Case Study № 01
How I rebuilt Chairish's auction business after the Chairish acquisition — new inventory pipelines, a homegrown AI cataloger, and the sell-through to prove it.
TL;DR — for the busy recruiter
By the Numbers
5.4% → 43%
Auction sell-through
+$78K / mo
Added GMV in 60 days
$155K → $342K
Quarterly GMV ramp
41 days → <3
Time-to-live for new lots
Chapter 01 — Situation
When Chairish was acquired, I stepped in to run full-service auctions. The business had inventory, a team, and a respected brand — but on the metric that actually matters in auctions, sell-through, it was deeply underwater.
The platform average for timed-plus auctions on Live Auctioneers was around 30%. Chairish auctions were converting at 5.4%. Plenty of activity. Not nearly enough outcomes.
Chapter 02 — Move 01
First move: stop treating Chairish and Live Auctioneers like two separate worlds. I built a method for moving Chairish dealer inventory onto Live Auctioneers through a dedicated membership profile.
That single change did three things at once: it turned a one-time fee into recurring membership revenue from dealers, gave their inventory a much larger audience, and lifted their sell-through. The GMV ramp speaks for itself.
$155,780
Q1 GMV
$282,533
Q2 GMV
$341,672
Q3 GMV
“Recurring revenue from the supply side is the most underrated growth lever in marketplaces.”
Chapter 03 — The Tension
GMV looked great in the deck. But underneath, the 5.4% sell-through hadn't moved much. We were running auctions full of inventory that was priced, positioned, and curated for a marketplace — not for a live auction audience.
If the answer was just "more of the same inventory," we'd already have it. The real problem was supply mix.
Chapter 04 — Move 02
I designed an auction management program to bring in new, true-consignment inventory: estates, downsizes, one-and-done consignors. Two ways in — work with an auction manager, or self-serve through a guided flow for the simpler cases.
It worked almost immediately. Sell-through climbed from 5.4% to 43% on average. In the first two months, the program added $78,000 per month in net-new GMV. Growth kept compounding even after I handed it off.
43%
New sell-through
+$78K
Monthly GMV added
2 months
Time to impact
“You don't fix a sell-through problem with more listings. You fix it with the right listings.”
Chapter 05 — Move 03
Supply was the next bottleneck. Lots were taking an average of 41 days to go from cataloging to live — too slow for the cadence auctions need.
So I built our own lot cataloger. Anyone on the team — or a consignor on self-serve — could use it. Under the hood it leaned on AI to draft titles, descriptions, and categorizations. Time-to-live dropped from 41 days to under 3.
More lots, faster, at a higher sell-through. That's the whole auction flywheel, finally turning the right direction.
“If a tool needs a manual, redesign the tool.”
The Receipts, Visual Edition
Marketplaces change. Screenshots don't. A small archive of the consignor-facing pages, the AI cataloger, and the internal ops console — click any image to enlarge.
Chapter I
A landing page that screamed the benefits of the program — and was SEO-tuned so we could pull new leads in alongside ones from Chairish, ESN, and LiveAuctioneers.
Chapter II
Upload photos, hit Generate, and a draft listing — title, description, measurements, low/high estimates — appears in seconds. Iterated based on real consignor feedback.
Chapter III
Where sales nudges engaged-but-stalled leads, and where production plans the next two months of auctions.
Chapter IV
Traffic on the new public surface — three months in, growing on every metric.
What this used
Epilogue
The program kept growing after launch. Leadership ultimately chose not to continue investing in it — a decision I respect, and one that doesn't change what the numbers proved: there is a real, repeatable auction business here when you align supply, speed, and sell-through.
Live Q&A
AI · trained on my résumé · under 4 sentences