effects separately or as a Combined view.

5. Can I analyse multiple products or effects at once?

No - the model supports one combination at a time. You can:

• Select a single product (e.g. Airpods)
• Choose one effect type (e.g. Cross)
• View one impact type (e.g. Short-Term Indirect)

To compare multiple setups, switch filters and use the export options or take notes for side-by-side analysis.

6. What does the “Long-Term Effect Period” preset do?

This time preset selects the period during which 90% of long-term indirect effects are expected to appear - based on model decay settings. It helps users focus on windows where brand equity matters most.

7. How is ROI calculated in the Portfolio Long-Term Model?

ROI is calculated by dividing media-generated profit by media investment, and can be viewed across:

• Individual products
• Different media effects (Direct, Cross, Halo)
• Time horizon (Observed vs Consolidated)
• Short-term and long-term attribution layers

8. How can I drill down easily into media performance by channel or publisher?

Use the Media Performance Data Table:

• Expand rows to navigate the hierarchy: Media Group → Channel → Publisher
• Reorder or filter rows using the table tools
• Click individual rows to explore ROI, profit, CPA at the most granular level

9. How many products can we have in a portfolio long term model?

The platform currently supports up to six product models within one Portfolio-Long-Term structure.

10. What should I do if the long-term data appears flat or missing?

Check the following:

• Confirm that brand metrics were available for the selected product(s)
• Ensure you are in Observed view and have selected Long-Term Indirect
• Verify that the selected time period includes sufficient weeks after media exposure

Some products may not have brand models applied - in which case long-term effects will not be calculated for them.

11. Can I switch between Observed and Consolidated views after selecting filters?

Yes. The Effect Toggle (Observed vs Consolidated) works independently of product or media effect selections. You can toggle between views at any time to see how the same impact is distributed differently across time.

12. How is brand equity calculated in the long-term model?

Brand equity is modelled using an autoregressive sub-model per product. It evolves over time as a function of:

• Past brand equity (with memory parameter)
• Media exposure (direct, cross, halo)
• External controls (if applicable)

This allows equity to decay gradually, representing real consumer memory.

13. What data is required to run a Portfolio Long-Term Model?

The model requires:

• Weekly media spend (tagged as direct, cross, halo)
• Weekly sales or KPI data per product
• Brand metrics (e.g. awareness, NPS) per product (recommended)
• Control variables (pricing, distribution, macroeconomic conditions)
• Profit margins per product (for ROI calculation)

A minimum of 3 years of weekly data is recommended.

14. What happens if brand data is missing for some products?

If only one product has brand metrics, long-term effects will be applied to that product only. The rest will use short-term logic.
This ensures model stability but means long-term equity comparisons across products may be uneven.

15. What’s the difference between Global and Halo media effects?

Both affect all products indirectly but are tagged and interpreted differently during setup.

16. Why can’t I view multiple impact types at once?

You can use the Combined option to see all effects rolled together, or switch filters manually for comparison.

17. Can I use this model for products with seasonal media bursts only?

Yes, but be aware:

• Long-term effects assume consistent media equity buildup.
• If media is too sporadic or brand equity decays quickly, long-term attribution may appear weaker.

You may need to adjust expectations or use the Observed view for more realistic phasing.

18. Is the Long-Term Effect Period fixed or configurable?

The preset is model-dependent - it highlights the window where ~90% of long-term effect is expected, based on decay parameters.