Four ways platforms here monetize. One has a $30M gravestone.
Before fixing on numbers, the honest question is: how does this industry already make money — and where is the gap. Here are the four archetypes, the live precedents, and where Frenzee sits.
The four ways. Each one is taken.
Every existing player picks one of these. Each works in some segment — each leaves a real gap.
The numbers, side by side.
What each player charges. Asian incumbents put it on suppliers; Western marketplaces put it on brands; agents put it on buyers. Frenzee splits across all three — for different work.
| Player | Who pays | What they pay | What's covered |
|---|---|---|---|
| Alibaba | Supplier | $3,499/yr Gold | Directory listing · lead-gen |
| IndiaMART | Supplier | ~$335-1,500/yr | Directory listing · lead-gen |
| Faire | Brand | 25% first / 15% reorder + $10 | Buyer discovery · net-60 underwriting |
| Pietra | Buyer (creator) | $29-49/mo + 3% | Sourcing AI · brand tools |
| Maker's Row | Both sides | $39-2,999/mo | Directory + tools (no service) |
| Sourcing agent | Buyer | 5-10% / 8% median | Full sourcing service · non-scalable |
| Accio (Alibaba AI) | Buyer | Free + ~$49/mo | AI sourcing assistant only |
| Tundra | Neither | 0% — free for all | Tried to be Faire-but-free · bankrupt 2023 |
| Frenzee | All 3 sides | $49-599/mo + 2-3% take | Project management + AI + supplier network · multi-side |
"Free for everyone" doesn't survive.
"Raised $30M+ to be a zero-commission B2B marketplace. Sold to Faire in 2023 — for what's essentially salvage value. The unit economics never penciled."
The trap: free is great for acquisition, fatal for retention without a paid layer underneath. Frenzee's free tier is the wedge, not the product — paid unlocks the work that actually saves operators time and money. This is the answer to the inevitable advisor question: "why not just be free?"
Where Frenzee sits.
Two axes that actually matter: how much service does the platform deliver, and which side(s) pay. The gap is in the top-right.
The top-right quadrant — high service intensity AND multi-side revenue — is empty. Every player who tried multi-side without service intensity (Tundra) collapsed. Every player with high service intensity is single-side (traditional agents) and can't scale. Frenzee is the first to combine project-management-grade service with multi-side network economics — and that's only possible because AI absorbs the per-engagement cost that historically forced sourcing agents into single-side commission.
Why all three sides pay without triple-dipping.
Each side pays because Frenzee replaces a real, current line item on their P&L. Same platform, three different jobs.
Frenzee is the first AI-augmented project-management platform for cross-border sourcing — combining the service intensity of a sourcing agent, the network economics of a marketplace, and the unit economics of SaaS.
The market has Alibaba-style directories, Faire-style marketplaces, and one-person sourcing agents. None of them deliver the full project end-to-end. None of them earn from all three sides honestly. None of them combine real service with real software leverage.
The reason no one's done this before is that the service intensity has always required human bandwidth one engagement at a time. AI is what makes the unit economics work — Frenzee can run 10 client engagements with one operator, where a traditional sourcing agent runs one.
That's the bet. Thursday, we'd love your read on whether it holds up.