Sri Mandir is India’s leading devotion platform — bringing temple rituals, chadhava and pujas to millions of devotees on their phones. Here are three projects where I worked across the funnel: building trust in an unseen service, organising festival worship into one destination, and turning a coin system into a real habit loop.
Teaching users what an online puja really includes — and converting chadhava devotees into puja bookers.
↗One destination for every tithi: special pujas, chadhava, kathas, daan and guides — instead of scattered discovery.
↗Designing earn–burn loops around digital assets to convert engagement into measurable retention.
Devotees were happily offering chadhava — but most had no idea what Sri Mandir’s Puja service actually included. I designed an education layer across the funnel that explains the ritual, builds trust, and moves users from awareness to booking.
A puja is an experience the user never witnesses in person — the Sankalp taken in their name, the full vidhi performed by pandits, the recording, the Prasad (Aashirwad box) shipped home. When none of this is explained, users default to what they know: they confuse Puja with Chadhava, treat the higher price as unjustified, and never engage with puja listings.
Only 2.8% of chadhava-transacting users activated on the Puja service within 30 days.
Puja L1 page conversion for non-activated users sat at a low 0.6%.
Puja L1 → L2 click-through was just 17% — users browsed but didn’t go deeper.
No single place in the app answered: “What exactly happens when I book a puja?”
Instead of pushing harder on listings, we built an awareness system: a dedicated Puja Awareness page that narrates the experience (Sankalp → Vidhi → Recording → Prasad delivery), surfaced through a one-time bottom-sheet journey and persistent banners across the surfaces where confused users already were.
On a user’s first visit to the Puja tab, an awareness bottom sheet slides up after a short, backend-configurable delay. Scrolling or tapping anywhere expands it into a full-screen story; the hero video plays inline with play/pause and mute controls.
A scrollable narrative built from modular sections — designed so the product team can iterate without an app release:
The page is only useful if confused users find it. We placed configurable entry banners exactly where the puja–chadhava confusion lives — shown only to users who haven’t transacted on puja yet (cohort logic configurable by transaction count).
I worked with the PM to define the full instrumentation — impression and action events for the bottom sheet, banner sources, 3s/25s video-view depth, and FAQ interactions — plus a guardrail: if App-open → Puja L1 conversion dipped, the bottom sheet would be rolled back.
Trust is a design material. For services users can’t physically witness, showing the operation — temples, pandits, the Aashirwad box — does more than any discount.
Design for iteration speed. Making the hero fold HTML-configurable meant the team could test narratives weekly instead of waiting on releases.
Guardrails make bold bets safe. Defining a rollback metric upfront let us ship an interruptive surface (the bottom sheet) with confidence.
On every special tithi, the app celebrated — festival darshans on the temple page, kathas in the library, special chadhava pushed through banners. But there was no single place where a devotee could find everything for that occasion. So we built one.
On a festival like Ekadashi or Mahashivratri, a devotee’s intent is singular — “What should I do today, and where?” But our answer was fragmented:
The temple page showed festival darshans and pushed chadhava — but only chadhava.
Special pujas for the tithi had to be searched for manually in listings.
Kathas, rituals and how-to guides for the occasion sat buried inside the Library.
Daan opportunities relevant to the festival had no discovery surface at all.
High festival intent was leaking out of the app simply because the content was never assembled in one place.
Explore brings together everything for the current tithi or festival: special pujas, chadhava offerings, daan, kathas, step-by-step puja guides and downloadable guide books — organised by what a devotee does first, next, and after.
For major festivals, we experimented with an animated hero section — turning the top fold into a moment of celebration rather than a static banner. It set the emotional tone and pulled users into the collection below.
Even the app logo adapted to the festival — a small but loved detail that made the whole product feel alive on auspicious days and signalled “something special is happening today” before a single tap.
Each collection wove devotional content (katha, vidhi, significance) directly into booking entry points — so learning about the occasion and acting on it became one continuous journey instead of two separate apps’ worth of behaviour.
Organise around intent, not inventory. Users think in occasions (“today is Ekadashi”), not in product categories — the IA should mirror that.
Delight compounds trust. The dynamic logo and animated heroes were “small” work with outsized emotional payoff for a devotional audience.
Content is the top of the commerce funnel. Kathas and guides weren’t engagement filler — they primed the booking decision.
Sri Mandir’s coin — the Punya Mudra — had a problem every gamified product fears: users earned endlessly, spent nothing, and the currency lost meaning. I worked on the engagement-side burn loop: making digital asset unlocks feel earned, celebrated, and worth coming back for.
Massive coin accumulation had eroded the perceived value of the Punya Mudra — earning felt meaningless.
Unlocking a digital asset didn’t actually deduct coins — it was a filter, not a transaction. All assets could be unlocked with near-zero effort.
When users became eligible to unlock something, the app never told them. Adoption sat at just 10% within 15 days.
Unlocking had no moment of delight — and only 18% of users ever unlocked again the following week.
The moment a user qualifies to unlock their next specific asset, a nudge appears on the temple screen — capped at once per day so multiple eligibilities never become spam. Tapping anywhere on it lands users with the ready-to-unlock item pre-selected, removing every step between desire and action.
An eligibility event triggers MoEngage push notifications with deep links straight into the unlock state — recovering the “lost opportunity” cohort who earned coins but never returned to spend them.
Unlocks now actually burn Punya Mudra. Scarcity returned meaning: each asset became something saved for and chosen, with the passbook reflecting every deduction transparently.
A confirmation pop-up turns the unlock into a moment of positive affirmation — making assets aspirational and seeding the next loop. Edge cases like coupon + coin stacking, refunds and subscription exclusions were resolved in design before build.
A reward without a cost isn’t a reward. Adding coin deduction made unlocks scarcer — and far more desirable.
Correlation is a starting gun, not a finish line. We used the retention correlation to justify a fast MoEngage experiment for causal signal before heavy investment.
Habit loops need closure. Signal, spend, and celebration each existed in isolation before — retention came from connecting them into one loop.
Across awareness, discovery and retention, the common thread was the same: meet devotees where their intent already is, make the invisible visible, and let delight do the heavy lifting. I’d love to walk you through the full process decks for any of these projects.