B2BVault's summary of:

How Kellanova uses AI to predict creative performance and drive KPIs

Published by:
Marketing Dive
Author:
Chris Kelly

Introduction

Kellanova is using AI to measure how ads perform before launch, helping brands like Pringles and Cheez-It boost ROI with data-driven creative choices.

What’s the problem it solves?

Marketers often guess which ad designs will work best, wasting money on poor creative. Kellanova’s AI tool removes that guesswork by predicting outcomes with high accuracy.

Quick Summary

Kellanova worked with Vidmob and MMA Global to test AI-powered “predictive impact scoring.” By analyzing over 443 ad assets and billions of impressions across Meta platforms, they identified 19 creative factors most linked to strong results, like showing brand logos early, using positive emotion, or adding text overlays.

The AI achieved 83% accuracy in predicting three-second view-through rates, doubled performance improvements, and drove an 11% ROI lift. Unlike broad creative rules, the findings are tailored to Kellanova’s brands, making them a unique competitive edge.

The process isn’t about rigid formulas. Instead, it gives creative teams a data-backed “recipe” to guide future ads. Kellanova now applies this scoring system across multiple platforms and funnel stages, updating criteria for Facebook, TikTok, and more. The system creates a feedback loop where agencies are held accountable for meeting AI-informed creative standards, and results are checked quarterly.

Key Takeaways

  • Nearly half of creative decisions can be linked to measurable outcomes.
  • Predictive scoring forecasted ad performance with 83% accuracy.
  • Campaigns improved over 2x and ROI grew 11%.
  • Criteria like emotion, logos, and text overlays proved most effective.
  • Insights are platform- and funnel-specific, not one-size-fits-all.
  • Agencies are now held accountable to meet these AI-driven creative benchmarks.

What to do

  • Audit your past creative assets and identify patterns tied to outcomes.
  • Use AI tools to test ad effectiveness before spending on media.
  • Build platform-specific scoring systems (what works on TikTok may not on Facebook).
  • Apply insights differently at each funnel stage (brand awareness vs. conversions).
  • Hold creative teams and agencies accountable with clear scorecards.
  • Keep updating criteria as algorithms, audiences, and goals evolve.

The B2B Vault delivers the best marketing, growth & sales content published by industry experts, in your inbox, every week.

Consumed every week by 4680+ B2B marketers from across the world

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Explore the rest of the B2B Vault