Behavioural data tells you what happened. Qualitative research tells you why. The researcher’s job is to take complexity and make it simple enough to decide on.
01The job, and a typical day
Fiftyfive5 is a qualitative research agency — words and feelings rather than numbers, behavioural shifts, and the human side of the consumer. It was an independent agency and has since been acquired. The client list spans commercial, government, education, tourism, fast-moving consumer goods and alcohol, including RTDs in New Zealand.
The work has one foot in the boardroom and one in the street, and the researcher has to code-switch between the two.
| Face | What happens there | Output |
|---|---|---|
| Street the coal face | Talking to a wide variety of people — physiotherapists, speech pathologists — about purchasing needs and behaviours, for instance to inform the marketing of a new training program. | Primary field data |
| Office internal facing | Analysis sessions and writing the debrief after fieldwork. | Analysis and brief |
| Boardroom client facing | Taking the brief, presenting the proposal, kicking off the project, delivering a fast-turnaround verbal topline, and running workshops to help clients actually use the research. | Proposal and topline findings |
- Qual is a team sport. The questions are always “what have we learnt” and “what does this mean for our client’s business.”
- Insights-into-action workshops exist because findings on their own don’t change anything — the application is the deliverable.
- Clients come from across the business — product, innovation, R&D, consumer strategy and consumer insights teams.
02How market research helps businesses grow
Three buckets, and they map onto the marketing process in order.
1. Identify growth strategies — where to play
- Demand frameworks and segmentation to identify, size and prioritise growth opportunities
- Market, cultural and category exploration to identify growth platforms
- Implications for acquisition, category drivers, portfolio planning, positioning, pricing and innovation
2. Execute against growth strategies — how to win in market
- Defining an insight-led customer value proposition (CVP), positioning and creative strategy — then developing and testing it
- Identification and sizing of innovation platforms; ideation and concept development
- Co-creation, development and validation of executional product, service and channel blueprints
- Customer experience strategy and design
- Pricing strategy
- Shopper, channel and retail strategy
3. Evaluate performance — metrics that matter
- Integrated performance-monitoring programs covering CX, category, brand and comms, and sentiment and reputation
03The three gaps generative AI can’t close
The tempting move is to “just add Gen AI.” The more useful framing is human in the lead, human in the loop — deciding which parts need a person driving and which only need a person checking.
| The gap | Why it persists |
|---|---|
| Human and cultural understanding | Meaning is cultural and contextual, and that is the part the model has least grip on. |
| The data doesn’t exist yet | AI is only as good as the data plugged into it. If nobody has asked the question, there is nothing to train on. |
| LLMs can’t handle the analysis | On-platform tools in particular make a lot of mistakes — and businesses are basing multi-million-dollar decisions on the output. The question becomes: where are the mistakes, and where do we need human quality control? |
Too much data creates complexity, not simplicity. CRMs, sales data, dashboards — none of it automatically produces a decision. Our job is to take complexity and make it simple.
What versus why
Behavioural data will tell you the what. Qualitative data tells you the why.
The example given: the data shows people are not drinking as much. It cannot tell you why — and the why is what determines what the business should do about it.
The same gap shows up in what people report about themselves. A consumer says “I want something cheaper,” but the real driver may be value for money, convenience, saving time or reducing effort. What people say is not always the reason behind their behaviour, so you keep asking why.
04The qualitative toolkit
The researcher’s job is to design the most effective solution against the brand and business objectives — a bespoke method for System 1 and System 2 thinking, drawing on behavioural, motivational, ethnographic and digital approaches.
Talking to people
Online and face-to-face discussion groups
Focus groups, typically six to ten people discussing a brand, product, ad or pack. Online versions are faster and cheaper and let you reach people in different cities or countries. In the room you observe reactions, emotions and how people respond to each other — not just what they say.
Conflict encounters
Deliberately putting people with opposing views together — Apple users against Samsung users, for example. Conflict surfaces strong opinions, underlying values and real attitudes that politeness would otherwise hide. One example discussed was emerging views on gender in the country.
Consumer juries
A group of consumers acts like a jury, evaluating an ad, product, brand idea or campaign, then delivering their ruling and recommendations from the consumer’s point of view.
Cognitive interviewing
Focused on how someone thinks through a decision, and on using memory to reach emotion. Why did you choose this? What were you comparing? What made you hesitate? What changed your mind?
Watching what people actually do
Shopping safaris
Go shopping with the consumer — physically or digitally — and watch what they really do. Where they go first, what catches their eye, what they pick up and put back, where they hesitate. Important because what people say they do is not what they do.
Ethnographic observation
Observing people in their real environment: home, workplace, kitchen, school, the shop. People often cannot fully explain their own behaviour, so it is better to watch it. The example given was baristas working at speed, then watching the footage back with them.
Mobile ethnography and digital co-discovery
Participants document their own lives on their own phones — photos, videos, diary entries — capturing experience in the moment rather than relying on recall.
Consumer immersions
Taking the client out of the office and into the consumer’s world. One example: putting clients on a bus into a lower socio-economic neighbourhood. The point is not to interview the consumer but to briefly become them.
Longer-running and secondary methods
Agile innovation
A fast, repeated loop — test, feedback, change, test again — for a company developing a new product or service, with the consumer brought into the process. Client workshops and consumer groups jump back and forth rather than waiting for something to be finished.
Digital communities
Longer-term online communities with set tasks, where researchers keep interacting with the same people over time rather than capturing one interview at one moment.
Case studies
Multiple people talking about the same thing from different angles. The example: understanding a university choice by interviewing the student’s parents, teachers and peers about why.
Social listening
Scraping conversations already happening online — TikTok, Instagram, Reddit, forums, reviews. The advantage is that nobody is answering because a researcher asked; you get the language people actually use. This is one area where AI genuinely helps.
05The disciplines behind the research
Good qualitative work is not just interviewing. It combines deep qualitative experience with complementary disciplines, pulled in and co-ordinated to solve the client’s problem — for businesses, but also for government and social enterprise.
| Discipline | What it contributes |
|---|---|
| Cultural insights | Culture, values and lifestyle — behaviour makes sense once you know the context. |
| Behavioural science | Why people behave as they do, and the gap between what they say and what they do. |
| Behavioural economics | Why consumers aren’t rational — loss aversion, anchoring, framing, nudges. |
| Psychology and social psychology | Emotion, memory, perception, motivation, decision-making — and how other people, social norms and group pressure shape choices. |
| Customer experience (CX) | The whole journey: before, during and after purchase. |
| User experience (UX) | The act of using the thing. Is it easy, confusing, convenient? Where does it break? |
| Cross-cultural analysis | You cannot copy and paste a strategy across markets. Culture, values, social expectations and habits differ. |
| Semiotic analysis | What colours, images, logos, fonts and packaging actually signal within a given culture. |
| Brand and comms strategy | Positioning, key messages, advertising strategy. |
| Creative development | Developing and testing campaign ideas, concepts and new products. |
| Client-side experience | Understanding the business problem, not just the research method — what does the client actually need to know and do? |
| Future forecasting | Possible future consumer trends, cultural shifts, industry change and opportunity. |
And it isn’t only consumers
Audiences include consumers from all walks of life, B2B buyers, trades, healthcare professionals, and culturally and linguistically diverse Australians.
06Four projects
Creative effectiveness
The problem. Brands spend millions filming an ad campaign, and want to know before they spend whether it will work, and where. An effective campaign has to do several things at once: stand out, deliver a clear message, engage its specific audience emotionally, and announce a product or build brand love.
The solution. A simple global scorecard against effectiveness criteria, nuanced by market; optimisation recommendations for the creative ideas; and detailed, agnostic identification of script problems in specific markets. In the campaign discussed, the ad relied on jokes — which is hard to make work across seven cultures. Increasingly enabled by AI moderation.
Mobile ethnography — baristas and plant-based milk
The problem. Consumer needs are often subconscious. Baristas working in a blur serving customers are not consciously thinking about packaging pain points.
The solution. Wearable body cameras. The work identified the truths, needs and tensions for professional baristas using plant-based milks; unpicked opening, storage, usage and disposal issues; trialled new packaging prototypes; and developed innovation platforms over Zoom — watching the footage back with the baristas and talking through their thoughts and feelings.
Cross-cultural analysis — a tea brand across seven markets
The problem. How does a company position a brand as one coherent story across multiple global markets when each market has a different set of competitors and a different retail environment? Markets included India, Saudi Arabia and the Philippines.
The solution. Qualitative research to uncover needs, brand perceptions and competitor perceptions, with cross-cultural analysis across the multi-market work. Running the markets together saves the company money — you are looking for the similarities, so you don’t end up making three separate campaigns.
Demand prediction tool — what will students want to study?
The problem. Universities need to respond quickly to shifts in student demand and competition, but insight today is fragmented and lagged, which makes emerging opportunities and demand gaps hard to see.
The solution. A single trusted view of education product demand, built by connecting data sources into a unified modelling layer.
| Market demand data | University performance data |
|---|---|
| Population demographics Student demand Skills demand Graduate outcomes Competitive intensity | Market share Student success Financial performance |
The output is two maps. For existing programs — university performance against attractiveness — to see what is still worth doing. For new programs — momentum against size — to see what should be brought out.
The point made about this one: you cannot read it off demographics and numbers alone. Cultural drivers, social factors, values, motivations and psychology explain why students make the education choices they do.
07Activating research with generative AI
Ubersight
Their own tool — a way to supercharge existing frameworks with more data, integrate different data sources into the work, and ensure new knowledge compounds rather than being lost at the end of each project. Used for data synthesis, competitor scoping, cultural intelligence and trend analysis.
| Layer | What it captures |
|---|---|
| Cultural drivers | The known and agreed future, defined by the hierarchy of global leaders and shifts in STEEP factors. |
| Macro trends | The overarching cultural narrative driving behaviour, beliefs and responses to products, services and brands. |
| Micro trends | The smaller, faster-moving expressions of that narrative. |
| Tension | Where the forces pull against each other — usually where the opportunity is. |
Dynamic personas
Their proprietary take on synthetic personas: customer segments loaded into an interactive, queryable interface, where the researcher controls the data set. A client might say “strip out everyone except the indulgent escapers” — and the persona is built from that client’s own qualitative research, journey mapping and survey data sitting in their knowledge bank. You can then ask it questions: if I did X, how would this segment respond?
What makes the approach different is the foundation — built on the right data and insight for that business specifically, and able to integrate multiple data sources into a single customer view.
Online task boards
Bespoke tasks designed for participants to complete — “getting to know you” exercises, weekday food diary entries. Often used when a client says “this is my target market, but I feel like I don’t really know them.”
Dovetail
A video platform with transcription, insight tagging and downloadable showreels. Video comes in, transcripts and clips come out, and an insight board supports the analysis — with on-platform AI tools now built in.
08Guardrails for AI in qualitative research
| The dos | The don’ts |
|---|---|
| Efficiency. Every project transcript housed in one spot, accessible to the whole team, and fast to group raw qual data by theme across interviews. | AI makes mistakes. And the decisions riding on the output are large. |
| Assurance. You can scan every transcript from every interview, so no insight is left behind. | Words are only one human cue. Tone, hesitation and body language don’t make it into a transcript. |
| Rich data outputs. Hit a button and get a video reel. | Some data environments are not secure, and video is not always OK to use. |
| Persuasive in a boardroom. One genuine consumer clip can be more convincing to a CEO than a researcher and a slide deck. | Transcription is not perfect, and translation treats language like maths. |
The residual worry, in her words: feeling like the child in The Emperor’s New Clothes — everyone agreeing the output is good when it may not be. Watch for preconception and bias, in the tool and in yourself.
09Key takeaways
- Data tells you what happened; qualitative research explains why.
- Research is about understanding people, not collecting data.
- What consumers say isn’t always what they do — which is why observation, ethnography and shopping safaris earn their place.
- Keep asking why. Curiosity is the core skill.
- Insight is not the end goal. The research has to help the business make a decision.
- Culture and context matter, especially across markets.
- AI is a tool, not a replacement. Human judgement and interpretation still do the analysis.
I would spend 55 minutes thinking about the right question.
Where the name Fiftyfive5 comes from