How to shut down a startup with integrity: lessons from Poppy's founder

Executive overview

Most startup shutdown advice focuses on the decision itself. The harder part is everything after: telling investors, letting go of your team, and preserving your identity when the company is gone.

Avni Patel Thompson built Poppy — a childcare marketplace in Seattle — for three and a half years before shutting it down in December 2018. The economics of the childcare category were margin-challenged from the start, and the unit economics shifted at every scale step.

The core insight: shutting down well is a skill, and founders who practice transparency throughout make the hardest conversations far easier.

Why the economics never worked at scale

• Childcare is inherently margin-challenged: the gap between what parents pay and what skilled caregivers need to earn is narrow.
• The business looked different at 10 bookings/week, 100, and 500 — each step changed the value proposition for caregivers.
• At scale, a personal, SMS-based "modern village" became another gig economy job — eroding the supply-side differentiation.
• Caregiver churn was structural: most were in transitional life phases (grad school, early careers) and eventually moved on.
• Parents wanted 6–8 trusted caregivers, but that pool churned faster than it could be replenished.
• Growth ran at 5–10% — solid for some marketplaces, but not enough to attract further investment at the required scale.

How the shutdown decision was made

• The decision was not a single moment — it built over roughly a year as evidence stacked up.
• Three options were evaluated: pivot, acquisition, or full shutdown.
• Acquisition conversations moved quickly: most potential acquirers said no immediately, which usefully closed that path.
• A pivot would have required cutting 75% of the team and effectively restarting — and should have happened six months earlier if at all.
• Runway acted as a forcing function: at six to eight months remaining, Avni stopped hedging and started deciding.
• Fred Wilson's "pivot or die" framing resonated: pivots are painful even when they work, and don't serve the employees or investors who signed up for the original mission.
• The decision crystallised around Thanksgiving; the announcement came December 3rd.

Running the shutdown

• Timing was deliberate: announcing in early December let the holiday period serve as forced recovery time for the team and founder.
• Conserving remaining capital for severance was prioritised over a drawn-out Hail Mary.
• The team was told in structured waves: lead team first, then everyone else — junior employees shouldn't carry ambiguous existential information they can't act on.
• Users were given tools to reconnect with caregivers independently after the service closed.
• Every team member's job landing was treated as a personal responsibility — "excruciating" by Avni's description.
• Avni set a hard personal deadline: December 17th, when she would stop working and start recovering.

The investor relationship

• Monthly investor updates had gone out since day one — every person on the team saw them, including bank balance and runway.
• Because investors were never surprised, the shutdown conversation was not a crisis.
• Transparency removes the sting: if hard news is never a surprise, it lands as context, not a shock.
• Almost universally, investors responded: "I'm sorry, but it's clear you did everything you could — what's next?"
• Choosing investors who believed in the mission paid off most visibly at this moment.
• The discipline of monthly updates was not built for the shutdown — it was built for accountability — and the shutdown benefit was a side effect.

The emotional process

• Identity and company were deeply fused: Poppy started when Avni's second daughter was four months old.
• During the wind-down, Avni didn't talk to friends — she knew she'd break down, and she needed to stay functional.
• Grief is real. Other founders' notes were the most meaningful: they understood the complexity without needing it explained.
• After December 17th, Avni imposed a hard boundary: no "what's next?" coffee until January 4th.
• Structuring recovery like a YC batch — a defined window where outside demands are blocked — made healing possible.
• Grief softened from acute pain to nostalgia over a few weeks; the jabs became less frequent.

Reframing what failure means

• Poppy served thousands of families and hundreds of caregivers and did that job well — the company achieved its local mission.
• Choosing not to pivot to a "high-end tech-enabled agency" was a deliberate values call, not a failure of imagination.
• The three and a half years were an education: hypothesis, test plan, learnings — not a dead end.
• "Shutting down" and "failing" are not the same thing; conflating them keeps founders trapped in bad situations too long.
• The more founders talk openly about shutdown, the less stigmatised the outcome becomes — which makes clean exits more likely.

On what comes next

• The passion for working parents and childcare didn't end with Poppy — the hypothesis ended, not the problem space.
• Avni is deliberately unstructured: coding classes, cooking, art, observational walks through retail spaces.
• The goal is to return to curiosity-led exploration before committing to a new hypothesis.
• Serial entrepreneurship is cumulative learning; the founders who "finally unlock the big thing" built on disparate prior tries.
• Starting another company on a timeline, rather than out of genuine compulsion, would be irresponsible.