PW

Product Strategy • UX Research

Crowd Meter

Project summary

Crowd Meter was a semester-long group project built around a single prompt: design an app that solves the problem of overcrowding. Working in a team of five as research lead, we took a fully research-driven approach — running secondary market research, user interviews, and a customer discovery survey before writing a single line of product spec.

The result was a full go-to-market pitch: a freemium mobile app that lets individuals check real-time crowd levels at any venue, while offering businesses a $99/month analytics tier to forecast demand and optimize operations. What made the project distinctive was the dual-audience model — one data network serving two very different customers — and a privacy-first architecture that addressed the core failure of every existing competitor.

  • Team of 5
  • Research lead
  • Market research
  • User interviews
  • Customer discovery survey
  • Business model
  • App pitch

At a glance

  • The challenge

    No existing tool gives people reliable, real-time crowd data without compromising their privacy. We were tasked with designing a solution from scratch, validated by real research.

  • Our approach

    Three rounds of research — desk research, qualitative interviews, and a quantitative survey — informed every product and business model decision before we pitched.

  • The outcome

    A fully scoped app concept with a privacy-safe data model, freemium pricing validated by survey data, and a dual B2C/B2B revenue strategy grounded in market research.

The problem

Overcrowded spaces cost people time, cause stress, and create real health risks — yet no product reliably tells you how busy a place is right now. Existing tools like Google Maps' "popular times" are based on historical data, not live feeds, and current crowd-tracking methods routinely compromise user privacy.

Current methods of collecting crowd data make it easy to identify and track individuals, so privacy has become one of the primary issues in crowd meters and crowd tracking.

Research finding (Koerhuis)

Research process

A three-phase discovery process

Before building anything, the team ran a structured research sprint to understand the problem space, validate assumptions, and find a solution grounded in real user needs. The process moved from broad secondary research to direct customer contact to quantitative validation.

  • Phase 1 — Secondary market research

    Mapped the competitive landscape, technology trends, and market size to understand what already existed and where the gaps were.

  • Phase 2 — User interviews

    Conducted one-on-one interviews to surface real behaviors, attitudes toward privacy, and what crowd information people actually want.

  • Phase 3 — Customer discovery survey

    Deployed a quantitative survey to validate interview insights at scale and test willingness to pay for premium features.

Methods in detail

Secondary market research

Desk research on industry trends, competitors, and the regulatory landscape.

What we examined

  • Existing products: Planet Fitness, Placemeter, Traf-Sys, SenSource, FootfallCam, RetailNext
  • Google Maps' "popular times" feature and its limitations
  • Privacy regulations and trends around location data collection
  • Crowd analytics market size and growth projections

Key outputs

  • Competitive matrix mapping each player's strengths and weaknesses
  • Identified the core tension: accuracy vs. privacy in crowd sensing
  • Market sized at $1.76B in 2023, projected to $12.48B by 2032
  • Confirmed no full-coverage, privacy-safe crowd meter exists

User interviews

Qualitative conversations to explore attitudes, behaviors, and mental models around crowds.

Interview goals

  • Understand how people currently deal with crowded spaces
  • Probe comfort with contributing anonymous location data
  • Identify what crowd information feels useful vs. intrusive
  • Uncover unmet needs current tools don't address

Notable findings

  • Most participants had never heard of any crowd meter
  • People prefer crowd data from a dedicated app, not embedded in maps
  • Unique request: not just headcount but crowd character (hostile, drunk, etc.)
  • Most people uncomfortable contributing data — privacy is a real barrier

Customer discovery survey

Quantitative survey to validate patterns found in interviews and size the opportunity.

Survey goals

  • Quantify how often people check for crowds before going somewhere
  • Identify which venue types matter most to users
  • Test willingness to contribute anonymous data
  • Gauge willingness to pay for a premium tier

Key results

  • 57% sometimes check for crowds; 14% always do
  • Top venues: grocery stores, gyms, restaurants, transit, libraries
  • 57% would pay $4.99/month for premium features
  • 57% willing to contribute anonymous location data
  • Users want wait times, real-time busyness, and safety checks

Cross-method findings

By triangulating across all three methods, we identified consistent signals that shaped the final solution.

  • Privacy is non-negotiable. Both interviews and secondary research confirmed that camera-based and always-on tracking are dealbreakers. Any solution must be opt-in and anonymous by design.
  • The awareness gap is a real opportunity. Most interview participants didn't know crowd meters existed — yet survey data showed latent demand. The market is undereducated, not disinterested.
  • Context matters more than raw numbers. Users want actionable information: is it worth going now, how long will I wait, is it safe? This shaped the "crowd character" features.
  • Businesses are an underserved secondary customer. Competitors like RetailNext focused on B2B analytics, but no solution bridged consumer and business value in one product.
  • Optional contribution unlocks supply. 57% would allow passive anonymous sharing — enough to build a viable data network if privacy is clearly communicated upfront.

Research to solution

Each research method directly shaped a product decision.

  • Secondary research → tech stack

    Camera and GPS tracking backlash led us to Wi-Fi signal detection and edge-processed sensor data as the core collection mechanism.

  • Interviews → feature set

    The request for crowd "character" and venue-specific detail directly informed the granular 1–10 score and venue-level breakdowns.

  • Survey → business model

    57% willingness to pay validated the freemium model. The $4.99/month price point matched stated willingness from respondents.

  • All three → dual audience

    Research confirmed individual users and businesses have different but compatible needs, enabling the B2C + B2B model.

Final solution

How Crowd Meter works

Crowd Meter provides a privacy-first solution for users looking to avoid crowded spaces and for businesses aiming to understand and manage foot traffic. By combining anonymous location tracking, Wi-Fi signal detection, and optional IoT sensors, it delivers real-time crowd insights without compromising user privacy.

  1. Data collection

    Passively collects anonymous data through optional location tracking, Wi-Fi signal detection, and IoT sensors at partnered venues. No personal data — only volume.

  2. Aggregation and anonymization

    All data is processed at the edge or in the cloud with no identifying information stored. Users can toggle data sharing on or off at any time.

  3. Crowd level estimation

    Algorithms categorize crowd levels into Low, Moderate, and Busy — plus a granular 1–10 score to guide decisions.

  4. Display to users

    Search specific locations, browse nearby spots, or set alerts for when a space hits your preferred crowd level.

Market research findings

  • Market

    No full-coverage solution exists

    Existing tools are inaccurate (Google's estimates), limited to one venue (Planet Fitness), or camera-based with ~80% accuracy and major privacy concerns.

  • Privacy

    Privacy is a dealbreaker

    33% of adults and 50% of teens disable location tracking. Any viable crowd-sensing solution must collect anonymous, opt-in data only.

  • Demand

    People want real-time, not historical

    85% of surveyed users want real-time crowd data. They check before going to restaurants, gyms, and grocery stores — but only if data can be trusted.

  • B2B

    Businesses are underserved

    Retailers, gyms, and venues already collect crowd data internally. There's an untapped opportunity to surface this as actionable analytics for operations.

Competitive landscape

  • Crowd Meter

    Our solution

    • Real-time and predictive accuracy
    • Privacy-first, anonymous data
    • Works across all venue types
    • Dual consumer and B2B value

  • Google Maps

    Strengths

    • Large data pool, built-in navigation

    Gaps

    • Inaccurate, delayed, no real-time data
    • Users can't contribute data

  • Planet Fitness / transit apps

    Strengths

    • Accurate and real-time

    Gaps

    • Limited to a single venue type
    • Not scalable across locations

Revenue model

One data network, two customers — the survey validated both price points.

  • $4.99 / month

    Consumer premium

    • Predictive crowd alerts for frequent spots
    • Notifications when a place quiets down
    • In-depth venue stay statistics
    • Free tier for basic real-time levels

  • $99 / month

    Business analytics tier

    • Hyper-specific foot traffic analytics
    • Demand forecasting for staffing and inventory
    • Weather, event and holiday trend integration
    • API and white-label partnership options

Target users

Two primary segments with a shared need for crowd information.

Busy professionals University students Introverts and crowd-averse users Parents managing schedules Health-conscious individuals Grocery stores and gyms Libraries and universities Restaurants and cafes

What I learned

This project pushed me to think like a product strategist, not just a student. A few things that stuck with me:

  • Research changes what you build

    We started with a vague brief about "overcrowding." It was only after talking to real users that we understood privacy was the actual barrier — not awareness or convenience. Without the research phase, we would have built the wrong thing.

  • Qualitative and quantitative need each other

    Interviews told us what people cared about. The survey told us how much. Neither alone would have been enough to make a confident product decision or a credible pitch. Using both together made our findings genuinely convincing.

  • The cold-start problem is a real challenge

    A crowd meter with no users has no data — and no data means no value. Thinking through how to bootstrap the network before launch was one of the hardest parts, and it taught me to think about supply and demand as equally important product problems.

  • Business model design is part of the product

    The freemium vs. B2B split wasn't just a pricing decision — it shaped the whole product architecture. Working through who pays, who benefits, and how those interests align made me think about products as systems, not just features.

  • Teams make better decisions than individuals

    Working across five people with different perspectives meant we caught assumptions we each would have missed alone. Learning to disagree productively and synthesize different viewpoints into one coherent pitch was as valuable as any research skill.

  • Data can validate intuition — or kill it

    Some of our early assumptions turned out to be wrong. Learning to let data override gut feel early is a discipline I'll carry into every future project.