Our latest Campaign for Real Innovation (C4RI) event ‘Agile Research in Action’ brought together client-side insight leaders and innovation specialists to discuss how AI and agile research are reshaping the industry. The conversation focused on a practical question: how can businesses use new tools to move faster, work more efficiently and make better decisions, while still protecting quality, trust and human judgement in high-stakes commercial environments.
Insight teams are under pressure to move faster
A clear theme from the session was that insight teams are being asked to do more, more quickly and often with fewer resources. Speed matters, but so does flexibility. Teams need approaches that can adapt when products change, priorities shift or stakeholders need answers quickly.
AI offers real potential, but it is not a silver bullet
AI was described as both an opportunity and a challenge. The hoped-for benefits included greater speed, lower costs, better use of data, scenario planning and stronger forecasting. At the same time, attendees raised concerns about accuracy, bias, hallucinated answers, governance, compliance and the effect on jobs and skills.
A recurring theme was that these limitations are often magnified in global work. AI does not perform equally well across all markets, languages and cultural contexts, which creates real risk if outputs are applied without local validation. As a result, AI-led insights need careful interpretation and validation, particularly when results are being used to inform decisions across regions. The discussion repeatedly returned to the same point: AI can add value, but only when combined with human judgement and local understanding.
Human expertise remains essential
One of the strongest messages from the event was that the future is not AI instead of people, but human-powered AI. While AI can process large volumes of data, automate tasks and surface patterns quickly, it does not replace experience, instinct, challenge or interpretation.
For research and insight teams, that human layer is still critical in deciding what findings mean and how confidently they should be used, especially when translating insight across markets with different cultural, regulatory and behavioural dynamics.
Innovation is becoming more practical and more targeted
The event highlighted three broad areas of innovation: agility, AI and better understanding of real-world product experience. Rather than pursuing technology for its own sake, the emphasis was on solving real client problems, whether that means faster product testing, smarter use of existing data, richer in-the-moment consumer feedback or better ways to model likely outcomes before committing bigger budgets.
Examples showed how this can work in practice
Several practical examples were used to bring the discussion to life. These included faster product testing to support quick stakeholder decisions, such as SprintAI; app-based methods like Intercept that capture consumer experiences in the moment; AI-supported qualitative approaches, including FortifyAI, that can increase scale while maintaining research structure; and modelling tools designed to show when existing data may already be strong enough to guide a decision.
Across all of these, the emphasis was on using tools selectively and responsibly, ensuring outputs are grounded in real consumer evidence rather than relying on automation alone.
The biggest barriers are not just technical
During the workshop discussion, attendees identified familiar barriers: limited time, lack of internal knowledge, mixed levels of confidence across organisations, compliance concerns, poor data readiness and uncertainty about which tools to trust. Stakeholder buy-in emerged as a major issue.
For many teams, the challenge is not only finding the right use case, but also ensuring tools, data and outputs are appropriate for different markets, and building the internal confidence needed to adopt new approaches responsibly at scale.
Progress will come through small, practical steps
Rather than trying to transform everything at once, the discussion pointed to a more realistic route forward: start with focused use cases, test ideas on a small scale, measure what works and build confidence from there. That could mean piloting a lower-risk application, improving prompting and training, validating AI-led outputs against real consumer evidence, or using initial efficiency gains to create time for deeper learning and experimentation.
What this means for clients
For clients, the key takeaway was clear: AI is already changing the way insight teams work, but successful adoption depends on using it with purpose, discipline and human expertise. Clients are looking for faster, more agile and more cost-effective ways to generate confidence, but they also want methods they can trust. The opportunity lies in balancing innovation with rigour.
How to move forward
If your team is exploring how to use AI or agile research more effectively, the most valuable next step may be to identify one practical challenge and test one focused solution. Start small, validate carefully and build from there.
To discuss how this could apply to your brand or category, please get in touch with Kate Binner at [email protected].
Stay In Touch