Insomnia Cookies: building a $350M late-night delivery brand from a dorm room

Executive overview

Seth Berkowitz started delivering cookies from his college house at Penn in 2003 because late-night options were pizza and Chinese food — nothing else. The idea was simple: warm cookies at your door in under 30 minutes. The execution took nearly a decade of losses, failed pivots, and a two-year financial crisis before the model clicked.

The core insight: strip everything back to the original promise — warm cookie, fast delivery, late-night focus — and the business works.

Delivery speed and product warmth are the product; everything else is a distraction.

Early growth and the frozen yogurt mistake

• Started with $100 in equipment, a cell phone, and a notebook to take orders
• Front-page article in the Penn student paper took orders from 5 a night to 80 overnight
• Partnered with co-founder Jared Barnett to cover weekends and expand
• Feared retail-only cookies would fail (Mrs. Fields, Great American Cookie had both struggled)
• Co-located Insomnia inside Tasty Delight frozen yogurt shops — gave away the brand to a fad business
• Seasonal college model left stores empty over summers and holiday breaks
• Yogurt worked in summer; students weren't there in summer — a fundamental mismatch

The pivot back to cookies

• Converted the Champaign, Illinois location to cookie-only in mid-2006: first profitable store
• Co-founder Jared departed; yogurt locations closed one by one
• Realized small footprints (250–500 sq ft) kept costs manageable
• Opened NYC Greenwich Village store — first city-only, non-campus location; immediate success
• Real estate rule: late-night areas only — bar and club districts

The crisis years (2009–2011)

• Committed to 10 food trucks; city ordinances and mechanical failures made them unviable
• Lehman collapse shut off investor capital entirely
• Cut team to himself and one finance associate; stopped paying himself for 18 months
• Drove cookie dough deliveries personally across the Northeast
• Moved office into a 50 sq ft closet in an investor's back office
• First child born March 2010 — forced clarity: figure it out or fold

The turning point

• iPhone App Store (2010 in earnest) transformed on-demand delivery from niche to mainstream
• Capital returned in 2011; one trusted investor convinced by a third party to cut a final check
• Store 22, Kent, Ohio (2012): first location funded entirely from cashflow — no outside capital needed
• Next 120 stores opened the same way, straight from profits
• Reached self-sustaining profitability by 2013

Scale and the Krispy Kreme deal

• Resisted third-party delivery (DoorDash, UberEats) until COVID forced the shift
• By 2016: 100 locations, profitable, growing from internal cashflow
• Late 2017: saw operational cracks — delivery times slipping, innovation slowing
• Hired an investment bank; met ~60 potential buyers
• Krispy Kreme (via JAB Holdings) purchased ~75% for $140M in 2018
• Partnership added financial discipline; helped Seth transition from operator-founder to CEO
• COVID hit just as they settled in — college campuses emptied, sales dropped ~50% overnight
• Extended delivery zones from 3 to 5 miles; sales recovered and exceeded prior levels
• 2024: Krispy Kreme sold majority stake to two private equity groups; Seth remains CEO

Lessons on model and competition

• Corporate-owned over franchise: preserved delivery as core competency, avoided the Mrs. Fields trap
• Crumble grew to 900+ franchise locations fast; Seth sees it as a different product (decorated, daytime) not a direct competitor
• Warm, fast, late-night delivery is the defensible position — hard to replicate in a franchise model
• Decisions driven by hope for 15 years; Krispy Kreme partnership introduced data discipline
• Healthy paranoia remains: monitors sales by hour, course-corrects constantly