AI-Powered Insights Conference 2027

08:45 Registration & coffee

09:45 Welcome from the Chair
Richard Owen, Head of Innovation, Firefish

10:00 Beyond the pilot: what it takes to scale synthetic customers
Synthetic customers are easy to pilot. Scaling them into a capability the business relies on requires buy-in, ownership, and sustained engagement.
This panel brings together E.ON Next, Leeds Building Society and No7 Beauty Company who have all made that leap. Each speaker will share the internal case that won first buy-in, the moment adoption nearly stalled, the unglamorous work including training, champions, quick wins, and where they see it heading next.
A session about AI impact and change management.
Chair:
Kylie Fitzpatrick, Business Director, Verve
Panellists:
Matthew Bickham, Head of Insight, E.ON Next
Steven Darby, Research Manager, Leeds Building Society
Rosie Hudson, Global Insights Manager, No7 Beauty Company

10:30 How agentic workflows are speeding time to insight at sky
Sky has a rich asset of brand, campaign and customer tracking data, but accessing the insight can be time-consuming. Working with Bayes Price, Sky has evolved from automated monthly dashboards to an agentic, conversational approach that puts data directly into the hands of a wider marketing audience. Teams can ask questions, build dashboards and reports, and get answers quickly, while agents proactively flag shifts in the data and offer prompts to diagnose them.
This session shows how Sky is moving from self-serve data access to an intelligent system that helps teams find, understand and act on insight faster.
Matt Gibbs, Managing Director, Bayes-Price
John-William Awbrey, Head of Brand & Campaign Insights, Sky

11:00 Morning refreshments

11:30 Joined-up insight: see the full consumer picture
Traditional research methods are increasingly challenged by fraud, bots, declining engagement and more guarded respondents. Meanwhile, consumers generate vast amounts of behavioural data through reviews, searches, social conversations and online interactions.
This session explores how AI can connect these fragmented data sources into a unified insight ecosystem, enabling brands to move beyond isolated research outputs and build a more continuous understanding of consumer attitudes, behaviours and experiences to support better business decision-making.
Paul Van Gendt, Chief Operating Officer, I-Genie.ai
Welmain Paige, Head Of Insights & Strategic Planning, Elida Beauty

12:00 Substance at speed: building the insight function of the future
As barriers to consumer understanding fall, organisations can generate insight at unprecedented speed and scale. Thanks to AI, rapid insight is quickly becoming the baseline. But when speed becomes abundant, how do we ensure we don't lose the substance that underpins high-stakes decisions? And what does that mean for the Insight function itself?
The panel will explore how Insight functions can adapt to this new reality, balancing the opportunities of speed and scale with the depth, rigour and judgement that consequential business decisions demand.
Chair:
Jamie Allsopp, Partner, CMRS, PA Consulting
Panellists to be announced

12:30  When the judge is an algorithm: catching hallucination and bias in AI-generated insight
LLM-as-a-judge is now standard in evaluation pipelines, but the judges have their own biases: position bias, verbosity bias, and self-enhancement, to name a few. When an algorithm grades AI-generated insight, these distortions can quietly shape what counts as good research, letting hallucinated or skewed findings pass as quality. This session maps the bias taxonomy of LLM judges. It also shows how to benchmark AI-generated insight against human judgment, where the two diverge, and how to set up evaluation loops that catch hallucination and bias before they reach stakeholders.
Delegates will leave with a practical checklist for keeping research quality high.
Saurabh Kurjekar, Data Scientist, Meta

13:00 Lunch

14:00 Panel: Synthetic or Real? How to choose, evaluate and trust synthetic research

• Which business decisions can synthetic data be trusted for?
• How to assess the quality, validity and limitations of synthetic data
• Practical frameworks for deciding when to use synthetic or real respondents

Chair: Kate Owen, Customer Insight Lead, TfL
Panellists:
Michelle Yaiser, Chief Product Officer, Savanta
Felix Wallis, Head of Research, Artificial Societies
Lori Noel, Senior Vice President, Client Services, Burke
Dionne Gordon, Head of Research & Market Insights, Correla
Sarah DeCaux, Head Of Research Team & Lead Analytics & Insight Manager, Co-Op

14:40 How PZ Cussons turned insight knowledge into an AI-powered innovation resource
PZ Cussons set out to accelerate St. Tropez’s US expansion by understanding where to innovate, which consumer needs to prioritise and how to turn insight into opportunity faster.
The insight existed, but was fragmented. PZ Cussons worked with Zappi to create a connected insight ecosystem whereby specialised agents interrogate knowledge across different studies, methodologies and partners bringing accumulated consumer understanding into one environment that innovation teams can continuously explore and apply.
Hear how they built this connected ecosystem, how it works in practice, and what they’ve learned so far about turning fragmented insight into a living resource for innovation.
Will Cooper, SVP Enterprise, Zappi
John Gurd, CEO, Market Measure
James Mackenzie, Head of Insights – Global, PZ Cussons

15:10 How do you build trustworthy insights with AI?
Charles Allison, Research & Insights Lead at Conveo
Sarah Riding, Head of Market Research at ITV

15:40 Afternoon refreshments

16:05 Crumbs don't lie: Using AI video analysis to understand snacking behaviour
Consumer-generated video captures the key behavioural moments that drive snacking enjoyment and product preference but that are rarely reported by consumers. Yet coding large volumes by hand has been slow and difficult to scale. So pladis partnered with MMR to test whether AI-enabled video analysis could close this gap and provide a faster, scalable solution to identifying drivers of product preference.
This talk shares what was learned about dunking, and a bigger lesson about method. Hear how AI scales the observation, but the researcher shapes the lens and makes the final call.
Alexandra Kuzmina, Global Innovation Director, MMR Research Worldwide
Tom Holmes, CTI Senior Technologist, pladis Global

16:35 Session TBC

17:05 Sponsored drinks reception