Vanity Growth or Real PMF? Tearing Down a Pre-Series-B Growth Memo
Objective: Read a subscription startup's investor growth memo (styled after the pattern that preceded Winc's 2022 bankruptcy, a company that scaled hard before proving retention) and separate the genuine PMF defects from the numbers that merely look alarming but aren't the actual problem.
A subscription wine-discovery startup is raising a follow-on round. The growth memo below is the kind of document that got startups like this funded in 2019-2021, styled after the pattern Winc's own trajectory followed (IPO'd November 2021, filed for Chapter 11 just over a year later, per its confirmed record). Your job: find what the memo is hiding, and don't get distracted by what merely looks bad. (This memo is a synthetic specimen written for this exercise, not a real internal Winc document.)
One specimen memo, several plausible-looking flags. Some of them are real PMF defects. Some are normal, healthy-looking noise a nervous reader might flag anyway. Sort them correctly and you'll see the pattern that preceded the real company's collapse.
Before you start
What you'll need
Free path (everything below is enough to finish)
Not in the tools directory, genuinely free, this project needs a document, not a data tool.
The process
Specimens to review
List every genuine PMF defect this memo is hiding, and explicitly name which of the listed numbers are NOT defects even though they might look concerning at first glance.
GROWTH MEMO — Series B Update (synthetic, modeled on the pattern of a real 2020-era subscription-commerce raise) Monthly signups: 8,200 -> 10,000 (+22% MoM) Sean Ellis score (last surveyed 14 months ago): "38%, in line with industry" Day-30 retention: 9% (down from 34% two cohorts ago) CAC: $61 -> $70 (+15% QoQ, blamed on "seasonal ad costs") App store rating: 4.2 stars, 1,900 reviews "Referral" signups: 1,400/month (all via a $15-credit referral bonus program) Monthly burn: $1.1M, runway 11 months Quote from the deck: "Our engaged super-users personalize their box every month and rave about us."
Specimen: synthetic, realistic
Final deliverable
A short memo-review write-up: the 4 real defects (with severity and lesson tie-back) and an explicit note on which flagged-looking numbers are actually distractors, not defects.
See a reference example
Real defects found: (1) Sean Ellis score is 14 months stale, (2) day-30 retention collapsed from 34% to 9%, (3) "referral" signups are entirely paid-incentive, not organic, (4) the deck quote cherry-picks super-users while median retention craters. Correctly NOT flagged: the +22% signup growth, the CAC rise, the app rating, and the runway figure, all real numbers, none of them a PMF defect on their own. The pattern here is the opposite of what builds durable subscription retention at a company like Chewy: real loyalty shows up in what customers do release after release, not in a stale survey score or an incentivized referral count standing in for it.
Success criteria
You're done when you can:
- Names all 4 defects with correct severity
- Does not flag any of the 4 distractors as defects
- Ties at least 3 of the 4 defects to a specific lesson section, not just a general "this looks bad"
- Final write-up explicitly separates "defect" from "concerning-looking but not a defect"