Research + Concept Design · 2025

Roti Kapda Makan

Order what you want. Eat who you are.
A research-led feature design for goal-based eating on food delivery platforms

Type

Research + Concept Design

Context

Academic · NID Interaction Design

Timeline

Apr 2025 – Nov 2025 · Ongoing

Research

With Averil Sybil Stephen (NID)

Hero — Lifetime of Food and Nutrition chart
0 interviewsTelephonic across mixed age groups and cities
0 respondentsPilot survey across 7 Indian cities
0 methodsPrimary research methods combined

Section 1

A lifetime of food, and one vulnerable decade

We began not with users, but with the system. Secondary research mapped the relationship between food, knowledge, environment, and institutional support across an Indian citizen's lifetime — tracking four variables simultaneously: food autonomy, nutritional knowledge, nutritional needs, and external care.

A specific pattern emerged in the 18–30 age window. External care drops sharply — for the first time, people are feeding themselves without family or institutional oversight. Food autonomy peaks. But nutritional knowledge is still catching up. The gap between autonomy and knowledge creates a period of heightened moldability: food preferences and habits are being formed without a strong foundation to form them on.

Lifetime of Food and Nutrition chart — 18–30 window highlighted

Migration compounds this at exactly the wrong moment. The environmental scaffolding that previously shaped food choices — family, regional cuisine, home cooking, familiar vendors — collapses simultaneously. What fills the gap is whatever is most convenient.

How might we reinforce the food habits of migrant urban youth for better nutritional practices?
Causal loop diagram — breakdown areas from system map

Section 2

Five methods, one territory

Before beginning primary research, we returned to the system map to identify where the breakdown around food actually happened. This consolidated into a three-part framework: shopping, consuming, and cooking — each existing within an environment of regional context and lifestyle influence.

We also made a deliberate decision about emotional data. Food is deeply tied to memory, identity, and emotional state — but we couldn't pre-define "emotional connection" as a research category without risking that participants would perform toward it. Instead, we stayed open.

The emotional layer wasn't planned for. It insisted on being found.
Research framework — Shopping, Consuming, Cooking, Regional Aspects

Pilot survey

A pilot survey covered food habits, regional influence, and knowledge about food and nutrition across seven cities: Kochi, Chennai, Pune, Bangalore, Vadodara, Ahmedabad, and Varanasi. 35 respondents. Two findings confirmed the problem was structural: migration created dramatic shifts in food behavior across all cities, and cost consistently outranked nutritional value in food decisions regardless of city, income, or food culture.

Survey data synthesis — 91%/56% shift, price vs. nutrition ranking

Group discussions

The discussions confirmed the quantitative direction but added texture — the social, emotional, and cultural dimensions of eating that a form couldn't capture. Food was comfort. Food was nostalgia. Food was guilt. Food was a proxy for home.

Observatory research

What we saw: once the decision about what to eat is made, the act of eating itself becomes passive. Attention goes elsewhere. Reflection about food happens only as feedback about taste or quantity — never about nutrition. The intention to eat well exists at the decision point. By the time of eating, it has already faded.

Observatory illustration — three figures eating passively

Journalist interviews

Ten participants, two fixed questions, no preparation. Three themes emerged: balancing diet as a concept people understood abstractly but struggled to operationalize, regional foods perceived as inherently healthy, and freshness and cleanliness as the primary markers of food quality.

Shopping cart analysis

Each participant was asked to build a shopping cart for their next meal on a delivery platform of their choice. The outcome was consistent with the survey: price and brand value dominated decisions. Nutritional content was rarely a factor. The gap between stated values and actual choices was quantified here in the most concrete way the research produced.

Shopping cart analysis — participant building cart on phone

Expert conversations

Two informal conversations — a general physician and a dietician. A common theme: people eat to fulfill energy needs while ignoring nutritional needs. These are not the same thing. The intention is consistently present. The action waits for a crisis.

Section 3

From research to patterns

Across eight touchpoints of primary and secondary research, a consistent shape emerged. The affinity mapping organized itself into three gap areas: Education and Awareness, Behavioral Change, and Food System and Sustainability.

Of these three, Behavioral Change sat at the highest-impact, highest-feasibility intersection of the prioritization matrix — making it the most actionable territory for design intervention.

Affinity mapping — three gap categories
Intervention area prioritization matrix — impact vs. feasibility

The causal loops from the system map, revisited in light of primary research findings, revealed where interventions could create meaningful change versus where they would be absorbed without effect. The loop that mattered most: the connection between ordering behavior, identity, and the guilt-reward cycle.

Causal loop diagrams — decision pivot points

Section 4

Three archetypes, one underlying tension

The research produced a spectrum — three distinct relationships with food and health awareness, each with a different entry point for intervention. These are not invented personas. They emerged directly from the pattern of primary research data across methods.

Archetype visual — three profiles side by side
01

Unaware and Unbothered

Lives with the flow. Takes spontaneous decisions about food without deliberation. Has fragmented knowledge about nutrition and fails at continuity when they try to implement healthier habits. Not resistant to change — but not actively seeking it. Food for them is pleasure, social connection, and convenience.

02

I Know, I Know

Knows everything — in theory. Stuck in a perpetual loop of learning more, optimizing more, and doing less. Knowledge has become a substitute for action. Every new piece of information is a reason to wait until the system is perfect before starting. Was once healthy without thinking about it, carried by an environment that supported good habits.

03

I Am a Healthy Person — I Once Was

The most important archetype for the design direction. Doesn't have the most knowledge, but has something more powerful: an identity. Remembers what it felt like to eat well. The challenge is that urban migration has disrupted the environmental conditions that made their habits automatic. They haven't abandoned the identity. They've lost the infrastructure that supported it.

The design opportunity isn't to educate the first archetype or complete the second archetype's knowledge. It's to give the third archetype back the environmental conditions their identity needs to reassert itself.

Urban youth aspire to live healthy, but they don't organise their food around it.

Section 5

The negotiation nobody notices

When people were observed eating, food was a background activity. Reflection came afterward — about taste, about how much was consumed, occasionally about whether the choice was a good one. That reflection carried emotional weight. Satisfaction. Mild guilt. The quiet resolution to do better next time.

Put the survey data, group discussions, observatory findings, and journalist interviews together and one organizing tension emerged: people have a genuine relationship with food, but the moment of eating has been disconnected from that relationship.

The negotiation, slowed down

Every food order contains a decision more complex than it appears. The surface-level decision is "I want biriyani." The actual decision, slowed down:

"I've been eating rice and sabji all week. This is my moment of indulgence. Nobody is telling me to follow a diet. I have time for a nap after. I'm getting the biriyani."

That internal monologue — balancing current health state, earned indulgence, occasion type, guilt calibration — happens in seconds, below conscious awareness. It is never surfaced by the ordering interface.

Guilt as a design resource

One emotional signal ran through every research method without being planned for: guilt. But in the research, guilt didn't look the way health interventions typically treat it — as a problem to be eliminated. It looked like a calibration mechanism. An indulgent meal was followed, consciously or unconsciously, by a corrective one.

Guilt swings both ways. If indulgence is one side of the scale, the healthier choice is the other. This isn't dysfunction — it's a natural behavioral rhythm that design can work with.

The key tension

Cities don't have a food problem. They have a health-convenience problem. People are not making bad choices. They are making convenient choices. Any intervention that asks people to sacrifice convenience for health will fail at scale.

Every food order is a negotiation we have with ourselves — between who we are, what we've earned, and what we want right now. Research made this visible. Design can surface it.

Section 6

Goal-Based Eating — a trust layer for food delivery

Any intervention requiring a new platform, a new habit, or a new identity will fail adoption. Food delivery platforms are already the infrastructure of urban eating. The intervention doesn't need to create a new context — it needs to insert itself into one that already works.

In late 2025, Zomato launched Health Mode. In early 2026, Swiggy launched Eat Right. Both confirm the direction this research pointed to independently in April 2025. But both are recommendation engines — they tell you what to order differently. Neither creates a moment of self-recognition. The gap they leave open is exactly the gap the research found.

Four functions

Remind

The platform holds the user's goal and surfaces it at contextually relevant moments. Not nagging — awareness. “You've eaten well all week. Tonight's yours.”

Guide

Goal-aligned meal suggestions rise to the top. New restaurants matching the goal profile appear. The full menu is always visible. Enablement, not restriction.

Track

Progress shown as a pattern narrative, not a performance score. The biriyani is logged. It counts. “You're not following a plan — you found your own rhythm.”

Flex

Goals can be changed at any time, with history retained. The commitment is to a direction, not a fixed destination.

Screen A — The Mirror Moment

The order is confirmed. Then the platform does something different: it speaks first. A Mirror Card surfaces — a contextual observation built from what was ordered, when, and behavioral patterns. No question is asked.

"Friday night. A long week. You ordered the Biryani. Some meals are just for this."

Two reactions: "Exactly this" — I eat well usually, tonight's mine. "Not quite" — I'll shape this myself. A third option validates the choice without shaming. Commitment is initiated by recognition, not interrogation.

Screen A — Order Confirmed + Mirror Moment

Screen B — Goal Setting

The goal panel unfolds inline after the user reacts. Four goal options presented as identities, not restrictions: Stay Balanced, Fuel Better, Go Lighter, Eat Cleaner. A commitment reframe panel appears first — explicitly stating what opens up, not what closes down.

Every restaurant stays open. Goal-aligned meals rise to the top — the rest don't disappear. Biryani nights count too. Balance includes them.
Screen B — Goal Setting panel

Screen C — Goal Tracker

Week view with day dots — filled for goal-aligned meals, outlined for today, empty ahead. The biriyani is already logged. It appears as part of the pattern, not outside it. The weekly insight is the most important element — not a score, a narrative:

"Every Friday you've ordered something indulgent. Every Saturday, something lighter. You're not following a plan — you found your own rhythm."
Screen C — Goal Tracker
Home tab — ongoing goal reminder and personalised suggestions

The tone

The word "healthy" rarely appears in the feature copy. The framing is always about how eating feels — balanced, fuelled, lighter, cleaner. People don't organise their food around health as an abstract goal. They organise it around identity, feeling, rhythm, and belonging.

The app becomes the all-seeing partner who nudges you toward your own goals — not a strict system watching for failures.
IA diagram — full information architecture across Order, My Goal, Home tabs

Section 7

What this project taught me

What worked

Starting with the system before starting with the user was the right call. The lifetime chart and the system map gave us a principled reason for our user group and intervention territory. The decision to stay open to emotional signals rather than pre-categorizing them was also right. The guilt insight, the love of food, the identity thread — none were planned for. They arrived because the research posture was receptive.

What I'd do differently

The research question evolved significantly between the system map phase and the primary research phase. This evolution was real and important, but it wasn't documented clearly as it happened. I'd also conduct more rigorous expert conversations — longer, more structured, and with a wider range of practitioners including nutritionists, behavioral scientists, and food platform product managers.

What I'm carrying forward

The Mirror Moment — the core of the feature — wasn't invented. It was found in the research. The observatory findings showed where reflection happens. The guilt cycle showed what's present at that moment. The archetype showed what kind of trigger would land. The design was the act of connecting those dots.

The best features aren't designed from imagination. They're designed from attention.