The Case for a Curated Hi-Fi Marketplace Built for the Die-Hards


This thread got me thinking, and I'll throw something out there that I suspect a few others have quietly considered.

What if we actually built something better?

Not a knock on Audiogon or USAM — both serve a purpose — but neither was really designed for the serious end of this hobby. The audience here includes people running $50K, $100K, $200K+ systems. We're talking CH Precision, VAC, Basis Audio, Wilson, dCS, Nagra. Gear that deserves more than a blurry iPhone photo, a two-line description, and a listing that auto-expires in 30 days next to an ad for a $79 Bluetooth speaker.

Here's what I'd actually want to see:

**Flat $15 listing fee. No tiers, no commissions, no time limits.** Pay once, it stays up until it sells or you pull it. The current commission model on high-dollar gear is indefensible — a percentage of a $15,000 DAC sale is not a service, it's a toll.

**High-resolution photography as a platform standard.** Not a feature — a requirement. Multiple angles, macro shots of any wear, photos of original packaging if included. If you can't show it properly, you can't list it. Serious buyers make decisions on photography. This single change would separate the platform from everything else out there.

**Brick and mortar dealers list free.** Always. A curated dealer presence — not banner ads, actual inventory — gives the platform legitimacy and gives small independent shops a fighting chance against the grey market. Their trade-ins and demo pieces are exactly what serious buyers are looking for: known provenance, usually impeccably maintained, often with remaining warranty.

**A dealer clearinghouse for demo and trade-in gear.** Separate, searchable, clearly flagged. This is untapped inventory that currently trickles out through inconsistent channels. Centralize it.

**A community board that actually requires skin in the game.** Read-only access for guests. Participation requires a verified listing history or a one-time membership fee. Keeps it from becoming a free audio consulting service for people who have no intention of buying anything.

The model only works if the community that populates it is self-selecting toward serious participants. The $15 listing fee isn't really about revenue — it's a filter. People who won't spend $15 to list a $5,000 phono stage aren't your target audience anyway. Neither are the folks who show up to ask how a cartridge sounds and disappear when it's time to actually pull the trigger.

I'm genuinely curious if there's appetite for this here. The talent, the taste, and frankly the gear to make something like this worth building clearly exists in this community. The question is whether anyone wants to be part of something more intentional than what we currently have.

Would love to hear what others think the non-negotiables would be.

73cuttysupreme

For the .01% of the .01%. Where's the profit? In that realm, slow to low sales (little churn) and hard to keep going for very long once the newness wears off.

@73cuttysupreme 

I agree 98%.  The 2% is your fee structure.  $15 is crazy high for my $50 Bluetooth speaker and is to low for me to maintain the servers for all the forums as well as the actual for sale listings. Maybe a flat 3%-5% up to 5K and 2.5%-3% from 5K and up?

My 2 cents.

I have the tech stack economics being worked out right now.  that 3% isn't much cheaper than agon today.  add up that 3% on 5k plus the haircut I may or may not take on accepting PayPal or CC if I am a dealer, plus tax ramifications (if it's a thing) and all of a sudden 12% of your listing just went poof....  maybe it's a flat fee, maybe it's a percentage, like maybe 1% regardless of price but then it's charged at end of transaction and the whole up front gatekeeping BS listers goes away. or maybe we just go to 20 or 25 per listing, no booster crap.  the per unit economics are lower than you think for running the infra.  I already have the base infra spun out. there will be some 'always-on' infra, but it's demand driven.  tech stack is as follows:

Compute & API

  • ECS Fargate — containerized Node.js API, auto-scaling, no EC2 to manage

Frontend & CDN

  • S3 — static hosting for the Next.js (React) frontend
  • CloudFront — global CDN, absorbs ~95% of read traffic before hitting origin, DDoS protection layer

Database

  • Aurora PostgreSQL Serverless v2 — primary relational store for listings, users, transactions, forum; scales to near-zero in off-peak

Search

  • OpenSearch (t3.small at launch) — full-text listing search with faceted filtering (brand, category, price, condition, dealer vs. private)

Caching & Sessions

  • ElastiCache Redis — session state, rate limiting, search result caching

Media

  • S3 — high-resolution photo storage with signed URL delivery
  • Lambda — thumbnail generation and EXIF stripping on photo upload

Auth

  • Cognito — user registration, JWT tokens, MFA, optional Google SSO

Email

  • SES — transactional email, gear alerts, dispute notifications

Security

  • WAF + Shield — edge protection
  • GuardDuty — threat detection
  • Secrets Manager — all credentials stored here, never in environment variables

CI/CD & Monitoring

  • GitHub Actions → ECR → CodeDeploy (blue/green deployments)
  • CloudWatch — logs, alarms, dashboards
  • Route 53 — DNS with health-check failover (Multi-AZ)

Payments (third-party, not AWS)

  • Stripe — listing fee collection
  • Stripe Identity — ID verification for $5K+ listings

 

 

The two biggest cost levers are Aurora Serverless v2 and ECS Fargate — both scale to near-zero during low traffic, which keeps burn rate contained during the growth phase. The architecture is designed to scale horizontally without structural changes as volume grows.


 

I’ll be closely following this as what you propose would be highly interesting to me for the reasons you state. 

Post removed