How AptDeco built a trusted furniture marketplace from a Craigslist pain point

Executive overview

Selling furniture on Craigslist is unreliable, sketchy, and logistically painful. Reham Fagiri and Kalam Dennis validated a simple fix — add managed pickup, delivery, and payments — and saw demand immediately. They built AptDeco, a New York-based peer-to-peer furniture marketplace, through YC W14.

Their edge came not from tech sophistication but from practical operator instincts: structured team management, rigorous analytics, and a logistics model built from scratch that avoids warehousing entirely.

The core insight: solving the full transaction — not just the listing — is what Craigslist never did, and that gap is the business.

Finding the idea and validating demand

• Both founders had the same bad experience selling furniture on Craigslist at the same time
• Early test: copied existing Craigslist listings and added "free delivery" to measure response — demand was immediate and unambiguous
• Furniture was the fifth most popular category on Craigslist — the scale of the opportunity was already there
• Got a transaction on day one of launching a bare-bones site
• YC acceptance was the psychological trigger to commit fully — one founder quit before the interview, the other declined a new job offer

Leaving stable careers

• Reham was happy at L'Oreal and on an upward trajectory — the fear of leaving was real, not manufactured
• Key mental unlock: both were confident they could return to corporate roles if the startup failed
• Kalam had already given notice before the YC interview, financing early development while moonlighting
• Family skepticism was common — immigrant parents equated stability with success; entrepreneurship read as recklessness

What business school actually gave them

• Financial modeling: building projections and scenario planning, used every day at AptDeco
• Analytical structure: knowing how to set up data foundations to evaluate what works and what doesn't
• People management and conflict resolution: structuring feedback, managing performance, knowing when and how to let someone go
• Creating operational structure early — weekly meetings, check-ins, KPI reviews — brought stability to a chaotic early-stage environment

Team management and performance systems

• Weekly one-on-ones with direct reports, focused on strategic projects tracked in a shared Google Sheet
• Daily stand-ups to keep the team engaged and aligned
• Weekly KPI meeting reviewing sales, returns, reviews, and other top metrics
• Quarterly 360-degree performance reviews (scale of 1–5, anonymous, 10 questions) — founders' results shared publicly with the team
• The firing standard: by the time someone is let go, they should already know it's coming — if it's a surprise, the manager failed
• When a role becomes redundant, involve the person early and give them time to find alternatives

Pricing and margins

• Initial transaction fee was 10–15%; now 19–29% after properly understanding cost structure
• Early negative margins were partly from not knowing all the cost variables, partly from a third-party delivery provider who kept cancelling jobs for higher-value clients
• Brought delivery in-house and ran A/B price testing to find the right balance between delivery fee and conversion rate
• Lesson: don't optimise for margin before product-market fit, but do build the foundation to measure it accurately once you have traction

The no-warehouse logistics model

• AptDeco picks up from the seller and delivers to the buyer the same day — no storage, no warehouse
• Decision was intuitive from the start: warehousing caps scale by physical space and pushes costs onto already price-sensitive buyers and sellers
• Multiple competitors tried the warehouse model and failed
• Proprietary scheduling software built from scratch — existing routing tools didn't fit the model
• Head of operations eventually hired from furniture manufacturing and supply chain; earlier hire would have helped

Customer acquisition

• First major channel was New York City subway advertising — 1 in 6 cars, one month, under $50k, traffic tripled
• Subway ads had a halo effect on other channels; people assumed AptDeco was much larger than it was
• Digital advertising took 3–4 years to figure out; early agency-managed campaigns failed
• Breakthrough came from revisiting old Facebook/Instagram campaigns, finding one testimonial ad that had performed well, and rebuilding from there
• Current acquisition mix: ~45–50% word-of-mouth referral, plus Facebook, Instagram, and Google
• Brand-driven SEO is significant — 65–70% of inventory is from seven key brands (IKEA, West Elm, CB2, Crate & Barrel, Room & Board, ABC Carpet & Home, Restoration Hardware)

Customer profile and marketplace dynamics

• Sellers: 30–55, higher disposable income, often making life transitions (new baby, moving); the "Upper West Side couple" archetype
• Buyers: 25–mid-30s, aspirational, first or second job; want West Elm or Restoration Hardware quality at IKEA prices
• Growing signal: sellers are becoming repeat buyers — indicates a shift in how people think about shopping secondhand
• New York's car-free culture makes delivery more valuable and buyers more willing to pay for it

Analytics and data infrastructure

• Early analysis was done via SQL queries — slow and dependent on technical staff
• Implemented Looker as a BI platform; transformed the team's ability to act on data
• Now tracks everything: clicks, usability sessions, customer journeys, internal search terms, landing page performance, unit economics per transaction
• Visibility into brand-by-category performance, time-to-sell, drop-off points, and cart abandonment drove both marketing and product decisions

What didn't work

• Photography service for sellers: Airbnb-style professional photos — no demand
• Cleaning services offered at a discount: negotiated deals, but customers wouldn't pay even at reduced rates
• Early Facebook/Instagram campaigns managed by an agency: failed across 15 campaigns before one testimonial ad eventually worked

YC advice that mattered

• "Talk to your customers" — not customer service, but getting in front of them directly; flying back to New York weekly during the batch was essential
• Take every suggestion from partners seriously, no matter how unscalable it sounds, and report back the following week with results
• Create structure with the partners: recurring weekly meetings, prepared agendas, tracked action items — manage up
• The group office hours revealed that all companies share the same problems — normalising the chaos