Projects GroupBima
Making health insurance feel less like a burden — and more like a decision you’re proud of.
00 At a glance
01 Challenge
GroupBima is an IRDAI-approved insurance brokerage founded in 2018 with a clear ambition: make insurance accessible, transparent, and digital-first for Indian consumers. The product dream was real. The experience problem was too.
The user pain was threefold. Most Indians understood insurance was necessary, yet consistently deferred buying it — online or offline. The reasons weren’t financial. They were experiential. Existing platforms buried users under jargon. Agents were perceived as self-serving. Policy comparisons were overwhelming without being helpful. Trust, the most critical currency in financial services, was in short supply.
The business problem was structural. The market already had established players — PolicyBazaar, Coverfox, Turtlemint — with years of brand equity and massive user bases. GroupBima needed a differentiated experience, not a me-too product, to earn its place.
The design brief, distilled Design an insurance platform that earns trust in the first minute, guides users to the right policy without overwhelming them, and scales a brand that feels like a warm, knowledgeable friend — not a corporate sales machine.
8 competitors.
0 of them solved trust.
02 Research
Research covered three distinct tracks: understanding the competitive landscape, mapping user sentiment, and defining brand positioning through direct input from potential users.
I audited eight direct competitors. The findings weren’t about features — they were about failure modes: inconsistent UX, scattered information hierarchies, and over-reliance on price comparison without emotional scaffolding. The gap wasn’t functionality. It was clarity, consistency, and emotional tone.
What they had
Robust platforms, scale, and ad spend. PolicyBazaar alone drove 40,000 daily users; Coverfox ~9,000.
What they lacked
Clarity, consistency, and emotional tone. None had an experience users trusted at a gut level.
Reduce cognitive load at every decision point
Design for a human being, not a transaction
A system that doesn’t confuse across flows
Make it impossible to get lost
Rather than jumping to feature preferences, I ran a brand-perception survey to surface emotional associations. It covered:
The emotional map was most revealingMost users associated insurance with frustration, indifference, and doubt — not security, care, or confidence.
Five distinct logo directions were presented across typography, mark style, and brand personality. User input informed the final brand decision — not the other way around.
The key tension: younger users were comfortable online but didn’t trust the process; older users trusted the process but feared going digital.
Valued comparison features. Worried about data security. Wanted T&C clarity.
Bought offline for decades; a helpful consultant changed the equation. Wants faster claims and fairer settlements.
Prefers online for savings but felt offline agents gave better service. The trade-off felt unresolved.
Valued transparent policy exploration. Distrusted agents. Found large-insurer premiums prohibitively high.
03 Key insights
Users arrive skeptical. Tone — visual, verbal, interactive — must signal safety before asking for any input. First impressions are functional, not cosmetic.
Everyone offered comparisons; none explained the so-what. The UI must act as an intelligent advisor, not a data dump.
Over 50% of target users access financial services on mobile exclusively. Mobile-first was a constraint, not a preference.
Positive experiences used emotional language — “I felt cared for.” Warmth needed to be woven into the logic, not just the visual skin.
Users didn’t distrust human guidance; they distrusted self-serving guidance. The opportunity: replicate the good agent, digitally.
Willingness existed. The blocker was execution: convoluted flows, ambiguous T&C, and weak visual hierarchy eroded intent.
04 Product strategy
I used these as actual filters — things to argue against in design reviews, not decoration on a slide.
Principle 1
Users were already disoriented by competitors. Any inconsistency would erode fragile trust. Novelty was a liability; reliability was a feature.
Decision impactSingle navigation pattern across all flows. Components behaved identically across screens.
Principle 2
Insurance is heavy with complexity. The worst response to complexity is more UI. Every element must earn its presence.
Decision impactStripped the quote comparison screen to the three most decision-relevant data points per policy.
Principle 3
Most users would never open a desktop browser. Designing for 375px first forced better prioritization.
Decision impactAll flows designed and validated on 375px screens before any desktop consideration.
Principle 4
Insurance involves significant user anxiety. Assumptions are expensive.
Decision impactBrand survey, logo testing, and interviews all preceded visual design. No screen was final until tested.
Principle 5
The category default was clinical and cold. An emotionally engaging experience is not soft — it converts.
Decision impactFirst-person warm micro-copy. Onboarding framed as a conversation. Expressive iconography.
05 Information architecture
GroupBima served the Customer — a first-time or switching buyer, anxious and skeptical — and the POS Agent, managing customers, commissions, recruitment, and claims. The same platform had to feel simple to a nervous buyer and powerful for an agent managing a portfolio: a B2B product hidden inside a B2C shell.
Each gate was a trust checkpoint. Dropping out at medical disclosure was the highest-risk moment — users feared being penalized for honesty. The IA reduced this anxiety through progressive disclosure and transparent copy at every step.
The B2C shell
Customer
A first-time or switching buyer, anxious and skeptical.
The B2B product inside
POS Agent
Managing customers, commissions, recruitment, and claims. Fully designed end-to-end, with 15+ flows documented.
Who’s covered
Drop-off risk: privacy concern
Framed as personalisation
Guided, not overwhelming
Trust gate → conversion
* The agent portal was also fully designed end-to-end (15+ flows documented).
06 Design exploration
Fifteen directions were explored across typography, symbol style, and brand register. The winner: an elephant mark in brand crimson, set in a rounded square, with the wordmark “GroupBima” and “Connect to Happiness.” Elephants carry deep symbolic weight in Indian culture — wisdom, loyalty, good fortune, familial bonds, and protection. All map directly onto what insurance should feel like: a protective presence you can trust. Strategic, not decorative.
Nurturer. Communitarian. Maverick.
The deliberate choice of cool colors (deep teal, deep crimson) over the category-default orange was a differentiation signal, not just an aesthetic one.
07 Design system
#042A37#F7F4E8#980042#760933Typeface — Lexend, chosen to reduce visual stress & improve comprehension
Heading — 32/40, Bold
Body — 16/24, Regular
CAPTION — 12/18, MEDIUM
The primary decision surface — insurer logo, coverage, premium, top benefit, CTA. Comparable side-by-side on 375px.
Top-of-screen progress for multi-step onboarding. Signals “you’re almost there,” reducing drop-off.
Checkbox multi-select with age input, built for single-thumb operation.
Pre-existing condition disclosure — the most carefully considered micro-copy in the product: non-judgmental and transparent.
Persistent, one-tap human support on quote screens for users who stall before purchase.
08 Final experience
Journey 1 — Onboarding
Get a user from app open to a personalised quote in under 3 minutes by reframing onboarding as a conversation.
Key moves
Journey 2 — Discovery
Transform raw inputs into options the user understands and trusts.
Key moves
Journey 3 — Trust gate
Move from intent to purchase without losing the user to anxiety.
Key moves
“We promise to respect your privacy and not bother you.”
09 Outcome
No hard conversion metrics were available at project close — GroupBima was an early-stage product moving from brand to MVP. But the outcomes were real.
End-to-end system for two user types — consumer onboarding and discovery, plus a full agent portal (commission, servicing, payment, recruitment, training).
A visually distinct, emotionally differentiated identity — validated through user surveys before a single screen was designed.
Addressed every structural weakness identified in research: inconsistency, scattered information, cognitive overload, and trust erosion.
Qualitative signal: research surfaced a consistent finding — users who wanted to buy insurance online were blocked by bad experiences, not bad intent. The product was designed to remove those blockers, one flow, one screen, one word at a time.
10 Reflection
Starting with brand before interface was the right call — it prevented UI that looks good in isolation but feels disconnected from what the product stands for. Running the brand survey before locking any visual direction gave every downstream decision a foundation in real user data. The dual-user architecture forced systems thinking that made the product more durable and surfaced edge cases in the consumer flows.
Trust is a design material — not just a principle. Every interaction pattern, every piece of micro-copy, every visual decision either builds or erodes it. In a high-stakes, low-trust category, the designer is an architect of credibility. I also learned to treat research artefacts as deliverables in their own right — that’s how you build alignment with stakeholders who need the why before the what.
The onboarding still asks for too much before delivering value — I’d add a “show me what’s possible” moment (a price estimate or sample comparison) before collecting any data. The agent portal needed its own design sprint separate from the consumer product. And the wireframe-to-high-fidelity jump was too fast; today I’d build and test a mid-fidelity prototype to keep feedback on the UX, not the aesthetics.
What I’d change
Let people see the product’s promise before asking them to invest in it.
With AI, today
Redesign discovery as a conversation — the right policy in under five exchanges, ranked by best fit, not highest commission.
AI would also reshape the agent’s role: surfacing at-risk customers, flagging claims anomalies, generating renewal nudges, and drafting support responses — shifting the problem from information architecture to interaction design for AI-mediated workflows.