Blog

Blog

Blog

AI in Market Research: Balancing Machine and Human Collaboration

Consumer Insight specialists globally are in equal parts captivated and apprehensive about the impact of AI in market research. Almost 2.5 years since the first ChatGPT release, Artificial Intelligence is inarguably transforming every aspect of the research industry.

On my podcast, “Research Revolutionaries,” two industry leaders – Babita Earle, Zappi’s Managing Director International, and Caroline Frankum, Global CEO of Kantar’s Profiles Division – discussed the roles of human and Artificial Intelligence in the future of consumer research. Babita and Caroline reveal their take on:

  • The current state of AI in market research and what the rise of the machines means for the industry’s future
  • Specific opportunities and challenges in store for Chief Insights Officers and CMI teams
  • Where leading businesses are harnessing machine-driven data strategies
  • How to maximize the power of ML generate game-changing insights.

Article sections:

The Buzz Around AI: A Double-Edged Sword for Insights Teams

We all know that AI is the talk of the town in market research circles. A “word bingo” game at any recent industry conference would have seen AI as the outright winner! Artificial Intelligence also dominated Greenbook’s 2024 Grit Report. but how are these many conversations translating to action for insights leaders?

The tantalizing potential of AI in research and insights is undeniable: AI-powered tools can automate tedious tasks, crunch massive datasets quickly, and uncover patterns that human analysts might overlook. In theory, Machine Learning can liberate researchers to focus on higher-level strategic work and enable faster, more responsive research processes.

However, AI’s hype is matched by chronic anxiety about its implications for the future of market research, including:

  • Job Security: Will AI render human researchers obsolete?
  • Data Bias: Could biased AI models lead to skewed or misleading insights?
  • Insight Quality: Can we trust the outputs of AI?
  • Research Rigor: Will overreliance on AI keep research standards in mind?

AI is here to stay. We’re already seeing how emerging technologies will reshape our industry in the years ahead – and we must change our mindset and adapt our practices accordingly.

Babita asserts, “If you have an innovative mindset, you’ll be comfortable with the potential discomfort AI will bring.”

Ray Wang, Founder of Constellation Research is even more bullish: “Organizations that fail to implement AI-driven smart services will fall behind. The value of insights and the ability to achieve situational awareness drives massive competitive advantage.”

Whilst the AI revolution may be daunting, it’s also an exhilarating opportunity for researchers who are willing to embrace change and push the boundaries of possibility.

AI Applications: From Predictive Insights to Fraud Detection

McKinsey reports that although 55% of Chief Insights Officers have already adopted some AI technologies, only a minority have fully embedded AI in insights practices, and even fewer are leveraging it to create substantial business impact. 92% of insights leaders are planning to increase investment in AI by 2028 – but with leaders struggling to apply the capabilities of technology, what does embracing AI look like in practice?

AI-Powered Predictive Insights to Inform Product Innovation

AI is becoming increasingly useful in generating predictive insights from brands’ vast troves of data. CPG CMI teams are harnessing predictive analytics to answer business-critical questions such as:

  • Which new product or marketing ideas are most likely to resonate with consumers?
  • How might different consumer segments respond to an ad campaign or brand message?
  • What are the key drivers of brand performance and market share in each category?

“We’ve got quite a few organizations that have amassed a great deal of data on a platform that is standardized at scale, and it’s good data,” Babita reveals. “It gives us the ability to overlay AI to generate new ideas. But what does this enable us to predict for certain categories and certain brands?

“This predictive use case for AI is incredibly powerful. Imagine being able to forecast consumer trends, anticipate market disruptions, or test new concepts before they even exist – all based on the patterns and insights hidden in your historical research data.

“However, some early adopters on the buy-side urge caution.  “Context is often missing when you rely on AI for prediction, and results can often be ‘just weird’. A recent analysis suggested that ‘milk pizza’ might be appealing to our target market!  Without specialist curation, AI can be misleading and undermine the role of consumer insight.” How can leaders balance automated analytics output with the human intelligence needed to deliver valuable predictive insights?

Global consumer goods leader Unilever leverages AI to predict changing consumer preferences and informs innovative products that meet future consumer needs. Unilever’s psycho-analysis tool measures app-based interactions to define customer sentiment around new products and brand interactions. Improved understanding of consumer behaviors and motivations helps Unilever anticipate their future needs and has already seen new product launch success rate increase by 30%.

Knowledge Management in Data Democratization to Empower Organization-Wide Decision-Making

Gartner reports that 32% of leaders believe they have made incorrect decisions because of a lack of insight awareness. Democratizing data across any business is critical to aid understanding of analytics across teams, informing faster decisions and better outcomes.

In the past, research requests would require familiarity with the repository of past research studies. New technology-driven capabilities are enabling more intuitive searchability of research portals to offer users specific answers to their questions.  AI search is fast eliminating the time-consuming and labor-intensive process of trawling through multiple PowerPoint decks to find valuable data. A senior industry leader describes this as a game changer: ‘We are only in the test phase, but AI search functionality on our research portal has dramatically decreased the number of hours spent on research requests.  Test users get the exact answers they need fast.’

Leading knowledge management platform Strativo warns of the dangers of organizational memory loss. Gartner divulges that 47% of knowledge workers have trouble finding information that is relevant to their roles. Several studies show that organizational memory loss can affect job satisfaction, reduce efficiency and increase the cost of doing business.

AI is already offering significant advances in data access and searchability but crucially maximizing the utility of costly insight generation. In the future, we are likely to see a more seamless way of working between AI & humans in knowledge management systems as well as more cross functional integration to enhance efficiency.

The Role of AI in Combatting Fraud to Preserve Quality Data

The rise of sophisticated fraud attacks and “bad actors” infiltrating panels make combatting fraud even more vital for tomorrow’s insights leader. AI-powered tools can detect and eliminate fraudulent responses at lightning speed, protecting business continuity and improving data quality.

“Fraud isn’t going away,” advises Caroline. “It’s about how we stay one step ahead of the fraudsters. We need to make it so challenging for bad actors to get into our surveys that they just don’t bother.”

By combining cutting-edge AI with decades of expertise in panel management, leading research firms are guaranteeing truthful, reliable data to inform critical decisions – even as fraudsters become savvier and ever-more determined.

Of course, getting to this level of AI sophistication requires a strong foundation of clean, consistent, and well-structured data.

Synthetic Data: The Answer to Overcoming Age-Old Industry Challenges or a Banana Skin in Disguise?

Non sample alternatives are definitively gaining traction amongst research buyers. Greenbook’s 2024 Grit Report survey data suggests that twice as many buy-side researchers are looking at non sample options compared to 2022. Synthetic sample is arguably the most controversial aspect of AI application right now.

Key Considerations for High-Quality Synthetic Data

Synthetic data is only as good as the real-world data it learns from. To maximize its effectiveness and ensure fairness, the following factors must be carefully managed:

Avoiding Bias Reinforcement
Avoid amplifying biases present in the original dataset by carefully evaluating synthetic data.  Deploying fairness-aware techniques in data synthesis will create more equitable machine learning models.

Ensuring Representativeness
Training models on representative synthetic datasets is non-negotiable if you want reliable, unbiased, and effective AI systems. Synthetic data should always accurately reflect the target population to ensure applicability.

Promoting Data Diversity
A diverse dataset enhances the performance, fairness, and robustness of machine learning models. Incorporate variations across different demographics, conditions, and edge cases to avoid generalization.

Prioritizing Data Quality
High-quality training data leads to better AI models. The accuracy, completeness, and relevance of synthetic data directly impact the effectiveness of AI-driven insights and predictions.

Great AI relies on great data. Investing in high-quality, diverse, and representative synthetic datasets is essential for building ethical, unbiased, and high-performing machine learning systems.

It is the early days in synthetic data experimentation but its potential to enhance quality and speed of research is alluring.  Kathyrn Topp, Founder & CEO of Yabble asserts ‘by generating data that mimics real user behaviors and attributes without utilizing actual user data, researchers can overcome many of the privacy and ethical issues that plague traditional research methods.  Additionally, it can grant access to very niche and previously difficult to reach audiences. ‘

The UK Market Research Society (MRS) are exploring the use of artificially generated data to enhance market research via their Delphi Group.  The findings from this initiative are expected to be available in July 2025. Watch this space!

The Future of Research Careers: AI as the Insights Teams’ Champion

The inevitable prominence of AI raises existential questions about the future of research careers. Will intelligent machines displace human insight professionals?

The consensus from my conversations is that the research professional will continue to serve in a vital role for insights generation. However, CMI teams must adapt to work alongside emerging technologies.  In the next few years, AI will change how we commission, analyze, curate and share insights – leaders are already thinking about the shape of their teams and the skills and training needed to maximize the application of AI-powered insights.

One respected industry leader told me: ‘I don’t need to be an expert in every technological solution, but I do need to promote a culture of innovation and an environment where my team can experiment and learn’. As AI is native to Generation Z, training for this demographic will focus on implementing their existing knowledge and growing their expertise effectively. Teams of all backgrounds and groups will require continual training to match their own knowledge with AI’s growing capabilities and build a symbiotic relationship with their technologies to scale impactful insights.

Rather than being replaced by AI, researchers and insights professionals are fast adding “prompt engineering” to their toolkits, using AI as “co-pilots” to marry both human and machine intelligence. As specialists, we are becoming more adept at asking the right questions from our AI tools and interpreting technology-driven outputs through a nuanced human lens. The unity of people and machines will require researchers to upskill in areas like data science, analytics, and storytelling where AI cannot deliver contextual, engaging translation. The fundamental human skills of critical thinking, creativity and empathy will be increasingly harnessed in the career of the future researcher to turn automated analysis into relevant recommendations.

“We fundamentally believe our sector is a people business,” confirms Caroline. “We source information from people, we sell information to people. We are about reflecting the truly diverse world that we have the accountability and the privilege to serve.”

Insights teams are increasingly expected to do more with less, provide more definitive answers and support faster decision-making across their entire business. Rapid and effective AI deployment is a means to the analytics end: a dynamic tool for augmenting and scaling human expertise, but not a substitute for insights specialists.

Leonard Murphy, Chief Advisor for Insights and Development at Greenbook, summarizes: “Insights teams are legitimately questioning whether skills related to process and experience carry as much weight as they used to in today’s AI-driven world. Intuition, creativity, context framing, influencing, communication and sense-making aren’t just “soft skills;” they may be the non-disruptable foundation insights professionals use to thrive in the future.”

The real value of AI-powered analytics will come from a close partnership between human and machine. The most effective researchers of tomorrow will continually upskill and adapt to apply their own consumer understanding to automated analysis and translate complex insights into actionable business recommendations.

The potential of AI is uncontested – but most insights teams have yet to fully safeguard against the looming risks and resulting business damage of emerging technologies.

Biased data will always deliver biased results. Inaccurate insights inform flawed decision-making and even discriminatory outcomes if left unchecked. The USC Information Sciences Institute recently found that over a third – 38% – of data used by AI contained bias. Gender and racial bias are routinely common in consumer data sets, presenting not only a risk of discriminating against large groups of consumers, but the danger that diverse consumers will be under-served or even ignored in product development recommendations.

As an industry, we’ll need robust processes to audit AI models for bias, ensure diverse and representative training data, and maintain human oversight of machine-generated insights.  Researchers and brands will rightly demand to know the “why” behind the “what”: transparency and interpretability in AI research must also be prioritized. Transparency in AI involves developing systems that are explainable, accountable and available for human interrogation.

Finally, we’ll need to grapple with the broader ethical implications of using AI in consumer research. How can we protect consumer privacy through automated data collection? How do we ensure responsible use of AI for the benefit of consumers, not just the bottom line?

Ethical use of AI will require leaders to:

  • Create Transparent Processes: Practice and disclose comprehensive data usage policies and the workings behind AI decision-making
  • Adopt Privacy Safeguards: Collect only necessary data, anonymize it, and obtain informed consent from consumers
  • Mitigate Bias: Regularly audit AI systems for fairness and inclusivity
  • Promote Accountability: Establish clear mechanisms and responsibilities for AI-driven decisions
  • Encourage Regulation: Support policies that balance innovation with consumer protection.

Complex problems are never solved with easy answers. As an industry, we’re just starting to confront the risks and challenges of emerging technologies. The leaders of tomorrow will navigate the brave new world of AI-powered research with a new approach to existing challenges and a growth mindset.

Looking Ahead: How AI Will Transform Market Research and Improve Insight Team Effectiveness

What does the future hold for AI in market research?

“Growth never comes from comfort zones,” shares Caroline.  “Researchers will need to lean into the discomfort, ask tough questions, and experiment fearlessly to realize the full potential of AI.”

Researchers are already experimenting with automated technologies to speed up and scale up insights through:

  • Automated data collection
  • Machine-powered analysis
  • Knowledge management
  • Synthetic data collection
  • Predictive modeling and insight generation.

As technology advances at breakneck speed, the possibilities of AI in research will only expand – and the researchers of today are balancing machine efficiency and human expertise to anticipate tomorrow’s decisions.

FMCG giant Nestlé is leveraging AI for innovative market research and future product development. Nestlé’s GenAI platform gathers data on consumer preferences and conducts rapid market trend analysis across restaurant items, home cooking habits, social media and a variety of online sources to accelerate product innovation. Nestlé uses AI to analyze vast amounts of data in short time frames and supports automated analysis with strategic human knowledge, to identify emerging consumer needs and validate new product concepts for the most successful market launch.

Fei-Fei Li, Co-Director of the Stanford Institute for Human-Centered AI predicts: “The future of Artificial Intelligence is not about man versus machine, but rather man with machine. Together, we can achieve unimaginable heights of innovation and progress.”

The successful integration of AI will require a careful unity between machine efficiency and human expertise, between speed and quality, and between innovation and responsibility. If we can become revolutionary innovators to meet the demands of the AI-driven world and the consumer needs of tomorrow, our industry’s brightest days – and better opportunities to drive business-critical decision-making – lie in wait for us.

Revolutionize Your Research: The Powerful Partnerships Behind Game-Changing Insights

Delineate help the world’s largest consumer brands to conduct faster research and deliver more accurate consumer insights that influence business-critical decisions. Coca-Cola, Ancestry.com and Nomad Foods employ our brand tracking and campaign evaluation technologies to mobilize the data-driven insights that create a competitive advantage.

Discover how Delineate can transform your insights capabilities to meet the consumer needs of tomorrow.

Join our Newsletter