Etsy CEO Evaluation — v4
OnethingNeededShop (shop 50156638) · a fresh, live-API read of where the shop actually stands, superseding etsy-ceo-eval and -v3. Every number below is pulled from the Etsy Open API v3, not remembered.
SOURCE: Etsy Open API v3 · https://api.etsy.com/v3/application/shops/50156638 (getShop, /listings, /receipts, /sections; OAuth) · pulled 2026-07-17 · local snapshot scratchpad/etsy_live_full.json · cross-read against 14 prior Etsy chat sessions
The one-line story
The shop has a real, accelerating funnel — 63 orders in ~9 weeks, growing 7 → 25 → 31 a month on 1,123 views — but it has been running as a $1 free-sample machine: 69 of 71 items ever sold went for $1, so lifetime gross is $95. Prices were just raised to $13–$38, so the shop is at its pivot moment — with zero full-price validation, a fulfillment backlog that threatens account health, and 121 dark listings behind a storefront that advertises products it no longer sells. The next 30 days decide whether the demand survives real prices.
71
Lifetime sales
volume 7→25→31/mo, rising
$95.02
Gross, ~9-wk window
AOV $1.51 · 97% sold at $1
7
Orders not fulfilled
oldest ~28 days · local report says 13
3
Reviews (avg 5.0)
4% of buyers left one
23
Active listings
all tarot/voice, one section
121
Dark listings
85 inactive · 34 draft · 2 other
1,123
Views on active
~6.3% crude view→sale
$13–38
Current price band
raised from $1 — untested
1 · The revenue reality
Orders are climbing — dollars are not
May '267 orders · $9.93
Jun '2625 orders · $27.95
Jul '2631 orders · $57.14
Read it carefully: the bars are order counts, and they are healthy and rising. The dollars beside them are near-zero because the price was $1. This is a demand-proven shop that has never actually been monetized.
Where the money came from
| Items sold at $1 | 69 / 71 |
| Items sold at $10 / $13 | 2 |
| Total discounts given | $1.50 |
| Gross from the two full-price orders | $23.00 |
| Gross from the 59 promo orders | $72.02 |
| Refunds issued | 1 |
It was not a coupon — the list price itself was $1. After Etsy's $0.20 listing fee + 6.5% + payment processing, a $1 order nets roughly break-even at best. The shop essentially gave the catalog away to buy velocity.
2 · The fulfillment liability — fix this before anything else
7 paid orders are not marked fulfilled in the API
| Order date | Age | Buyer | Item | Paid |
|---|
| 2026-06-17 | ~28d | Karen M. | Voice Tarot 12-Min Monthly Check-In | $1.15 |
| 2026-06-26 | ~19d | Lamyaa B. | 15-Minute Voice Tarot Reading | $1.00 |
| 2026-07-01 | ~15d | TT Chan | Voice Tarot 12-Min Monthly Check-In | $1.00 |
| 2026-07-02 | ~13d | JoAnne N. | Voice Tarot 45-Min Multi-Spread | $10.00 |
| 2026-07-07 | ~8d | Leah O. | Voice Tarot 12-Min Monthly Check-In | $0.75 |
| 2026-07-13 | ~3d | Gabbryanna S. | 15-Minute Voice Tarot Reading | $1.13 |
| 2026-07-13 | ~2d | PM | Voice Tarot 12-Min Monthly Check-In | $1.18 |
Why a CEO cares even at $1: these are made-to-order readings a real person paid for and has not received. On Etsy that path runs order → buyer message → case → account-health hit → search suppression. You cannot run a price-raise experiment on a store whose search ranking is being dragged down by open cases. Deliver all seven this week; cancel only any you genuinely cannot fulfill.
Provenance note: the API flags these as is_paid=true, is_shipped=false. Today's local fulfillment report flagged 13. For message-delivered readings the "shipped" flag can lag actual delivery, so treat 7–13 as the range and reconcile each against your sent-files log before cancelling.
3 · Demand signal — what actually sells
The relationship / psychic-hook listings outsell generic "tarot reading" by a wide margin. Click a column to sort. Sales = units seen in the receipt pull.
| Listing |
List $ |
Sales |
Views |
Favs |
Signal |
| Are They Watching You · Psychic Connection | 13 | 11 | 154 | 5 | top seller |
| Psychic Reading No Cards · Real Intuition | 13 | 9 | 96 | 2 | top seller |
| Quick Tarot · One Question, Voice Reply | 20 | 7 | 203 | 9 | most viewed |
| Your Timeline Reading · Extremely Accurate | 13 | 7 | 91 | 2 | strong |
| What Happens Next · Future Insight | 13 | 5 | 52 | 4 | strong |
| Their Hidden Thoughts · Love Telepathy | 13 | 4 | 46 | 4 | solid |
| The Message They Cannot Say | 13 | 4 | 67 | 2 | solid |
| Soulmate Energy Check · Love Reading | 13 | 4 | 38 | 2 | solid |
| Voice Tarot 12-Min · Monthly Check-In | 16 | 4 | 68 | 3 | solid |
| 15-Minute Voice Tarot Reading | 20 | 2 | 108 | 1 | viewed, low conv. |
| Why They Went Quiet · Psychic Insight | 13 | 2 | 27 | 0 | emerging |
| One Question Full Answer | 13 | 2 | 35 | 1 | emerging |
| In-Depth 30-Minute Reading | 38 | 0 | 25 | 2 | no sales yet |
| No Contact · Read The Silence | 20 | 0 | 6 | 2 | no sales yet |
| Apology You Deserved · Healing | 28 | 0 | 0 | 0 | no traffic |
| Income Leak Named · Career/Money | 20 | 0 | 1 | 0 | no traffic |
Product-market signal: "Are They Watching You," "Psychic Reading No Cards," "Their Hidden Thoughts" — the love / telepathy / no-contact / psychic angle wins. The generic "tarot reading" framing and the high-ticket $28–38 items have little or no traction. Build the catalog and any ad spend around the psychic-relationship hook, not around the word "tarot."
4 · The storefront is telling two stories
The announcement vs the catalog
Live shop announcement still advertises:
- "multitrack MIDI song packs for producers" — 0 active
- "printable music theory guides" — 0 active
- "faith devotionals" — 0 active
- "Notion productivity templates" — 0 active
Four of the four non-tarot lines named in your own storefront copy are dark. A shopper who lands on the announcement and searches the shop finds only tarot. Either relight those lines or rewrite the copy — a storefront that promises what it can't show erodes trust.
121 built listings sitting idle
| Active | 23 |
| Inactive | 85 |
| Draft | 34 |
| Expired / sold-out | 2 |
| Sections with 0 active | 3 of 4 |
Faith & Devotionals, Music/MIDI/Books, and Templates & Misc are all empty of active listings. This is either a second revenue leg already built and switched off (a decision to make deliberately) or dead inventory to purge. Right now it is neither — it is drift.
5 · The five CEO moves
1
Clear the fulfillment debt — this week, before anything else
Deliver the 7 (up to 13) open orders, cancelling only any you truly can't fulfill, and reconcile each against your sent-files log. Open cases suppress search; you cannot test higher prices on a suppressed shop.
Do firstEffort: hoursProtects: account health
2
Validate the $13–$38 prices — you have zero evidence they convert
Watch the next ~20 orders at real prices. If conversion holds, you just 13×'d revenue per order. If velocity craters, introduce a $3–$5 entry tier (a single-question reading) as the new front door instead of jumping buyers straight from $1 to $13.
Highest upsideEffort: monitor + 1 listing
3
Farm reviews — 3 reviews from 71 sales is your #1 conversion leak
For a "real human, no AI" reader, reviews are the funnel. Add an automated post-delivery review ask to every order. You wasted 68 chances on the $1 buyers; don't waste the full-price ones. Target: 20+ reviews in 60 days.
CompoundsEffort: template + habit
4
Resolve the storefront contradiction — decide, don't drift
Either relight the MIDI / music-theory / faith / template lines (121 listings already built) as a real second leg, or rewrite the announcement to the tarot-only truth. Pick one this month.
Trust + focusEffort: 1 decision
5
Concentrate on the psychic-relationship hook that already wins
"Are They Watching You," "Psychic Reading No Cards," "Their Hidden Thoughts," "Your Timeline." Make these the hero listings, mirror their SEO across new listings, and retire or reprice the $28–38 items with no traction.
Data-backedEffort: SEO + catalog trim
6 · Risk register
What can hurt you
- Account health: open/undelivered orders → Etsy cases → search suppression.
- Price cliff: the $1→$13 jump is untested; velocity could collapse.
- Single point of failure: one human records every reading. July's 31 orders + manual fulfillment is exactly how the backlog was born — and volume is still climbing.
- Trust erosion: storefront advertising 4 dead product lines.
What's working (protect it)
- Proven demand: 63 orders in 9 weeks, accelerating, no ad spend cited.
- A clear winning niche: psychic / love / no-contact readings.
- Perfect rating: 5.0 across the 3 reviews you do have.
- Built inventory: 121 listings ready to deploy the moment you choose a second leg.
- Clean API pipeline: full read/write toolchain already in place for pricing, SEO, and fulfillment automation.
7 · The agent stack — turn the moves into running help
Six operational agents that run the shop off the API pipeline already in place. Each maps to a finding above. These are blueprints, not running. Per standing policy nothing here runs autonomously until you authorize a specific agent. "Buildable now" means the read/write endpoints it needs are already working; "spec" means it needs a human-approved send.
Buildable now
Fulfillment Watchdog
Guards the "same day" promise you just published, before it turns into cases.
- Cadence
- every 30 min
- Reads
GET /shops/{id}/receipts — is_paid, is_shipped, create_timestamp- Does
- flags any paid order past ~18h unfulfilled, with buyer, item, and age
Addresses Fire #1 (7 open orders) and keeps the new same-day copy honest.
Buildable now
Price Sentinel
Watches whether $13 to $38 actually converts now that the $1 era is over.
- Cadence
- daily
- Reads
- receipts (grandtotal) + listing views
- Does
- tracks AOV and view-to-sale on the next 20 full-price orders; alerts if velocity drops; proposes a $3 to $5 entry tier
Addresses Move #2 (the untested price jump).
Spec — needs approval to send
Review Harvester
Turns delivered orders into the reviews the storefront is starving for.
- Cadence
- on delivery, then day 3
- Reads
- receipts marked delivered
- Does
- drafts the post-delivery review ask; you approve each send
Addresses Move #3 (3 reviews from 71 sales).
Buildable now
SEO Mirror
Spreads what already sells to the listings that do not.
- Cadence
- on demand
- Reads
- top-seller tags and titles vs weak listings
- Does
- proposes tag and title edits (≤13 tags, ≤20 chars) mirroring the psychic and love winners; PATCH on approval
Addresses Move #5 (concentrate on the winning hook).
Buildable now
Catalog Reconciler
Keeps the storefront from promising what it cannot show.
- Cadence
- weekly
- Reads
- shop announcement vs active listings and sections
- Does
- flags copy advertising dead lines; lists the 121 dark listings by state for relight or purge
Addresses Move #4 (the storefront contradiction).
Buildable now
Token & API Sentinel
Keeps the pipeline alive so every agent above can run.
- Cadence
- hourly
- Reads
- OAuth expiry in
.env
- Does
- refreshes the access token before it expires (
refresh_token.py)
Infrastructure for the whole stack; the token expired 18h before this pull and had to be refreshed by hand.