We’re excited to introduce the Augmented Data Programme—a transformative leap in research methodology that combines advanced statistical modeling with collected insights. By integrating statistically generated data, this program empowers smarter, more impactful decision-making.
Why do we need augmented data?
Surveying brand performance across thousands of brands poses unique challenges. To ensure statistical representation and reliability, we’ve developed a systematic approach to optimize respondent input without compromising data quality.
Initial approach: streamlined sampling
Several times per year, we collect critical performance data on thousands of brands across 40+ countries, surveying individuals aged 0 to 65 years old.
Each respondent is presented with approximately 60 brands, but expectations for brand familiarity vary greatly. After all, you wouldn’t expect a 7-year-old boy to know, love, and desire the same brands as a 24-year-old female adult!
However, even among respondents familiar with many brands—on average, recognizing 24 of the brands presented—requesting detailed responses for all brands would be impractical. This could overwhelm participants and potentially impact the accuracy and reliability of the data collected.
To address this, we implemented a looping mechanism that ensures respondents focus on a manageable number of brands, specifically 13 brands selected through a low-bucket statistical procedure. This method prioritizes brands the respondent is familiar with while meeting remaining survey quota requirements.
This targeted sampling ensures respondents provide thorough answers for a manageable subset of brands. Aggregating data across respondents then creates a complete picture of brand performance.
- Proportional reliability: Lesser-known brands receive higher interview rates (IR) to ensure sufficient data coverage and statistical base.
- Robust insights: Every brand reaches a respectable observation count, enabling reliable analysis of key metrics such as affinity and other performance levels.
While this method has been effective, the evolving needs of modern market research demand even greater precision and depth—particularly for niche groups and lesser-known brands. That’s where our new AI-enhanced approach comes in.
Our new approach: AI-enhanced insights
Leveraging cutting-edge statistical AI modeling, we address gaps in collected data by generating synthetic responses. This innovation provides a comprehensive, 100% dataset for every brand, seamlessly combining direct responses with AI-calculated projections for richer and more actionable insights.
Key features of the AI-enhanced approach:
- Adaptive interview rate allocation
The AI dynamically adjusts interview rates across respondent segments by learning from demographic and survey data mapped to the full brand list. This ensures proportional coverage for lesser-known brands and underrepresented groups, improving data balance across all segments. - Pattern extrapolation across segments
By training on adjacent respondent segments, the model extrapolates behavioral patterns and preferences. This triples the effective sample size for smaller or niche groups—validated through rigorous parallel testing to maintain reliability and accuracy. - Maximizing statistical efficiency
This methodology enhances data reliability without increasing respondent burden. By improving sample representation in underrepresented segments, we achieve comprehensive insights while preserving efficiency.
Unmatched benefits of the Augmented Data Programme
- Amplified data depth: By seamlessly combining synthetic data with real collected data, under-sampled respondent groups experience a threefold increase in representation, delivering deeper, more nuanced insights across diverse audience segments.
- Unparalleled accuracy: Rigorous testing of our synthetic data—which enhances the information collected directly from respondents—demonstrates that 86% of all brands achieve data accuracy within ±2 points of their total collected affinity scores, with 70% achieving an impressive ±1 point. This ensures confidence in every decision based on our data.
The table below illustrates, with just a few examples, that the augmentation process has minimal to no effect on overall affinity percentages. This demonstrates the effectiveness of the Augmented Data Programme, ensuring it preserves the integrity of the current data.
- Sharper granular insights: Developed in collaboration with AI experts Pulse Partners and powered by our proprietary BrandTrends dataset, this program offers a new level of analytical depth. By delivering precise, brand-specific insights, it equips you to identify untapped opportunities, refine your strategies, and optimize outcomes with unmatched precision.
Why this matters
This breakthrough methodology tackles the inherent challenges of analyzing diverse audiences across thousands of brands. By improving representation for niche groups and lesser-known brands, we deliver more precise, actionable insights tailored to every stakeholder’s needs.
Our commitment to innovation and excellence in understanding brand performance is at the heart of this approach. Whether identifying opportunities, refining strategies, or making high-stakes decisions, this enhanced methodology ensures every brand gets the attention it deserves.
