03 Research & Strategy

Research & Strategy

Research is not about proving how much work you have done — it is about helping the product make fewer wrong turns.

This section includes competitive analysis, content-led growth research, and product judgment around onboarding and activation. Every study must ultimately answer three questions: what is worth doing now, what should wait, and what to validate next.

Petch Competitive Analysis

Independent product research · Competitive strategy · Pet social category

Petch seems to have everything: matching, maps, messaging, events, AI assistance, rewards, and subscription monetization. What really holds it back is not a lack of features, but that these features arrive too early.

Users are asked to fill out information, understand rules, navigate incentive mechanics, and even face payment before they have figured out whether the product is worth staying for. The more features pile up, the harder the core experience becomes to see.

Core Diagnosis

Petch did not lack features. Its feature set moved faster than user trust could develop.

Product maturity is not about completing features as early as possible, but about making each feature appear at the moment users are ready to accept it.

Main Issues I Found

Product Path

Complete, But Not Clear

Petch has many feature layers, but once inside the product it is hard for users to quickly understand what to do next and which path is the core experience.

Onboarding

Input Before Value

The product collects too much information at an early stage, increasing the mental and operational burden of first use.

Monetization

Monetization Before Trust

Ads, subscriptions, and paid benefits appear too early. Users are asked to pay for value they have not yet felt.

Incentives

Gamification Mismatch

Points, rewards, and badges can turn relationship-building into task-completion, weakening the light, natural community experience.

Matching

Swipe Logic Assumptions

Pet social matching should consider distance, personality, play style, energy level, and offline feasibility — not just swipe speed.

Product Tone

Functional, But Cold

For a product that helps users build local, low-pressure relationships, whether it feels safe, natural, and low-pressure matters as much as whether it works.

Pawly's Opportunity

Light

Reduce the operational and mental friction of first use as much as possible.

Smooth

Organize the experience around a natural core path: map discovery → event participation → offline meetup → repeat connection.

Warm

Make the product feel more like a trustworthy, friendly local community than a cold match or transaction.

Restrained

Validate core behavior and local network density before gradually introducing more complex features and monetization.

Read Full Competitive Analysis

A deeper PDF analysis covering product rhythm, onboarding burden, monetization timing, gamification mechanics, matching logic, product tone, and strategic implications for Pawly.

Pawly Content & Account Research

Pawly · Product Management Intern · Growth & Strategy

Content-led growth research · Creator pattern analysis · Account archetype strategy · Activation hypotheses · Brand positioning

I systematically reviewed the pet content ecosystem and representative accounts, then mapped different content formats to specific Pawly product goals: which content drives attention, which builds trust, which pushes profile completion, and which helps users understand the product.

Core Thesis

Pawly should not chase pet-content traffic alone.

The more important role of content is to gradually turn pet-parent emotional resonance into product understanding, trust, and actual usage.

Content Types I Focused On

Storytelling / Meme

Easier to gain exposure, create resonance, and be shared at low cost.

POV / Educational

Build professional trust through concrete, practical scenarios while explaining the problem Pawly solves.

UGC Prompts

Lower participation barriers, making it easier for users to contribute and gradually form sustainable community interaction.

Pet Personality

Drive profile completion and identity formation, bridging content browsing to product activation.

Strategic Insight

Storytelling content makes users see Pawly. POV and educational content help users understand and trust Pawly. Pet personality content further carries profile completion and identity, while UGC topics help the community build sustained participation.

Pawly's Content Strategy

Hover a card to see an example — click to keep it flipped.

Visibility

Create highly relatable, shareable content around the everyday emotions, awkward moments, and small frustrations of pet ownership.

Example

“Taking your dog to socialize for the first time—and realizing you’re the socially anxious one.”

Use familiar pet-parent moments to create emotional resonance and reach people already experiencing the need Pawly addresses.

Trust

Use concrete scenarios to show how pet compatibility, meetup safety, and low-pressure social interactions can work in practice.

Example

“Your first meetup doesn’t need to be a big dog party.”

Show how a first interaction can begin with a nearby park, a small group, and a short meetup, making offline socializing feel safer and easier to picture.

Activation

Use pet-personality content to move users naturally from consuming content to completing their profile, then toward identification with their Pawly Type.

Example

“Is your dog a Social Butterfly or a Quiet Observer?”

Turn a personality-themed post into an entry point for the Pet Personality Assessment, then translate the result into a Pawly Type that becomes part of the pet’s profile.

Community Loops

Design low-friction UGC prompts that turn small moments from everyday pet life into repeatable community participation and interaction.

Example

“What did your pet do today that only you found ridiculously funny?”

A photo and one sentence are enough to participate, while comments and responses invite other pet parents to contribute their own versions and keep the loop going.

Read Full Content Research

A deeper PDF appendix covering account archetypes, creator behavior, content-led growth strategy, activation hypotheses, and implications for Pawly brand and product positioning.

03 Data & Behavioral Research

Behavioral Analytics Capstone Poster Preview

Data & Behavioral Research

Findings & Practical Implications

Depression risk is shaped not only by individual symptoms, but also by economic conditions, family relationships, and access to social support.

In this study, greater financial stability and stronger social ties were more often associated with lower risk. Financial strain, relationship disruption, and barriers to communication or support appeared more frequently among higher-risk groups.

These findings indicate association rather than causation. Still, they suggest a clear direction: mental health risk assessment should look beyond symptom scores and consider the person’s broader life context.

Potential Directions for Action

— Monitor the sustained impact of financial strain and major family changes
— Offer support or referral earlier when social connection begins to weaken
— Reduce barriers to seeking help rather than waiting for risk to become severe

Data & Product Research

Behavioral Analytics Capstone: Depression Risk Prediction

This study combines actigraphy and survey data to analyze behavioral signals related to depression risk, and to compare what each data type is best suited to answer:

One dataset is minute-level actigraphy from 55 participants — small in sample size but rich enough to capture activity rhythms, day-night variation, and sleep proxies. The other is 4,836 NHANES survey records, coarser in behavioral dimensions but better suited for establishing a stable benchmark. I aggregated minute-level records into daily and individual-level features, then engineered and selected features around daytime activity, behavioral variability, relative amplitude, activity peak timing, and nighttime inactivity.

What This Shows

— Transforming raw behavioral records into rhythmic and activity features usable for modeling
— Identifying temporal signals from circadian rhythms, behavioral variability, and sleep proxies
— Removing redundant variables and reducing the feature space from 108 to 42 dimensions, and from 70 to 54 dimensions
— Judging the appropriate uses and limitations of small-sample high-dimensional behavioral data versus large-sample population survey data

Open PDF

Academic capstone output.

All materials shown have been reviewed and limited to publicly shareable, non-confidential content, while preserving the original product reasoning, documentation structure, and decision logic as fully as possible.

Other product experience and team collaboration results can be found in the resume. Some internal documents are not displayed due to confidentiality requirements.