Authored by Richard Heath, Chief Innovation Officer and lead of Curionblue, Curion
AI has transformed front-end innovation. Back-end innovation, where real prototypes meet real consumers, is next. Curion Chief Innovation Officer Richard Heath shares what he learned at MRS Innovation Pipelines 2026 in London
Who’s Leading Product Innovation Research In 2026
Sometimes it feels like fate is on your side.
I was 4 months into setting up a new innovation function. Long enough to know what our priorities are, but still constantly scanning the market for what everyone else is up to.
I’ve always been competitive. But when it comes to innovation, being “first” is a lot more satisfying.
So, when I was invited to go to London to the MRS Innovation Pipelines 2026 and listen to 4-5 of my biggest competitors talk about their strategies, clients, tools and tech back-to-back, I was in!
Assume Your Competitors Are At Least As Good As You
I have often felt the UK’s lead market research trade organization largely ignored the R&D field of market research. But no more! They have created a fantastic full-day event entirely dedicated to sharing innovation best practice, thinking and technology. If you’re in R&D or innovation research, this is an excellent event to attend.
The competition is strong, and nothing is more motivating than someone setting a high bar.
MMR, Ipsos, Zappi, Kantar, NIQ, and so on… they have smart people, improving tech, top-tier clients, and are investing to push product and innovation research forward. And Curion fits directly into this, too! I was attending the greatest geeky party! Frankly, one key takeaway for me was that Curion needs to be involved in these conversations next year.
Front End Innovation Trends: AI and Early-Stage Testing
Overall, there were two big trends on show here.
- Trend one: AI was almost a dirty word. More than once, AI was acknowledged as a useful tool being used, but only in the way you might acknowledge that a Coke Zero has aspartame in it: it’s important, but we’re not proud of it.
- Trend two: Early-stage innovation was very much the focus. It’s tempting to say this is because it comes first in the process. But really, it’s because it’s easier to innovate in. Researching concepts was already easier, cheaper and faster than later stage work.
These two trends are related, of course. Early stage is about ideas and filtering. Ideas can be expressed in simple language, and in digital form. This is bread and butter for AI LLMs.
Indeed, AI has clearly washed over front-end innovation. If you’re reading this and you still spend 6 weeks creating, filtering, refining and testing new ideas, stop reading now and give us a call!
So, why the reticence over AI in Front-End research? Is it because of the known link between many AI engines and online panel sample and its tricky tendency to be a little more “robot” than you’d ideally like? Who knows! One thing I do know: R&D people won’t accept so easily.
A Proven Model For AI-Enabled Product Testing
About 10 years ago, I had the pleasure of being in the same room as Elaine Rodrigo. Maybe she remembers me! Back then, she was an insight leader for Mondelez. I was wrestling with the APAC leg of a truly complex global brand tracking study. She was the leader we needed when it felt like no one seemed to know what to do. She walked in, listened, made the right call, and moved on. I was in awe.
And now, Reckitt is the fortunate business to benefit from her pragmatism and clarity. All conference stories are tidied up versions of reality, but Elaine and her NIQ Bases counterpart, Antoine Saunier, presented the template to follow when driving AI enabled change.
- Start with the problem: Testing takes too long and it’s too expensive.
- Find the right partner: The one with the right people and the right data.
- Lay down the gauntlet: A big hairy audacious goal.
- Support: Even the best agency won’t get it right the first time.
- Have faith. Enough said.
Who doesn’t want to cut testing times from weeks to days, or even hours? The beautiful thing was that Reckitt and NIQ have made it look easy. Maybe it’s when Elaine is in charge!
Why Back-End Innovation Is Harder (And Worth It)
Back-end innovation is the later stage of product development, where working prototypes are tested with real consumers before launch. It covers stage gates, prototype testing and large-scale validation. By this point, money and manufacturing are committed, so a product has to prove it performs with real people before it goes to market at scale.
But at this conference, so far so Front-End. Where are the big ideas to disrupt later stage testing, stage gates, and large-scale validation?
The innovation team at Curion, called Curionblue, is building tech-enabled product research methods that give product leaders faster, more predictive answers at every stage of innovation. And one goal we have is to bring the best of modern front-end innovation to back-end.
The thing about late-stage research is that it is hard. Outside of medical research, it’s probably the hardest research. Real working prototypes need real messy people in the real messy world. Even in facilities, you’re dealing with valuable prototypes, precise tests and the randomness of real people.
It is precisely this that makes it worth doing. High stakes need high research.
Late-stage also carries more jeopardy than concept stage ideas. Money has been sunk. Factories have been deployed. The latter stages are where things get real.
As Mandy Seal, Global CMI Director at Mars Pet Care mentioned… this is where we enter the ‘Squid Game of Innovation’.
The beauty of the work being done at Front-End is that it gives us more chance of success in these Squid Games. But it’s unlikely to remove “The Games” altogether.
The stakes are simply too high to remove the stage gates.
Companies consistently spend millions here, every year. But this pales into insignificance compared to the risk of allowing untested ideas to be loose at scale.
The goal is not to remove. The goal is to streamline. And the opportunity is real.
There is no reason we can’t speed up Back-End innovation in the way we have with Front-End.
How to Speed Up Back-End Innovation
There are 3 levels to what we need to do:
- Automate: There is no reason the phases before and after can’t get the same treatment Zappi has given concept and ad testing (if you know R&D the way Zappi knows advertising).
- Simulate: Curion has SO MUCH DATA! Curion has the world’s only validated metric of product repeat purchase. Top tier clients have been testing so consistently with us for so long that it is now possible to predict performance for them with little or no new data. This is not about synthetic. It’s about using AI and machine learning to create truly powerful simulations of what would happen if product A were tested in market B.
- Formulate: What happens when you combine 50 years of product testing data with now easily accessible nutrition and ingredient data? You can create a real formulation and know whether people will engage with it.
Easy, right? Well… not quite. There is much to learn from Elaine’s model at Reckitt. Set the challenge; find the partner; truly, partner; and have faith.
This Time, Next Year
I’ve been with Curion 6 months. We have the data. We have the analytical smarts. We have the AI tech partners. We have the will. And we certainly have the clients willing to come on the journey.
So, this time next year, the team and I are looking forward to contributing with a little more Back-End!