How Care.com used data at every stage to scale

Executive overview

Most founders collect the wrong data, at the wrong time, at the wrong cost. Sheila Lirio Marcelo built Care.com into a public company by doing the opposite: gathering cheap, targeted data at each phase and acting on it fast.

The core discipline is matching data ambition to company stage. Cheap signals early, richer instrumentation later.

No data, no scale — but not all data is created equal.

Why most early data misleads

• Friends and family surveys return false positives; they want you to succeed.
• Surveying the wrong population (e.g., diabetics for a cake pop business) is equally worthless.
• Commissioning city-wide research before product-market fit is expensive and slow.
• Useful early data is cheap, fast, and targeted to the actual decision you need to make.
• Asking smart advisors "what's wrong with my idea?" is the best-value data available to a pre-product founder.

How Sheila found her market before building

• Observed which personal services were moving online — job boards, dating, social media — as a proxy signal that care could follow.
• Hired 20 college interns to analyze Craigslist listings across U.S. cities: no complex technology, results in weeks.
• Used that analysis to identify the top 20 metros where supply and demand for caregivers was strongest.
• Mapped the ratio of caregivers to families needed to produce real hiring outcomes (~12–15 caregivers per family to generate 3–4 interview choices).

Building the marketplace with data guardrails

• Launched free in target metros, then activated payment only once caregiver-to-family ratios hit the threshold — turning data into a product trigger.
• Multiple verticals (childcare, senior care, pet care, tutoring, housekeeping) provided insurance against misreading any one market.
• Childcare timing was right; senior care was too early by a decade — having the other verticals carried the company through.
• Free-to-paid sequencing prioritized marketplace liquidity and data accumulation over early monetization.

The scale customer is not the same as your customer

• Sheila's own "sandwich" experience (simultaneous childcare and elder care) motivated the company but didn't define the scale market.
• Personal pain is valid signal for a problem worth solving; it is not a substitute for user data about who will drive volume.
• Senior care demand existed but the consumer mindset wasn't ready — user data revealed the gap that personal conviction had obscured.
• The sooner that picture comes into focus, the sooner you can redirect resources.

What Care.com missed — and why it matters

• Mobile and on-demand behavior shifted faster than Sheila anticipated; a people-services business couldn't pivot as quickly as a pure software platform.
• The lesson: track macro technology trends (mobile, cloud, AI) as leading indicators, not just user metrics.
• Companies that adjusted to mobile thrived; those that didn't suffered — the pattern held across many tech businesses of that era.

Data on both sides of a two-sided marketplace

• Empowering families by exploiting caregivers would have destroyed one side of the business and contradicted the social mission.
• Sheila partnered with the National Domestic Workers Alliance to learn fast what caregiver protections to build in — using existing research rather than starting from scratch.
• Practical outcomes: platform refused listings below minimum wage; sweep accounts linked to MasterCard let caregivers use earnings for groceries and pharmacy; Stride Health partnership enabled healthcare access.
• Caregiver retention is supply-side product quality. Treating it as "extra" investment is a false economy.

Applying the same principles beyond for-profit companies

• Art Fields, Lake City SC: commissioned an economic impact study in year one; gathered business-by-business sales comparisons annually. Result: $5.4 million in documented economic impact from the first event.
• The Asian American Foundation (TAF) used for-profit, scrappy, data-driven principles inside a non-profit structure to accelerate launch.
• Non-profits must track impact data as rigorously as for-profits track revenue — the measurement discipline is the same, only the metric changes.