PetPal: a companion-led discovery app for local food, travel & exploration
A local and travel discovery app where collectible pet companions help users find places to eat, explore, and experience, blending recommendations, character interaction, and lightweight gamification into a more personal way to discover the world around you.
Utility and connection, solved together
Local discovery still feels like homework: ideas live in one app, routes in another, and restaurant searches in a third. None of them understand the kind of day you're having.
PetPal combines useful recommendations with a warm, collectible companion that reaches out first, before you even ask.
One family, four scenarios, endless collectibility
Rather than one mascot, a whole pet family covers the range of user preferences, each member has a distinct personality suited to a different scenario. You adopt a main buddy first, Shiba recommended by default, then unlock the rest as you explore.
Mochi
Loyal, warm and reliable, the buddy who always knows the way.
Sugar
Lazy, healing and clingy, made for quiet spots and late nights.
Bao
Soft, dopey and curious, chasing check-ins and photo spots.
Pudding
Round, lovable and hungry, first to find the best local eats.
Three things that set PetPal apart
Apps like Character.AI and Replika are built around the conversation itself. PetPal's companion talks, but its job is to get you out the door somewhere good. The payoff shows up in real life, not just in the chat.
Finch ties a pet's growth to self-care habits, which is great for routine but doesn't help you decide where to eat tonight. PetPal ties that same growth loop to actually useful local and travel decisions, so the companionship has something to do.
Most companion apps give you a single character to build a relationship with. PetPal's four buddies each specialize in a different kind of outing: Mochi for everyday navigation, Sugar for quiet evenings, Bao for exploring new places, and Pudding for finding good food nearby.
Three principles, one companion doing the work
Every feature decision traces back to three principles that keep the product useful without losing its warmth.
The nearby, daily stuff: answers "what's good nearby?" in one tap and quietly becomes your everyday companion.
Kicks in once you leave home, builds a check-in map, points out photo spots and local-only secrets, plans half-day routes.
Surfaces picks proactively, reading time, weather and location as implicit signals rather than waiting for you to type a search.
Tell PetPal what you need, or just open the app
Get 1 to 3 picks, no feed to scroll
See why: taste match, weather, trust score
Visit, check in, and watch your buddy grow
From a generic list to a companion that talks back
The first pass looked like a generic recommendation feed. Testing showed the pet needed to be part of the interaction, so twelve screens mapped the full flow by hand, and the static pet icon became a buddy that speaks up, reacts to taps, grows through XP, and dresses for the moment through a zodiac-themed skin collection.
Clear a mission, claim the look
Each buddy gets its own set of zodiac-themed skins, unlocked through missions, rarity, and progress.
Onboarding to discovery, one continuous flow
Adopt a buddy → set your taste → get trusted picks → check in → grow your pet. It should feel like hanging out with a companion, not filling out a search form. Below: an early prototype of the onboarding-to-home flow.
Choosing a buddy, then landing on the home feed. Click to view larger.
What I'm learning so far
Designing a companion-first product is different from designing a tool: every screen has to answer "what would the pet say here?" as much as "what does the user need to see?" Balancing a genuinely useful recommendation engine against a kawaii, emotional layer, without either one undercutting the other, has been the central design tension of this project, and one I'm still actively working through.
"Does pet-IP companionship plus proactive personalized recommendation significantly outperform traditional search tools on retention and acceptance rate?"
"Moving away from a search tool people compare and get burned by, toward a companion that actually looks out for them."
The pet models already move, but the rigging isn't there yet: right now they only play a handful of canned motions instead of reacting naturally to what's happening.
Where PetPal goes next
Piloted in one NYC neighborhood, built to scale beyond it.
An early pulse-check before scaling
A short concept survey sent to early testers ahead of expanding beyond the first neighborhood, paired with a prototype link so people could click through the flow themselves before answering. The goal was to capture how it actually felt to use, not just whether the idea sounded good on paper. Small sample, but a useful early signal.
Proactive recommendations are the weakest signal in the data, sitting at 40% versus 80% for companionship and consolidation. Closing that gap is the real work for Phase 2.
"Maybe the pets could do some more, they seem to just loop through a few messages that are just there. Since you're using ai you could probably make it more customizable but yeah."
"I'm also not sure if I feel the need to use this app because I already have Google Maps to recommend me some spots and TripAdvisor for activities."
MVP ✓ Concept Validated
Single NYC neighborhood. Buddy + right-now picks + check-in loop. Next: validate retention with real users.
Phase 2 In Progress
Testing and refining with real users: travel mode, smart routes, the check-in map and zodiac skins are all in that loop now.
Phase 3
Deeper personalization: taste profiles that get smarter with every check-in, more cities and membership, plus a voice interaction pilot.
Phase 4
IP collabs & merch, open ecosystem, real-visit data moat, plus better rigging for live character reactions.
How it sustains itself, later
Retention comes first, and monetization follows once the core loop actually works: zodiac skins you can earn through missions or just buy outright if they're rare, merchant listings with a redemption share, and brand or IP collabs with food and beverage, travel, and pet merchandise partners. The goal is for revenue to grow out of that trust, not compete with it, so PetPal doesn't get commercialized before the core experience has proven itself.
Want the full plan? Let's chat →