AI trends for insights: From promise to reality

In recent years, Artificial Intelligence (AI) has been heralded as the ultimate transformative force set to reshape the insights sector, and the world. The result of this promise has been a rushed application in all its forms, the development of new AI-based solutions, and basically every product somehow receiving a stamp of “infused with AI.” Surely clients must be demanding all this, right?

Esomar’s Global Users & Buyers of Insights 2025 report shed some light on this. Out of all – potentially unfulfilled – demands from agencies, the most salient one with a quarter of all responses was that of activation of insights, which is a wish to receive increased guidance and advisory on how to implement the generated insights within the company in order to navigate this complex and fast-changing world. One had to look hard to find a new category, sitting below with 7% of mentions, of clients requesting more technology use, AI and innovation from their suppliers.

Interestingly, users and buyers declared having AI implementation for research at the top of their agendas and considered it to be their main short-term concern, a decision presumably driven from the management board of companies as part of their strategy.

Looking elsewhere, Esomar’s Global Prices Study for 2025 dug further to try and understand the scale of AI’s impact in the agencies’ research services. At the time, the impact on both cost and efficiency was still minimal. Incidental third-party research in the broader economy seems to point in the same direction, with AI taking longer than expected to translate impact into gains.

However, the push remains, and most companies have AI implementation front and centre in their development plans. A working hypothesis, supported by the strong investment in the most technology-enabled side of the industry, points towards a different kind of stakeholder driving a substantial portion of the demand: investment companies seeking to capitalise on the valuation boost that markets are giving to companies that embrace AI. Being investors of both the client as well as the agency sides of research, this could help explain this push despite the lack of obvious client demand. After having cooled down in recent years, Esomar’s Global Top Insights Companies 2026 ranking has shown a recent uptake in industry investment.

Are we yet to see AI’s full potential?

One of AI’s clear abilities is to support researchers in their daily activities and, according to Esomar’s recent consultation on AI Trends, it appears to be how the average professional is using it today: text and image analysis, and quality control. This may appear to contrast with the perception of a whole array of companies that claim to have seamlessly integrated AI in their systems, the extraordinary capabilities displayed at conferences, the firms at the industry forefront that receive the awards or the global assessment of AI development such as the AI adoption curve developed by some key industry players. The daily life of the average researcher, however, may not be as bombastic.

AI’s potential is becoming clearer, though. That once abstract promise is now gaining a agreed vocabulary, the different techniques and applications are being evaluated, and with this a much more realistic view of AI’s capabilities and limitations is emerging. This much became manifest during the – ongoing – Community Circle series organised by Esomar to discuss, among others, evaluation methods on several types of synthetic data.

These efforts and open-floor discussions may be particularly relevant for small and medium enterprises, which constitute the majority of the industry’s players but recognised in Esomar’s AI Trends study their struggle to adapt to a fragmented regulatory AI landscape with their limited resources.

Normalising the trend

Regardless of AI’s ultimate potential to transform the industry, the current reality is the desire to embrace it, and with implementation comes governance. Esomar’s exploration of AI Trends uncovered a worrying governance gap in its first iteration in 2025 where, while AI had already been adopted by most respondents, only around half had some sort of governance framework such as internal guidelines or policies. Less than a year later, this gap has almost completely closed, showing the industry’s high regard for responsible AI development.

Some of AI’s perceived risks remain at the top of the AI users’ minds, such as bias, discrimination, model accountability and data privacy. Those familiar with the ICC/Esomar International Code of Conduct may grin at this finding, since these were included in the lists of risks to address in the revision of the document ultimately approved by the Esomar membership in 2025. This strong alignment with the industry’s realities demonstrates the suitability of the ICC/Esomar International Code to effectively guide self-regulation in the sector and promote responsible AI development.

Being a central theme for its strategy 2030, Esomar is currently working to set up and coordinate the creation of an expert working group focused on AI. The result will focus on three main pillars: professional standards to guide AI development, case studies to show its capabilities and properly evaluate it, and lastly showcasing AI at events and awards. Expected for Esomar’s Congress in Valencia, stay tuned to hear much more about Esomar’s efforts to support the industry in its AI journey.

Ultimately, the time freed by employees through AI adoption can be fully invested in the vital added value of insights activation. It remains up to companies, however, to decide whether they want to focus on facilitating the execution of routine work through AI as a cost-saving measure and shaving FTEs off, or if they want to use this freedom to empower their workforce to work on the activation of those insights as demanded by their clients.

Follow the ongoing discussion at the interactive Community Circles focused on AI so you can pour your opinion, ask questions and create global consensus around this topic.