Kantar and Realeyes have achieved another significant milestone by winning the I-COM Data Creativity Award in the ID Resolution Category. This prestigious recognition underscores their latest innovation in Verify, a lightweight user verification technology, which address critical issues in the rapidly changing landscape of sample quality.
In winning this award, Kantar and Realeyes demonstrate the industry's potential to innovate and improve, setting a new benchmark for data integrity and positive business impact.
The Industry Problem
The surge in survey fraud presents a growing challenge, our industry report shows that a third of surveys are low quality: infiltrated by deceitful participants or inaccurately qualified respondents to duplicate profiles and automated bots from the open exchange. This pervasive issue directly impacts data quality, casting doubt on the reliability of essential business insights.
Over recent years, the incidence of fake surveys has tripled, yet quality issues remain inadequately addressed. Sample buyers have traditionally emphasized lower prices and quicker turnaround times, leading suppliers to prioritize speed and cost over data quality. Considering that $1 trillion in business and advertising decisions hinge on the accuracy of market research, this situation is unacceptable.
Kantar’s Commitment to Quality
As a leading authority in marketing data and analytics, Kantar remains committed to delivering the highest quality data via their proprietary Premium Panels by leveraging decades of expertise managing panels and advanced, proprietary machine learning anti-fraud tech.
Their investment in cutting-edge AI technology strengthens panel quality and combats fraud across third-party panel sources. This unwavering dedication to maintaining the highest standards of data integrity not only distinguishes them from their peers but also solidifies their partnerships with over 90% of the world's major advertisers.
Addressing Fraud Head On
As technology evolves, the landscape of deceptive data and fraudulent practices continues to expand, ranging from sophisticated bot farms to intentional and inadvertent duplicates. These intricate challenges call for robust and innovative solutions.
Realeyes' Verify harnesses the power of lightweight facial verification to combat bots and user fraud with superior effectiveness compared to traditional methods like CAPTCHA. Verify is adaptable, reflecting real user behavior globally without needing personal identification from panel respondents. This computer vision approach enhances data quality while safeguarding respondent anonymity.
The ability to detect and eliminate fraudulent activity on the fly opens up the opportunity for the audience panel testing industry to ensure their data integrity, providing accurate, high-quality samples every time.
Improving Respondent Experience
Fraud concerns has led to increased security checks, putting the burden on the respondent experience that can lead to drop-offs. Verify removes this friction by enabling respondents to confirm that they're real and not a duplicate user, simply by using the camera on their device.
Positive Business Impact
With Verify, Realeyes are looking to establish industry benchmarks, reduce data fraud, improve measurement effectiveness and customer experience for all round business impact. This accolade from I-COM highlights how Realeyes and Kantar are leading the charge for reliable and trustworthy sample data.
Data Quality Pledge
We believe that reliable data quality has to be the central pillar of the consumer data ecosystem and panel survey industry. Realeyes have created a Data Quality Pledge as a commitment to:
• Transparently publishing indicators of quality
• Ensuring all staff, and suppliers are trained in data quality best practices
• Sharing examples of bad practices
Companies such as Kantar, Flowers Foods, and SAGO have signed the Data Quality Pledge to contribute to greater transparency around survey quality. Every pledging company gets access to see the quality of their sample compared to industry average.