Now that the dust has settled since presenting BBH USA ‘Marketing’s Most Expensive Oversight’ at Esomar Congress 2026, I’ve had time to reflect on the many presentations and meetings I had.
Below are my top takeaways, thematically grouped, but also, I’ve had a think about how you can practically apply this to your workflow.
Theme #1 – Keep things messy, embrace contradictions
A number of the presentations touched upon the idea that too often, when we go into reporting mode, insight professionals seek consistency.
There are several reasons why that is the case:
- Objectives – Our primary role is to find answers to specific research questions/ objectives, which naturally narrows our vision. Connected to this, often we’re having to find a role/ purpose for the brand we’re working for.
- Time – Often we’re working towards tight deadlines. To meet them, you have to be tactical in how you analyse the data you have.
- Human nature – We’re humans; we love to “put things in boxes”. That’s because our brain hates contradictions, which leads to anxiety. This is compounded by how we’re educated, where we’re taught to challenge inconsistencies.
Instead, many talks championed the idea of embracing the contradictions and also keeping things messy.
What they mean by this is giving yourself space to explore the insights that spike your interest, but also freeing yourself from pre-defined reporting flows, specific frameworks, and classic research ways of presenting insight (e.g., resolving tensions, clearly defined needs etc).
Putting this theme into practice:
- Spaces – Create a space for yourself (or team) to be continually noodling on ideas, thoughts, or back pocket ideas which sit outside of the focus of the study. Whether that’s a Miro board, a channel (think Slack, Teams, etc) or just a shared Word document – whatever is easiest for you and others to quickly share.
- Contextualise – Let’s bring back the focus on contextualising our work. What I mean by that is just a slide that says – 10 Things We Learnt – before you then go full steam into your presentation. Keep in mind, those gems you talk about at the start often will impact the way consumers react to the concept you explored, or the way they engage with a product daily
- Step back – You’ve turned that topline report around in one day! Well done, but now, force yourself to take a step back and reflect on all things you decided not to include in that topline. Now write them down and append them to that report. It doesn’t need to be pretty. This gives you and your stakeholder the chance to experience these learnings (instead of them being completely forgotten!)
Theme #2 – Focus on the edges
I am a firm embracer of AI in qual (I’ve been championing it since 2020 at Qual360).
But in the wrong hands, it can lead to thinking and final recommendations becoming generic. That’s because most AI tools are built to look for patterns, or the tool is designed to review data in a particular way.
It’s because of this that several presentations highlighted that some of the best insights or newest thinking often comes from the ‘edges’ – that’s either spotting something radically different to what everyone else is doing, or, it’s investigating the ‘outliers/ niche groups’ – looking to understand what we can learn from them.
But this isn’t just about where you look; it also has a profound effect on how you do things too.
Putting this theme into practice:
- Disrupt – When designing research (whether qual or quant), how can you disrupt people’s thinking – forcing them out of social conformity and generic responses?
- Diverse – During our presentation, we championed the idea of ‘Design for ADHD, to design better for everyone’. This insight came from speaking to an audience that has for too long been overlooked. So, think about which audiences you often ignore – and speak to them.
- Off-topic – Create spaces in your research questioning that allows people to share what they want. Quite often, the most profound insights I’ve captured begin with a simple question such as ‘What’s on your mind?’
Theme #3 – Synthetic data
How could I not talk about synthetic data!
A dominant theme during the conference (however, interestingly, Simon Patterson from QRI Consulting) reminded us that we’ve been exploring/ debating this topic for at least 7 years!
If you’re not aware of what synthetic data is, ChatGPT told me it’s: ‘Synthetic data is artificially generated information that imitates real-world data for testing, training, or analysis.’
We could debate how accurate it is (depending on which talk, accuracy ranges from 70% to 85% – make of that what you will), but my takeaway is that its use is ‘additive’ to more traditional methods (particularly qualitative research), and typically it’s used to quickly generate or test new ideas.
Putting this theme into practice:
- Build – If you have years of data on a specific group of people talking about a particular product, start exploring how best to build your own synthetic data. This can be achieved via ChatGPT, Claude, but also Copilot. But word of warning: make sure this is done in a closed environment
- Rehearse – Once built, use simulated participants to practise interviews and refine lines of questioning. Treat their responses as prompts for better questions, rather than customer insights.
- Benchmark – If you want to go all in, first run a small experiment alongside traditional study to compare and contrast findings. Look at where synthetic responses align and where they diverge.
Theme #4 – Highlight our value
AI is rapidly evolving our industry. It’s empowering us to scale what we do, but also to speed up specific tasks.
But as many presentations highlighted, it also raises profound questions about what exactly the role of researchers is. For example:
- How do we further avoid commoditising what we do?
- How do we look to differentiate ourselves from our programmed alternatives?
- How do we clearly highlight the value we bring to a project or workstream?
Putting this theme into practice:
- Frame – Think carefully about how you talk about yourself (and how that is different to whatever AI tool you use). Beyond simply “being the human,” what do you actually bring to the table? Is it professionalism, a consultative nature, sharp judgement, the ability to ask the right question at the right moment – or something else entirely?
- Clarity – There are 175 zettabytes of data on the internet. 5 focus groups will generate around 90000 words. In short, we’re all drowning in data. Therefore, our role should squarely focus on turning this abundance of information into clear, actionable recommendations.
A final thought…
To round off this piece, I’ll leave you with a profound question posed by our keynote speaker, Dr. Knatokie Ford: how do you prepare and educate children for jobs that don’t yet exist?
This really struck a chord with me.
At Shape, we’re engaged in endless debate about how we’ll be conducting research in the future – what tools we’ll use, what skills will matter, what “expertise” will even mean.
Dr. Ford’s question suggests the same uncertainty applies just as much to the people we’re raising and training today in market research.
If we don’t know what the work will look and feel like in the future, our job isn’t to predict it – instead, it’s to:
- Train people to be adaptable
- Fuel their curiosity in the world around them
- Equip them with foundational thinking skills
- Teach them to ‘read the room’/ use judgement
- Embrace the messiness, the contradictions, non-linear thinking
I believe these types of skills will outlast any one job description, but also are AI-proof.