Pitch deck
Pre-seed · $2.5M · San Jose, California
Own less.
Access more.
Turn the things you own into the things you need — instantly, with AI.
Sony A7 III body + 28–70mm
1.1 mi away · Like new
S$660or S$46/week
ShareHouseSan Jose
Herman Miller Aeron, size B
0.3 mi · Excellent
S$345S$24/week
iPad Pro 11" with Pencil
0.3 mi · Excellent
S$405S$28/week
Haworthia in ceramic planter
0.3 mi · Like new
S$15S$3/week
KitchenAid Artisan stand mixer
0.4 mi · Excellent
S$170S$12/week
Pre-seed · $2.5M · San Jose, California
Own less.
Access more.
Turn the things you own into the things you need — instantly, with AI.
Sony A7 III body + 28–70mm
1.1 mi away · Like new
S$660or S$46/week
ShareHouseSan JoseHerman Miller Aeron, size B
0.3 mi · Excellent
S$345S$24/week
iPad Pro 11" with Pencil
0.3 mi · Excellent
S$405S$28/week
Haworthia in ceramic planter
0.3 mi · Like new
S$15S$3/week
KitchenAid Artisan stand mixer
0.4 mi · Excellent
S$170S$12/week
{02}The problem
Every home is sitting on thousands of dollars it cannot spend.
The average US household
$4,267
of unused things — 161 items it no longer touches.
$559.8B
sitting idle across American homes
9.7M tons
of furniture landfilled in a year, 80% of all that is thrown out
{03}The insight
Barter fails because Alice needs Bob to want her chair.
Every marketplace
Days to never
The wait is where people give up.
ShareHouse
Two minutes
The credits arrive before the buyer does.
An AI appraisal priced in a common unit removes the coincidence of wants. Alice’s chair becomes buying power.
{04}The product
Snap it. Get credits. Claim what you need nearby.
- 01
Snap it
Four photos of anything you no longer use. No description, no price to guess.
- 02
AI appraises
Exact model, recent used-market sales, condition from the photos, local liquidity.
- 03
Credits land
Immediately. The item stays in your home, committed, until a neighbor claims it.
- 04
Spend nearby
Buy it outright or rent by the week. Local pickup, QR handoff, both confirm.
Credits are not money: not cashable, not transferable, not for sale. A credit only claims goods.
Herman Miller Aeron · size BShare Value
0
Good evening
Your Share Wallet
Available to spend
448
S$213 more unlocks when your items are claimed
In the ShareHouse
Fujifilm X-T3 body
Reserved
Trek commuter bike, 56cm
On the map
Activity
Sony A7 III body + 28–70mm
1.1 mi away · Like new
S$660or S$46/week
{05}Share Value
We price what it will clear for here — not what it cost.
We start from real used-market sales of the exact model, then multiply by the condition we can see, how sure we are, and how quickly things like it get claimed near you.
The appraisal is the ceiling, not the advance. A new member spends 25% of it at once and the rest as the item clears — which is the credit line, and the first thing that keeps the currency honest.
- 25%
- spendable at once · New
- 40%
- spendable at once · Established
- 50%
- spendable at once · Trusted
Sony A7 III body + 28–70mm
Used-market comp $860
Share ValueS$660
Velvet three-seat sofa, forest
Used-market comp $780
Share ValueS$300
Same sticker, different Share Value. The liquidity term is what every previous credit economy was missing.
{06}Live today
Built, deployed, and appraising real photographs.
AI appraisal in production — vision plus live market search
Map marketplace, buy or rent, credits + USD checkout
Postgres ledger: one open claim per item, by construction
An assistant that can drive the whole app
11s
Measured on a real listingDeWalt drill — correct model, wear identified, four market comps.
Good evening
Your Share Wallet
Available to spend
448
S$213 more unlocks when your items are claimed
In the ShareHouse
Fujifilm X-T3 body
Reserved
Trek commuter bike, 56cm
On the map
Activity
Herman Miller Aeron · size BShare Value
0
{07}Market
A trillion-dollar stock of dead capital, entered one neighborhood at a time.
Why it is a stock, not a flow
$400
is the emergency expense a large minority of US adults cannot cover in cash.
Many of them own a garage worth several thousand. Liquidity, not wealth, is the binding constraint — and household goods are the one asset almost everybody has. The wedge is the 39.6M Americans who move house each year and need both sides of this network in the same month.
{08}Why now
Four things became true at once.
~$0.05
AI can run the pricing
Pricing is the monetary policy of a credit economy. A vision model now identifies an exact item and grades condition for about five cents — the job that used to need an expert, per item, per neighborhood, re-estimated nightly.
93%
Secondhand went mainstream
93% of Americans bought used last year and the stigma is gone. We are not asking anyone to change their behaviour, only to be paid sooner.
1 phone
Everyone carries the evidence
A modern camera resolves a model label, a serial plate, the wear on a chuck. A decade ago the same photograph could not have supported a binding valuation.
Later
The credit layer can go further
Postgres now, because fraud and inflation are easier to fix when an entry can be reversed. Then verifiable public issuance, then independent settlement, then transferability — last, never first.
Every failed precedent made its currency liquid before it had an economy. Transferability is the last feature, not the first.
{09}Competition
Plenty of competitors for the item. None for the moment.
Shipped, then inspected
Handed over on the spot
Still at home
Never paid
Paid at the sale
Paid before a buyer
Nobody still operating pays at listing while the item stays at home.
Pay-before-sale is big where goods are standard — ATRenew does US$3.0B on phones. It dies wherever it has to take bulky, idiosyncratic things into a warehouse.
AI pricing is a commodity.
Meta’s Seller app prices 430M items a month for free. The money attached to the appraisal is the company.
{10}Precedents
Tried twice, in 2017. What broke them is designed out.
CASH · Tokyo
¥360M in 16 hours
advanced against photos, in cash, before a single item arrived. Sold for ¥7B ($62.2M) as a six-person company. Under 1 in 10 failed to deliver.
Mercari NOW · the incumbent
~40 minutes
for its ¥10M daily cap to run out. Priced from its own listings, paid on the spot — and closed after 9 months.
What broke — and where it lives in ShareHouse
- Paid full value, in cash
- 25–50%, in credit only spendable here
- ¥360M in sixteen hours
- Daily network cap and coverage breakers, enforced at issuance
- Photos of ramen
- Appraisal refuses food, people, screenshots, stock images
- Trucks and warehouses (Kaiyo)
- Never takes possession
- Credits sold for dollars (Listia)
- Credits cannot be bought
- Wallets frozen (Bunz)
- Backed by members' goods, never by our cash
The one we have not solved: when Listia let credits become cash, listings rose 50% and values doubled. Closed-loop credit attracts worse goods — so we measure it by category from the first listing.
{11}Defensibility
What compounds, when software itself is free.
The clearing data
A model knows what an Aeron sells for. It cannot know what one clears for in Willow Glen, in eleven days. That is not on the internet — it is a byproduct of operating.
Reputation
Handoffs completed, as-described rate, reviews from neighbours who met. Not a badge — a credit limit. It does not travel.
Density
A rival cannot enter “the US”. They must beat us in San Jose, then Willow Glen, paying full supply-seeding cost each time.
The balance
Spendable in one place. Poshmark sellers reinvested 48% of earnings they were free to take home in cash.
The appraisal is not the moat. The correction to it is.
Comp data can be bought. Clearing data cannot — no other platform issues value before a sale, so none of them needs to know how fast things move.
Economies of scale, literally
- One team, amortised across twelve metros rather than one
- Pricing off observed clearance rather than a prior
- Disputes patterned rather than novel
- Supply from referral rather than paid
At a 5% take, one metro cannot carry a team and twelve comfortably can — which is why the projections behave as they do.
{12}The credit economy
One identity decides how many credits may exist.
In steady state
B/A = 1 / (a/p + 1−a)
Credits granted per dollar appraised, given how much we pay up front (a) and how often listings actually clear (p). Everything else follows from it.
What it told us
Paying 100% up front is solvent only if everything clears. The closest comparable operator clears 43% / 69% at thirty and ninety days — having physically inspected every item first.
So the front-load is capped at 50% for every tier, and tier buys a smaller haircut instead. The reward lever moves; the solvency lever does not.
Coverage target, published live
125%
Inventory over credits outstanding. Tighten at 115%; mint on claim only below 105%.
And a sink
4% of every claim is charged in credits and destroyed. Ithaca HOURS died with 33% of its money supply held by one shop that then refused it. Circulation is not a sink.
What will be attacked
Wash trading reached 94.5% of volume where per-trade rewards were paid. We mint on listing, so the mint is the reward. The counterparty graph is the only defence with real ground truth — and it must exist before the credits do.
{13}Business model
Goods trade in credits. We are paid in dollars.
One claim, year three
| Item value, in credits | S$130 |
| Our fee (5%, in USD) | $6.50 |
| Sales tax, collected and remitted | $12 |
| Card processing | ($0.83) |
| AI appraisal | ($0.20) |
| Trust and disputes | ($0.78) |
Contribution per claim$4.69
72%
contribution margin on the fee
3.3×
LTV to CAC
Dollars never touch the goods — only the fee and the tax. That keeps the credit loop closed, and it is why the AI cost (about five cents) is immaterial while card fixed fees are the real margin question.
{14}Plan
Prove one metro clears. Then open five.
Now
Product live
Appraisal, marketplace, credit ledger, assistant — all deployed.
Month 6
San Jose at 2,000 listings
First measured sell-through by category, replacing our estimated liquidity discount with real data.
Month 12
$2.6M GMV in year one
7K active members, coverage held above 110% for a full year, 1099-B infrastructure live.
Month 18
Series A
110K active members on the year-two ramp, repeat rate above 40%, a credit economy that has not inflated.
One metro has to clear before a second is worth opening — the liquidity discount is an estimate until a neighborhood has sold through a season of it.
{15}Forecast
$101M of revenue in year five.
Revenue, with goods exchanged
$142K
$2.6M GMV
$2.2M
$41M GMV
$11M
$190M GMV
$39M
$579M GMV
$101M
$1.3B GMV
Y1 · +1 metro
7K members
Y2 · +5 metro
110K members
Y3 · +14 metro
514K members
Y4 · +27 metro
1.6M members
Y5 · +38 metro
3.6M members
$101M
revenue in year five, on $1.3B of goods
78%
gross margin
On $213M raised across 5 rounds, each gated on the last. We are asking for the first. Most sensitive to sell-through, then to CAC outside California — both measured before the round that depends on them.
{16}Scale
What $500M would take, from where the plan lands.
The forecast
$101M
year five, on 3.6M active members
5×
to $500M. Capital buys members; it does not buy activity per member or take rate. Those have to be earned.
Revenue is members × what each transacts × what we keep. A target fixes a surface, not a plan — so here are three points on it, none a variation of the others.
Reach
$498M
- Members
- 14.0M
- Each spends
- $400
- We keep
- 9%
Faster than Airbnb, Poshmark or Vinted ever grew.
Depth
$500M
- Members
- 6.0M
- Each spends
- $925
- We keep
- 9%
Four times Poshmark per member. Tests the thesis directly.
Rate
$500M
- Members
- 9.0M
- Each spends
- $500
- We keep
- 11%
Half of revenue from products not yet built. Inspection is closed to us.
What we are claiming is not that we reach $500M. It is that the business has a shape at that scale needing nothing absurd of any single number — and that every route runs through the same two constraints as the plan itself: sell-through and density. A $500M company is a metro playbook that works, run a hundred times. We are asking to be funded to run it once.
{17}Risks
What would break this, and what we do about it.
Collusive minting
Wash trading hit 94.5% of volume where per-trade rewards were paid. Counterparty-graph instrumentation before credits ship; diversity required for the top tiers.
Accumulation, not inflation
Ithaca died with a third of its money held by one shop. A credit-denominated fee that destroys, plus published concentration limits.
Adverse selection
When Listia let credits become cash, listings rose 50% and values doubled: a closed loop attracts worse goods. It cannot be designed out, so it is measured by category from the first listing.
The appraiser is wrong at scale
Zillow wrote down $407.9M. Calibration watched per decile, and a model-drift breaker that suspends automated minting.
Running out of dollars
Bunz froze wallets over a fiat cashflow problem while its credit system was solvent. USD reserve against the float, disclosed.
Meta
Free, enormous, now AI-priced. They will not issue value against uninspected goods — that is the liability we designed around.
{18} The ask
$2.5M to make
a neighborhood clear.
45%
Engineering & product
25%
Supply seeding & growth
15%
Trust, safety & support
15%
Legal, compliance & reserve
Everything you own has Share Value.