Methodology · Semiotics · Advertising

Researching a Joke

A method for testing humour without killing it

Methodology · 2018–2019 · Semiofest, ESOMAR APAC, Qual360 Europe

"But how does one research a joke?"

At a glance

  • Type: Methodological innovation, applied in two brand studies
  • Applications: A tyre brand's road-safety communication (CEAT) and a food-delivery platform's campaign (Swiggy)
  • My role: Created the method; research lead and lead author
  • Methods: The Unfocused Group, semiotic decoding, in-depth interviews, social listening
  • Presented at: Semiofest 2018, ESOMAR Asia Pacific 2019 (published paper), Qual360 Europe 2019

Context

Early in my career, I presented the results of an ad test to the head of marketing of a leading FMCG company. The focus groups had trashed the film. He listened, visibly bored, and at the end asked one question: "But how does one research a joke?"

His point stayed with me for years. If that ad had come on in the middle of something people were watching, they would probably have laughed, and remembered it. But put them in a room, play the ad and ask "Did you like it? Did it make sense?", and they switch into judges. Nobody wants to be the one who laughed at the silly ad.

The question

Humour is one of advertising's most powerful tools, and one of the hardest to research. So the question was methodological before it was commercial: can we create research conditions where people react to humour the way they actually would, instead of the way they think they should?

And once we could see humour clearly, could we decode how it works in India, where we laugh at ourselves easily but some subjects remain firmly off-limits?

My role

I conceived and designed the method, led the research and analysis, and was lead author and presenter of the work across three international conferences.

Approach

  1. 01

    The Unfocused Group.

    We kept the group but reversed almost everything else about it. Instead of a facility with one-way mirrors, cameras and stock art on the walls, we met in a warm, homely space. Instead of a moderator asking questions, a host. Instead of a discussion guide, something closer to an adda or a kitty party: friends chatting, with the ad played as part of the conversation. The moderator observed, let talk flow as stream of consciousness, and guided only lightly. We were watching for the laugh, the groan and the retelling, not for the verdict.

  2. 02

    Semiotic decoding.

    Alongside the groups, we decoded how humour works in Indian advertising, mapping campaigns within each category and well beyond it against the major theories of why people laugh.

  3. 03

    Inside out, then outside in.

    We started with in-depth interviews and analysis of past communication to understand the codes, then tested which themes resonated with consumers. Social media conversations added another layer, because that's where people are at their most candid.

Focus group vs. unfocused group
Focus groupUnfocused group
FacilityHome
Moderator asksHost observes
Evaluate the adLive with the ad
PosturingConversation

What we found

  1. 1

    Indian humour is cultural, conversational and a little caricature-ish.

    Brands lean on "only in India" situations, wordplay and stereotypes, and the codes were shifting fast towards meme culture and more outward-looking humour. We laugh at ourselves easily, and our jokes can be edgy. But there are lines: don't joke about our gods or our holy cows, actual and metaphorical.

  2. 2

    Different humour does different work.

    We mapped Indian advertising against four theories of humour:

    Relief:

    laughter that releases tension, letting even dark themes like death and injury be discussed.

    Superiority:

    laughing at someone else's misfortune, often slapstick, and often a cautionary tale.

    Incongruity:

    putting together things that don't belong, for surprise and memorability.

    Benign violation:

    breaking a norm in a way that stays safe, funny because it's true.

    The right humour depends on what the brand needs the laugh to do.

  3. 3

    For road safety, benign violation fit best.

    For CEAT, a challenger tyre brand in a category that used little humour, the opportunity was to stand out without trivialising safety. Shaming humour felt wrong for a serious subject. Benign violation let the brand name a truth every driver recognises, that the road is full of idiots, without lecturing.

  4. 4

    Masculinity was the other code to crack.

    Category advertising leaned on familiar masculine personas: the charmer, the dependable boy next door, the controlling alpha, the lone risk-taker. The market leader owned the alpha. We found room for a different man: not a macho hero, not quite a "woke bro", but a self-aware observer who questions the world's foolishness with a wry smile.

  5. 5

    Humour can restore dignity.

    For Swiggy, the same method turned to identity. Customers routinely called delivery partners "Swiggy", erasing the person behind the uniform. Humour let the brand make that point without guilt-tripping anyone.

What we recommended

For CEAT, three building blocks for the narrative:

Promise: a brand that is your ally against everything that's wrong on the road.
Persona: a funny observer of life, self-aware rather than macho.
Packaging: humour that gently pokes fun at others without offending.

For Swiggy, a story built on a deliberate absurdity: imagine a delivery partner whose parents actually named him Swiggy, from the "Happy Birthday, Swiggy" cake at age five to an awkward moment with a girl he likes. It ends at a customer's door, when she finally calls him by his real name. The message: our delivery partners have names, so use them.

Impact

The research shaped communication for both brands: a road-safety platform built on wry, self-aware humour for CEAT, and, for Swiggy, an ad humanising delivery partners that was conceptualised directly from our findings.

The method itself travelled further than either brief. It was presented at Semiofest in 2018, then at ESOMAR Asia Pacific and Qual360 Europe in 2019, where the paper "Researching a Joke, Are You Kidding Me?" was published in the ESOMAR conference proceedings.

The hard parts

  • Unlearning the focus group. Moderators, clients and participants all arrive with habits about how a group should run. Letting go of the guide, and trusting that the insight would come from conversation, took deliberate design and nerve.
  • Analysing what isn't said. In an unfocused group, the data is laughter, interruptions and retellings, not answers. Making that rigorous meant watching closely for reactions, not just listening for opinions.
  • Working near the line. Humour in India can delight or offend in the same breath, and safety, masculinity and identity are all sensitive subjects. Every recommendation had to be funny without punching down.

Reflection

That marketing head's question was dismissive, but it was also the right question. Some things lose their meaning the moment you ask people to judge them, and humour is one. The answer wasn't to stop researching jokes. It was to stop asking people to perform as respondents.

I still use that lesson in AI research today. The most honest data rarely comes from asking people what they think. It comes from creating conditions where they can simply be themselves.