From data sorting to competitive edge: How large-scale AI models are powering next-gen insights

Advanced AI models can now process and interpret vast quantities of consumer data, marking the dawn of a new chapter in the commercial adoption of Large Language Models for market insights. General-purpose AI models are widely available, whereas advanced systems that combine deep learning with extensive proprietary datasets are stepping up to a league of their own to deliver next-gen insights.  

According to Kantar’s recent report, GenAI in Marketing: Fear or FOMO?, AI-driven insights are now a key decision-making factor for brands, with 57% of marketing professionals viewing them as fundamental to understanding consumer behavior (BrandZ, 2024). Indeed, 62% of those surveyed are set to increase their investment in AI thanks to advanced solutions now capable of predicting market shifts, evaluating advertising effectiveness, and identifying growth opportunities quickly and accurately.

For smart, data-driven brands listening to the industry headwinds, this technical leap means faster insights and more accurate predictions for better decision-making that drives growth.

From Early Data Analysis to Advanced Predictive Models

Market research analytics has progressed phenomenally since the 1980s from multivariate statistics and algorithmic forecasting in the early years to machine learning at scale by 2013. Fast forward to today where advanced AI models process billions of data points across multiple formats, combining behavioral patterns, attitudinal insights, and performance metrics to predict market trends with unprecedented accuracy.

This technical progression is evident in solutions like ConceptEvaluate AI, which analyses nearly 6 million consumer product evaluations to forecast market success. The system processes vast quantities of historical concept testing data, validated against real-world historical outcomes, such as sales and penetration via Kantar Worldpanel data.

For instance, when Iceland Foods aimed to pioneer AI-created ready meals, they used ConceptEvaluate AI to examine concepts for their new wellness range. The AI-powered evaluation helped Iceland identify the most promising products and energy-focused benefits, guiding their product development strategy and accelerating innovation while reducing market risk.

Advanced Models Plus Processing Power for Precise Insights

AI models and their capabilities now center on processing power and sophisticated data architecture. Kantar's AI assistant (KaiA) connects multiple proprietary datasets through Large Language Models to answer complex marketing questions in natural language. Users can ask "Why is my brand power declining?" or explore competitive metrics, receiving instant insights from Kantar’s wealth of consolidated structured and unstructured marketing, consumer, and sales data. A leading global brewer with over 500 brands experienced this first-hand when they faced mounting challenges with time-intensive data analysis and rigid tools that hindered quick insights.

Behind KaiA's natural language interface lies an intricate technical framework of computational infrastructure and sophisticated data architecture that processes multiple data streams simultaneously. For the global brewer, this allowed their brand teams to perform advanced brand equity analysis through a chat-based interface, while the system automated previously manual quarterly reports.

The brewing company's adoption of KaiA illustrates the processing power required for enterprise-scale AI solutions. In maintaining data quality across vast datasets, KaiA handled more than 14,000 questions from 120 users while automating 17 report types. The Senior Insights Manager highlighted, "Instead of manual tasks, teams can focus their time and energies on the things that really matter: the ‘so what for the business?’ questions."

This processing power extends across Kantar's AI solutions. LINK AI, for example, processes 35 million consumer responses across 250,000 ad tests to deliver creative effectiveness predictions within minutes. When LINK AI processed 11,000 ads for Google in under a month, it demonstrated how AI architectures can scale rapidly without compromising accuracy.

Developing Continually Learning Multi-Modal AI Models

AI models are fast integrating across the marketing lifecycle. LINK AI, for example, moves beyond finished ad testing to guide creative decisions from early concept stages. TrendAI's market analysis combines with LINK AI's creative assessment to deliver compelling campaign insights, while ConceptEvaluate AI adds innovation testing capabilities.

Looking ahead, AI models are refining natural language processing, improving pattern recognition, and developing intuitive interfaces between users and complex datasets. For example, Text2Topics analyses extensive open-ended survey responses and streamlines qualitative data to help clients identify patterns in large datasets and better understand consumer sentiments. Similarly, KaiA tackles complex brand tracking queries through natural language conversations, instantly converting unstructured questions like “Why is my brand power declining?” into precise insights drawn from consolidated marketing data.

Advanced Reasoning for Clear Decision-Making with Next-Gen AI Models

The next generation of AI insights platforms puts sophisticated analytics at your fingertips. By combining advanced AI processing power with decades of market research expertise, these tools help brands accurately predict market shifts and quickly spot opportunities. The result: actionable insights that drive an inimitable competitive edge that competitors can’t readily replicate.