Why AI

Technology has had a huge impact on market research, creating a new generation of faster and easier-to-use tools that help brands discover what...

Why AI: How machine learning enhances your consumer insights

Explore Kantar’s Expanding Portfolio of Artificial Intelligence and Machine Learning Capabilities

Technology has had a huge impact on market research, creating a new generation of faster and easier-to-use tools that help brands discover what consumers truly think, while drastically shortening the time between putting a survey in-field and retrieving valuable consumer feedback.

Deliverables that once took weeks can now be completed in as little as a few hours. Automation has also streamlined process by making survey research more repeatable – brands can get feedback on innovation concepts or test creative by inputting a few simple parameters into a templated online survey, rather than recreating the survey each time.

The combination of our technology and longstanding data collection can reduce your time and spend.
Using Machine Learning to create prediction algorithms allows us to better understand data that will enhance your brand's future.

Impact of technology on market research

AI is used in many Kantar offerings, including our Entertainment on Demand and ComTech business units, to help our marquee clients better anticipate consumer behavior and understand the propensity of customers to abandon their purchase decision.

The aim of the propensity model is to measure how likely consumers are to cancel or stick with a product that they had acquired. We can compare and contrast against their competitors to see how the client base is made up.

An additional benefit is the ability to manipulate sample size.

As a result, increasing the sample size significantly allows for further, more overarching, analysis. For example, what areas make ‘churners’ leave the brands.

Kantar’s data is longitudinal

We collect data from the same respondents every period and can track usage changes, satisfaction drivers, etc., over time.

This allows us to assess the accuracy of the model over time and retrain the AI to improve results.

Look at a scale of 0-100 on how likely the consumers are going to churn/remain loyal. It will be able to show the client how ‘healthy’/’unhealthy’ your base is and what proportion it makes up vs. your competitors.

Microsoft Customer Story-Kantar uses Azure Machine Learning to help streaming services drive market growth)

In a nutshell...

Purchase

Usage

Satisfaction

Consideration

Loyalty

Sales
Tariff
Details
Retail channel & Name
Reasons for purchase
Key Influences
Demographics
Marketing touch points
Media recall
retail experience

Usage
Bill payer

Satisfactions
Recommendation/net promoter score
Experience factors

Consideration
Preference
Intention to purchase
Brand imagery
Decision factors
Preferred features

Loyalty/churn
Switch/churn
Drivers of loyalty
Drivers of churn

...we train the model based on past behaviors and apply it to our current base.

On Kantar Marketplace: Link AI

Link AI uses machine learning to gain creative insights across larger sets of ads without having to worry about traditional cost-per-test expenses getting in the way.

A rapid go/no-go video content measurement solution that is designed for fast pace and high-volume ad development.

Kantar Marketplace is an automated market research platform designed for insights professionals, marketers and agencies who want to test, learn and move faster. It is the only market research platform with solutions that have been independently validated to predict sales and brand growth.

Learn more about Kantar Marketplace

Kantar’s AI solutions for creative measurement complement survey-based testing

Link AI is built from Link, the world’s largest normative advertising database

Market Research Platform | Kantar Marketplace

Behavioral and creative predictions

Eyesquare data across Kantar solutions contributed to the model set for behavioral predictions (metrics such as Skip Time and Interaction)
Virtual Reality & Eye Tracking

Link for digital and TV videos and Kantar surveys were used to train the machine for creative predictions.

Metrics such as Power, Impact, Branding, Enjoyment, Persuasion, and Affinity
‘Action Likelihood’ is a composite metric of creative predictions that relates to a user behavior
Link AI feature generation models are used to expand the breadth of ad features

Case Study: Digital Creative Intelligence for Unilever

Case Study: Link AI enabled creative effectiveness data from over 11,000 ads to be analyzed in a short period of time for Google

Improving an ad’s creative quality from “average” to “best” leads to 30% increase in ROI

Additional Case Studies:

Conversational AI

Conversational AI for Brand Guidance taps into a part of our DNA which is all about telling stories

Our AI-powered bot dialogues are carefully designed:

To let a person’s story about their experiences and preferences in a category unfold as naturally as possible
Through a guided, exploratory one-on-one conversation on a messenger platform.

Firmly embedded in our understanding of people and how brands grow, we offer two AI-powered scaled bot solutions:

Conversational AI for Imagery: Triggered by shifts related to brand imagery, geared to identify ways to own and grow specific brand benefits
Conversational AI for Equity: Triggered by shifts in brand equity metrics, geared to identify opportunities to unlock growth

True business impact through AI-powered conversational design at scale

When combined with artificial intelligence, these stories:

Directly inform your brand strategy by leveraging the right brand growth benefits.
Sharpen, refresh, and amplify your brand articulation.
Create more powerful brand communication that connects more deeply with people.on a deeper level.
Reveal opportunities for growth, potentially also through innovation.

How does it work?

Neuro/Facial Coding/Emotion AI

Pre-testing ad campaigns will anticipate people’s emotional response

Our automated research solutions allow marketers to get fast, and actionable feedback on how people respond to their advertising.

Whether testing before or after launch, it is critical that the test accurately reflects the breadth and depth of people’s emotional response.

Better identifying emotional responses

To ensure that expressions can be recognized irrespective of gender or ethnicity, our facial coding partner Affectiva has re-trained their A.I. to better identify the emotional response to advertising using Deep Learning for greater accuracy and consistency across demographic groups.

Based on 5 billion facial frames, 90 countries, and 11 million faces the coding system can now recognize a smile with an accuracy of 95% or higher for each of five groups:

Africans
Caucasians
East Asians
South Asians
Latins

Better representation means less chance of missing a minority response.

Focus AI

Through various ways, including Kantar’s proprietary methods and third-party relationships, the Advertising Insights platform collects and categorizes millions of ads in all markets across all media types.

Focus AI enables clients to generate data stories in just moments, highlighting key trends in ad spend shifts, but also and indicating what is likely attributed to the shifts in trends changes (ex: Executive Summary, Conclusions, Change Drivers).

It goes beyond just data reports, and sheds light on what is contributing to the changes and what advertisers should must pay attention to.

Insights drawn from Focus AI

Contextualize the key drivers of change at a macro level.

How do Advertiser/Brand Share of Voice rankings change over time?

Period over Period Changes:

Advertising activity at the Advertiser/Brand level impacting macro-level change in your custom-defined category
Advertising activity at the Media and Media Type levels impacting macro-level change
Advertising activity at the National and Local Media and Media Type levels impacting macro-level change
New and Existing Advertisers, Brands, Products & Creatives
New advertising influxes having an impact on current period spending

To learn more about the data stories you can create with Focus AI, please visit: https://info.kantarmedia.com/you-dont-need-data

Emerging Trends

Emerging Trends, the first cross-category search prediction tool

Consumer brands can apply Kantar’s Emerging Trends toolkit equipped with top of the line artificial intelligence and advanced analytics using:

18 years of search data
Thousands of categories
Close to two hundred markets
Billions of data points

The approach is machine led yet human-guided based on Kantar’s understanding of global trend cycles.

Understand the market

This is a completely new way to understand shifting consumer demand globally and across categories, with speed and at scale. Kantar’s proprietary toolkit identifies, classifies and quantitatively visualizes candidate trends across thousands of categories across all languages and all markets. We also explore if candidate trends are likely to stick or fade to ensure relevancy.

Example trends that consumer brands will be able to identify include ingredients, products features, needs, moments, occasions and sustainability movements as well as understanding which retailers, brands, influencers and celebrities could trend. Understanding if a trend will likely travel across a market or a category is something that consumer brands have been asking us to achieve for the last couple of years. We have researched different approaches and different data sources, but the scale just wasn’t there. Emerging Trends finally provides marketers with more ammunition to make the right investment decisions, using the power of artificial intelligence and advanced analytics. They will be able to address tomorrow’s consumer needs today.

Stella Wilson-Haffenden, Product Development Director for Digital Analytics, Kantar

The combination of human and machine intelligence (Trends 3.0) is the new normal when it comes to successful innovation and brand strategy. I hope we can join you in your journey to accelerate growth through big data and AI. Emerging Trends is the result of deep understanding data and analytics combined with the creativity, deep knowledge on trends and lateral thinking of our diverse team. I am sure you will find it ground-breaking.

Cynthia Vega, Global Product Director, Kantar

Find out more:
https://www.kantar.com/expertise/analytics/innovation-analytics