The NOVA Partnership Quarterly Newsletter July 2024
Sensory bot launch, AI personas, and a critical look at trending tech
The Nova Team
28 Jun, 2024 | 5 mins
It is official – Sensory Bot will launch at the end of September!
At MMR, we’re not just purely building solutions; we are fostering the evolution of AI with value – and our Sensory Bot is no exception. Our bright-eyed AI toddler has been learning fast and we’ve been developing and training it with the same care and expertise that goes into every MMR project. Just like a lively toddler, it’s been exploring the world of tastes and textures, learning all the sensory subtleties which differentiate and define the product experience.
In exciting news, our Sensory Bot pre-launch beta pilot programme has been extended due to overwhelming interest and participation.
Here are the key updates:
- New pilots: Thanks to your enthusiastic interest, we're working closely with many of you across several new pilots to further refine how our Sensory Bot captures the sensory subtleties we know define the very best product experiences, setting a new standard for product experience research.
- New insights: Our pilots are increasingly revealing the extraordinary insights potential of the truly multi-dimensional data we can collect in this way – insights we know will matter to you.
- New launch date: Mark your calendars for the global launch of the Sensory Bot at the end of September.
We’re excited to share these developments with you and will provide a sneak peek of our pilot results soon. Your support on this journey makes all the difference, and we're grateful for your continued engagement and partnership.
Are AI personas a viable tool for early-stage R&D ideation?
Jeff Bezos is known for his practice of leaving one chair empty at the conference table, instructing his staff to consider the seat occupied by their consumer: the most important person in the room. But does it actually bring the consumer’s perspective into the room?
What if CPG brands had something more powerful, reliable, and cost-effective at their fingertips?
Innovation in the consumer-packaged goods industry requires constant ideation and testing, with reliable feedback being essential. Traditionally, this feedback comes from real consumers, but there are constraints: limited project scope, time restrictions, and the inability to address questions that arise later. This challenge prompted us to explore whether AI personas could accelerate early-stage R&D.
What is an AI persona? It is a chat interface to your data. We use segmentation data from consumer studies to feed into a machine learning model. We then use a dialog-based interface to interact with the AI persona which represents segments from the segmentation study. AI personas are grounded in responses of real people but have the knowledge and reasoning of the base LLM to draw from when it comes to suggestions. This combination allows us to extend our insights beyond the data that went into either ideate or ask it to react to new ideas and project concepts.
In collaboration with Gosh! Food, we embarked on an experimental journey to evaluate AI personas. In this experiment, we deployed a two-cell approach to test a family-oriented, plant-based concept from Gosh!. We compared reactions from 500 real consumers with outputs from AI personas, created based on MMR's comprehensive segmentation study, called ‘Culture’.
Consumers provided short, open-ended answers, while AI personas delivered detailed, multi-paragraph responses. Through meticulous thematic analysis, we discovered that with precise prompting, AI personas are able to generate feedback as detailed, nuanced, and actionable as that of real consumers.
This finding is profoundly encouraging. The AI personas were trained using MMR’s segmentation rather than Gosh!’s primary data, yet they still provided robust insights. The potential to continually enhance these personas with a brand’s primary data means that brands can leverage them to bypass the need for primary research from scratch every time a new consumer question arises. While AI personas are not a replacement for consumer research, because primary research data is required to build them, this approach is swift, cost-effective, and extends the relevance of initial consumer research, maximizing the return on investment by making it more accessible and encouraging frequent use across the organization.
Jake Schneider, Chief Innovation Officer at Gosh! Food, shares, “We are really pleased with the results of this collaboration. Having access to our segments 24/7 via a dialog-based interface means unlocking an additional access point to the consumer point of view. This provides smaller businesses, which might be constrained in their research capabilities, with valuable insights. This approach allows us to include consumer-led sense checks in our primary research, supporting internal decisions and offering confidence that we are on the right path. Furthermore, it is cost-effective and provides immediate access to consumer perspectives, allowing us to innovate quicker.”
The concept of the empty chair serves as an abstraction of a customer, often allowing assumptions and biases to fill the void. However, AI personas represent the next evolution in the relationship between brands and their consumer data. With AI personas, the need for an empty chair is obsolete. Instead, we have a dynamic ‘always on’ chat function, delivering continuous, reliable consumer insights. This technological leap ensures that brands can drive innovation and enhance consumer satisfaction with real-time, actionable feedback.
We presented the learnings from this experiment at IIEX Europe. If you missed it, let us know and we can re-run it for you and your team.
Debunking the Hype: A critical look at neuro- and biometric methods & synthetic data in consumer research
When it comes to sifting through the plethora of opportunities that new tech suppliers promise, Nova does the hard graft so that you don’t have to! We’ve recently completed experiments in two areas getting lots of airtime at present – see below for our considered reviews.
The role of Neuro- and Bio- metric approaches in consumer and sensory research
Hallowed by some as the holy grail in consumer understanding, we were keen to assess whether new solutions in the Neuroscience space could unlock new insights for our clients. However, our recent experimentation has left us underwhelmed by the potential, neither seamlessly integrable, cost-effective nor actionable! We firmly subscribe to the view that brain activity (or any physiological reactions for that matter) cannot be used to decode emotion (if you’re interested in the psychology behind this and the strong conclusion in 2022 by The Information Commissioner’s Office (an arm of the UK Government) that the likes of facial coding are based on ‘pseudoscience’ - see links below).
Despite this, we were keen to investigate if these technologies* could accurately measure consumer engagement across their product usage experience. However, based on our experiences, we’re yet to find an option that addresses this need better than our existing Sensory Deconstruction approaches (which have the benefit of both tracking the experience and providing granular feedback on fit with consumer needs and optimisation opportunities as well as an engagement profile for the product).
*It’s worth noting that we prioritised practical solutions that will work in the moment of consumption, which does limit our exploration away from medical / university grade equipment that needs to be administered under highly controlled environments (and typically at a very high cost).
While we’re eager to harness technology in this realm, the cost-benefit imbalance, coupled with data inconsistencies and lack of diagnostic direction, hinders its application. Our commitment remains to meet consumers in their authentic moments of consumption to capture genuine experiences and opportunities which again limits the practical application of these solutions to unlock insights to power product development.
MMR Research | Predicting Emotions from Facial Expressions (mmr-research.com)
Facial Expressions Do Not Reveal Emotions | Scientific American
The Guardian Article – Information Commissioner warns firms over ‘emotional analysis’ technologies
Does Synthetic Data hold any weight?
In case you missed it, see here for our Global CEO’s thoughts on the role of Synthetic Data, including important watch outs if you’re tempted to try a new supplier in this space.
We’re keenly working with both third parties and our own Data Sciences team to explore how we can leverage Generative AI to create Synthetic Respondents – a space where we see much more potential than the ‘glorified weighting’ to boost small base sizes that sounds too good to be true (and in some of the cases we’ve explored to date, really is).
Getting through to Gen Z: Innovative Engagement with Qualitative AI chatbot
Discover how Gen Z – South Africa's largest consumer segment – can be effectively engaged using innovative qualitative chatbots. In this pioneering study, Jacqui Horsley and Brett Marshall explore the use of Conversational AI to delve into the motivations, needs, and frustrations of this digital-native generation. Where traditional survey methods fall short, Q.Bot emerges as an engaging, scalable solution that fosters trust and delivers deep, actionable insights.
Q.Bot – MMR’s qualitative exploratory chatbot – offers a way to access Gen Z’s stories at scale. After all, when we really want to understand people, we turn to their stories. Stories are engrained in our DNA; they are how we learn, discover, and share our world. Harnessing the power of Conversational AI, Q.Bot fosters one-on-one engagements to get to the heart of Gen Z’s lived experiences. Whether it is discussing their favourite brands, lifestyle choices, or social causes, Q.Bot captures the nuances that matter.
Dive into their stories and learn how brands in South Africa and beyond can decode the Gen Z mindset and truly connect with this dynamic demographic that are shaping markets everywhere.
Exploring Multimodal Engagement: Voice and more
In line with our passion for research that truly connects with consumers, MMR’s Q.Bot is expanding beyond text interactions. We are actively exploring innovative approaches that allow consumers to share their stories in the most natural way possible, making research more intuitive and inclusive.
In this exciting evolution, Qbot experiments with voice, video and image response features. By giving consumers the choice to communicate in their preferred format, we empower them to share authentically. Such a personalised approach also stands to provide clients like you with deeper, richer insights into consumer experiences – the kind of nuance that text alone cannot capture. Stay tuned for updates on our pilot results as Q.Bot’s expanding capabilities shape the future of consumer research!
See NOVA present at upcoming Events:
Eurosense, Science 8-11 September 2024, Dublin, Ireland, Alexandra Kuzmina with a workshop on “Exploring the Intersection of New Technologies and Sensory & Consumer”
Eurosense, Science 8-11 September 2024, Dublin, Ireland, Ansie Collier and Phiala Mehring with “Finding the connections: Leveraging rich consumer conversations to deliver a multi-dimensional understanding of the product experience”
MRS Behavioural Science summit, 26 September, London, Alexandra Kuzmina with “Fast forward: using behavioural science to build a more accurate understanding of future behaviour”
Society of Sensory Professionals Conference 2024, Pittsburgh, Pennsylvania, October 2–4, 2024, Alexandra Kuzmina with “Future-Oriented Concept Testing: An Experiment with Conversation AI and Behavioral Science”
Society of Sensory Professionals Conference 2024, Pittsburgh, Pennsylvania, October 2–4, 2024, Ansie Collier and Phiala Mehring, with “Bridging the gap: Leveraging a trained Sensory Chatbot for enhanced product understanding”
ESOMAR Asia Pacific Insights Ignited 2024, Bangkok 6-8 November 2024, Ansie Collier with Kathy Mendoza, Consumer Science and Product Experience Global COP Lead at Haleon, with “Decoding the thing and I with AI: Turning consumers’ bubbles and brushstrokes into more than data points”
Catch up on past events
IIEX Europe, Amsterdam, 25-26 June Alexandra Kuzmina and Jake Schneider, Chief Innovation Officer Gosh! Food Ltd, with “Synthetic Sparks: Accelerating R&D with AI Personas” and more exciting sessions from MMR.
NewMR webinar Luisa Gibbons with “Identifying where AI truly adds value”
Didn't get a chance to attend these events? Don’t worry! If you’re feeling the FOMO, just let us know. We're happy to arrange an exclusive re-run for you and your team, subject to availability.
Stay up to date
Subscribe to our Quarterly Newsletter powered by NOVA, MMR's Tech Innovation Team