Three kinds of operators · three motions · one engine
Three different operators. Three different ways Frenzee earns its keep.
Frenzee serves three kinds of operators day-one. Each one comes in through a different door, takes a different first action, and pays at a different moment. The work the engine does underneath is the same — the surface adapts to how each operator already works.
The three: migrating-from-Shopify shop owners, professional sourcing operators (you), and people with ideas + capital + no production network. Below, each one mapped end-to-end.
Migrating shop owners
Shopify, Shopee, Stan migrants
Small-shop operators plateaued on existing platforms. They already set everything up once. They won't do it again unless it's painless.
How they describe their own problem
"I already set everything up on Shopify. I'm not going to do that again. If you can't make this painless, I'm not migrating."
"I'm already paying for Shopify every month. Show me the value before you charge me — and make the migration a one-click thing, not a setup project."
Workflow · discovery → first revenue
From a paste-the-URL onboarding to a monthly retainer the migrant doesn't want to leave.
Discovery
Trigger · outreach
Cold outreach to plateau-stuck Shopify operator
Lead with the migration painlessness, not the AI. "You already have Shopify. Here's the one-click way to add a production manager."
First action · turnkey onboard
Step 01 · 30 sec
Paste shop URL
Operator pastes Shopify / Shopee / Stan URL into the Frenzee onboarding flow. No re-setup. No exporting CSVs.
Step 02 · auto-extract
Frenzee imports the catalog
Products, prices, contacts, brand voice — all extracted from the existing site. "You paste your URL. You get a branded shopfront."
Second action · the production manager wakes up
Output · profile
Branded Frenzee profile generated
A full operator profile auto-built from the imported catalog. Foreign buyers can find it. Auto-translate handles language reach.
Step 03 · onboard
AI production manager opens the first chat
Bot reads the catalog, then opens: "Let's figure out your next product order." First-touch is operational, not promotional.
Daily action · the value loop
Channel connect
WhatsApp + email linked
Operator wires their existing ops channels. Frenzee captures vendor chats both sides. Linked-device pattern.
Always-on
Communication audit
Drift detection on vendor commitments — spec, price, timeline. Surfaces inline when a vendor reply doesn't match what was promised. The framing here is "communication audit" — catch the moment something slips.
Output · 5pm
Daily 5pm summary
"Daily action service. 5PM summary. Source against the network. Get paid through Frenzee." Five things still open. Operator just decides.
First revenue
Revenue moment · subscription
Monthly retainer kicks in
Target ~$1K/mo if we prove sales lift past plateau. Multiple entry points eventually: subscription, one-stop project budget, or % cut per garment.
How they pay
Subscription-first. Transaction percentage later, for proven accounts only.
| Stream | Trigger | Day-1 | At scale |
|---|---|---|---|
| Monthly retainer | 30-day trial → subscription | ~$300–500/mo introprice-discover on first 10 paying | ~$1,000/moif proved sales lift past plateau |
| Project budget | Operator earmarks a campaign budget | Frenzee admin fee (flat or 5%) | 15–20% take on managed spend |
| Transaction % | Sale closes through Frenzee-sourced production | Not enabled"this is the riskiest. Only certain people." | 1–3% cut, opt-in onlyvetted accounts who can't walk halfway |
Why the multi-entry-point matters. Emily on the call: "We need to have different entry points for people. Subscription, like a monthly retainer. One-stop — okay, I wanna spend $100,000 on my project. Or once the sale is confirmed, you get a percentage cut of every garment."
Communication audit · how it surfaces here
Visible. Inline flags on vendor chats and a dedicated section in the 5pm digest. The framing is "don't get ripped off" — catch the moment a supplier reply drifts from what was agreed, before the operator misses it.
Professional sourcing operators
Sourcing + procurement operators
Professional sourcing + procurement operators running real client books. The agent replaces the production-manager or merchandiser seat — not augments it.
What the operator brings · what we add
"My brand clients want an AI that can replace the production-manager or merchandiser seat in their team. If I can offer them that, on top of my sourcing relationships, I become much harder to switch off."
"Suppliers in China respond on WeChat or email — mostly email now. The time-saving of someone reading + sorting + drafting replies in that flow is huge. I want my time spent on the calls that need me, not on triaging inbox."
Workflow · discovery → first revenue
Multi-channel chaos becomes a procurement CRM the operator can hand to a brand.
Discovery
Trigger · network intro
Direct intro through James + B-tier supplier network
Not paid acquisition — direct intro through the operator's own network. A handful of B-tier suppliers and one designer onboard alongside one existing client engagement, so the workflow gets pressure-tested against real production work.
First action · set the watch list
Step 01 · connect
Wire WhatsApp + email to bot
Operator links primary ops channels. Linked-device capture on WhatsApp. Email forwarding rule for inbound supplier replies.
Step 02 · mark
Mark 3–5 active vendor relationships
Operator flags which suppliers to "watch." Bot starts tracking commitments only for those — no broad sweep, no surveillance of every contact.
Second action · operator forwards what suppliers actually say
Channel · operator-routed
WeChat supplier reply
Operator screenshots / pastes the Mandarin reply into the bot DM. "Operator pastes/forwards/screenshots into the bot DM as canonical commitment of record."
Channel · operator-routed
Email supplier reply
Operator forwards the English email into the bot. Same canonical record regardless of origin channel.
Channel · direct
WhatsApp direct (today)
Already captured directly via linked-device on both sides. No routing needed. Phase-1 covers WA ↔ WA natively.
The bot does its silent work
Always-on · agent
Extract → normalize → audit
Bot extracts commitments (spec, price, timeline) from Mandarin / Hindi / English source. Normalizes to canonical units — USD, days, GSM. Audits drift in normalized space. Flag returns in the operator's language with link/screenshot to the source. Built on memory + grounding + citation guards tuned to refuse to hallucinate.
Daily surfaces · the two artifacts James actually wants
Surface · CRM
Procurement-sourcing CRM
Supplier list · current status · last touch · drift flags. The operator's working table. Can be screenshotted into a Friday brand update.
Surface · 5pm digest
Supplier follow-up section
"Vendor X promised sample by Mon. No message since Wed." The drift flag in operator language — the supplier-follow-up section that anchors the daily digest.
Physical QC chain · Jason's IP
v1 · Hong Kong intake
Sample arrives in HK → Emily + James inspect
Cheap shipping vs Sydney. Sample QC v1. Phase-2: multi-country checker chain ("the product may reach in another transit") — anti-bribery through distributed inspection. Jason: "this is our IP."
First revenue
Revenue moment · per-client engagement
Per-client engagement OR monthly
James charges his brand client. Frenzee licenses per active client engagement, OR a flat monthly. Pitch language to the brand: "I have an AI that can replace your current production manager."
How they pay
Per-client engagement. The agent earns its keep by replacing a seat in the operator's client team.
| Stream | Trigger | Day-1 | At scale |
|---|---|---|---|
| Per-client license | Operator engages Frenzee on a client | ~$500–1,500/client/moframed as "replacing a production-manager seat" | $2K–5K/client/moscaling with deal value, not seat count |
| Operator subscription | Active operator with 3+ engagements | Bundled into per-client | ~$2,000/mo flatoperator-side dashboard + CRM across client portfolio |
| Physical QC fee | Sample inspection in HK or chain | Pass-through cost + flat feev1 manual inspection | Tiered fee on multi-country chainPhase-2 distributed inspection |
Why the brand pays. Emily on the call: "You have an agent that can fully replace what some of them can do. Which makes sense to them." James gets margin uplift. Brand reduces a seat. Frenzee charges per replaced seat. No three-sided economics gymnastics.
Supplier follow-up · how it surfaces here
Central to the operator's day. The 5pm digest is built around the supplier-follow-up section. When a vendor goes silent past their promised date, when a price drifts, when a spec changes mid-thread — flags surface in the operator's language regardless of which channel (WeChat, email, WhatsApp) the message came through. This is the engine.
Ideas with capital
Capital + ideas · no production network
Capital and ideas, no production know-how. They sit on great product ideas for years because step one is too hard. Frenzee does the heavy lifting.
The customer's own words
"I've had this idea for two years. I just never moved on it because the sourcing piece alone feels like a six-month research project."
"A lot of my friends in finance want a side business — they have capital, they have product ideas, but they hit a wall the moment it's time to find a factory or a designer."
"By the time I've figured out the supplier, the designer, the MOQ math, the trademark — I've spent more on consultants than the first run is worth. So I don't start."
Workflow · discovery → first revenue
An idea sitting in someone's head for two years becomes 100 units, branded and trademarked, in 2–3 months.
Discovery
Trigger · referral
Friend shares a Frenzee referral link
Acquisition is conversational, not paid. "You're like — just go into Frenzee. Here's my referral link. We'll both get the money, whatever." Both referrer and friend share an incentive.
First action · idea in, pricing anchor out
Step 01 · idea
Type the idea into chat
"Oh, this is my idea. I wanna make mahjong-shaped pet toys." Free-text first message. No form, no field list, no setup.
Step 02 · anchor
Pricing anchor calculated backward
Bot asks: "How much do you want to sell at?" User: "$20." Frenzee calculates backward from retail price to MOQ cost. Margin shape locks early — James's framing.
Second action · Frenzee proposes the production stack
Output · factory
Vetted factory + sample images
"We're gonna help you find a vetted factory that produces quality stuff. They've produced for people before. Here, look at some sample images."
Output · designer
Indonesian designer rolodex + style picker
"We have designers in Indonesia. They're really good and cheap. Whole portfolio list. Do you wanna look through it or tell me what you like?" User picks a reference style (e.g. Spider-Man comic). Rolodex narrows.
Output · IP + trademark
IP transfer + trademark filing
"We'll handle the legal. That designer in Indonesia can't screw you over. He's gonna hand over all the IP. We can even help you with trademark filing in your desired market."
The number lands · decision point
Step 03 · cost
Cost calculation
"$6 per product at MOQ 100. So $600 plus Frenzee admin fee." Frenzee shows the math openly — no surprise fees. "$700 USD."
Decision · pay or walk
Pay deposit → produce. Or walk.
"Are you good? Done. Pay deposit." Single decision moment. ~$2,000 deposit, refundable based on production volume. The friction is here, on purpose — everything before is free.
Production · 2–3 months
Step 04 · sample
First sample arrives
Through Hong Kong intake (shared with the professional-operator QC chain). Buyer approves before bulk production. "Estimated 2-3 months depending on how fast you reply."
Step 05 · 100 units
First 100 produced + branded
"Your Mahjong pet toy can get produced. 100% with unique branding visuals that you own. With trademark in place if you want to."
Distribution · we know people who will carry it
Distribution · real shops
Distributor network test
"Can I sell 20 in Canada? 20 in US? 20 in Sydney? 20 in Hong Kong? Test out which is the hottest market." Real-life shops on a no-commitment trial. Jason's HK + Emily's network seed it.
Distribution · creators
Influencer affiliate sales
"We don't do AI-automated social media — that doesn't work very well. What we have are networks of distribution and creators." Real influencers, affiliate cut on confirmed sales only.
First revenue · and the ongoing relationship
Revenue · transaction-time
Deposit + admin fee captured
~$2K deposit (refundable on production volume) + flat admin fee OR 2% cut. Plus affiliate cut on distributor + influencer sales.
Revenue · ongoing
"From here on out, we're your production manager."
"We'll help you manage everything in terms of operations and production. You only have to make and figure out how to sell — but we'll help you." Monthly subscription OR transaction percentage after first run.
How they pay
Friction is at the deposit. Everything before is free, by design.
| Stream | Trigger | Day-1 | At scale |
|---|---|---|---|
| Deposit | User confirms quote | ~$2,000"refunded based on how much you produce" | Tiered by project size |
| Admin fee | At quote / sample stage | Flat $200"or 2%. We have to figure that out." | 2–5% transaction cut |
| Trademark + IP services | User opts in | Pass-through + flat fee | Tiered legal package |
| Distribution affiliate | Sale closes via Frenzee referral | Not enabled | 10–20% on each confirmed sale"affiliate cut on creators + distributors" |
| Ongoing production manager | After first run delivers | Optional add-on | Monthly OR transaction %user chooses — "what makes most sense" |
Why the friction sits at the deposit, not before. Emily: "Before, I would sit here forever. I'll never take a step forward because the heavy lifting is I need to do all that calculation. No one's talking me through it. But now you're telling me today — if I wanna make that mahjong pet toy, this is the quotation, and we've done this as cheap as possible." The product's promise is that the first decision is "pay or walk," not "fill in 47 fields."
Behind the scenes
Silent. The customer doesn't see the audit machinery. It runs underneath against the factory's and designer's commitments. The only time anything surfaces to the customer is if a deadline drift threatens the deposit-to-delivery promise they paid for — at which point they get a clear notification with what's slipping and what we're doing about it.