# Kitse Seller Activation — Plan to Improve Chef-Lead Conversion

> **In one minute:** Chefs apply excited, then quit during the slow permit/paperwork middle. The fix is
> a friendly onboarding page that splits steps into **“required to sell” (legal — must do)** and
> **“gets more orders” (optional)**, plus reminder emails that only say *“you can’t sell yet”* when it’s
> actually true. We never fake-block someone over a profile photo — we use an honest tip instead. Goal:
> take roughly **15% of applicants reaching their first order up to 30%+**. The working demo is the
> onboarding page in this folder.

---


**Goal:** move more applicants from *“applied”* to *“first order”* without weakening any of the legal
requirements chefs must meet to sell — and lift the activation **rate** measurably.

**The product** is the new onboarding page — open **`onboarding.html`** (it's the folder's landing
page). Complete the steps and preview your storefront. `emails.html` holds the lifecycle email
examples it triggers. *This file is the plan in text; the onboarding page is the deliverable.*

---

## 1. What Kitse actually is (so the plan fits the product)

Kitse is a California marketplace for **permitted home chefs** (Cottage Food / MEHKO). Sellers keep
75% of food sales + 100% of delivery fees, paid weekly. The published seller funnel is:

> **Apply → Onboard (permit + food-handler card) → Add menu → Go live → Register with state**

This matters because Kitse is *compliance-gated by law*. Unlike a normal e-commerce signup, a chef
**cannot legally take an order** until specific real-world steps are done. The plan’s whole strategy is
to make those unavoidable steps fast and visible — and to use them, not fake claims, as our urgency.

---

## 2. Where we’re losing Chef leads (the funnel)

Illustrative funnel (replace with real analytics — see §7):

| Stage | % of starters | Note |
|---|---|---|
| Started application | 100% | |
| Submitted application | 72% | form friction, but not the main problem |
| **Permit + food-handler done** | **34%** | **🔴 biggest single leak (−38 pts)** |
| Profile + menu complete | 23% | 🔴 storefront never finished |
| Payout connected | 19% | 🔴 can’t be paid → can’t sell |
| First order received | 15% | target: **30%+** |

The drop-off is **not** at “Apply.” Chefs are excited to apply. They die in the **compliance-heavy
middle**, during a ~2-week silent gap where nothing visibly moves and no one helps.

---

## 3. The four leaks and their fixes

### Leak 1 — The permit wall & the silent gap (biggest)
Chefs hit “go get a Cottage Food or MEHKO permit” and stall for weeks. No guidance, no visible
progress, no human. They ghost.

**Fix**
- A **live onboarding dashboard** (see `onboarding.html`): progress ring, a clear “what’s blocking you
  *right now*” banner, and a permit status tracker.
- Make *“we’ll help you get a permit”* concrete: decide Cottage Food vs MEHKO for them, pre-fill the
  county application, human-review before submission so it’s approved the first time.
- Kill the silent gap with the lifecycle emails in §5.

### Leak 2 — Profiles that never go live
No photo, no story, no hours → storefront is invisible or can’t take checkout. The chef thinks they’re
“done” and earns nothing.

**Fix**
- Split the checklist into **Required to sell** (legal) vs **Boosts orders** (quality) so a chef always
  knows what actually blocks money vs. what just helps.
- Block “go-live” only on the legal items; never on cosmetic ones.

### Leak 3 — No lifecycle nudges
A stalled chef hears nothing. The original idea — *email them that their profile isn’t done* — is the
right instinct, but only works if the message is **true** (see §4).

**Fix** — a 6-stage lifecycle email system (§5), each tied to real onboarding state.

### Leak 4 — High-value leads treated like everyone (e.g. **Canné LA**)
A marquee lead like **Canné** — the LA cannelé baker — gets the same generic drip as a casual
applicant, and we risk losing a flagship storefront.

**Fix — Concierge activation track** for priority leads:
- A **named onboarding specialist** (Maya in the demo) instead of automation.
- **Expedited permit help** and a **free food-photography shoot**.
- A 1:1 plain-text intro email (see `emails/concierge.html`), not a templated blast.
- Trigger criteria: strong cuisine/brand, high follower count, press, or a category we want in a market.

---

## 4. ⚖️ The honesty guardrail (important)

The proposed nudge — *“your orders are being blocked because you didn’t finish your profile picture”* —
**should not ship as written.** A profile photo does **not** legally block an order, so claiming it does is
false. That’s a deceptive-urgency / dark-pattern issue with real downsides:

- **Legal:** false statements in marketing email invite FTC “unfair or deceptive practices” and
  CAN-SPAM scrutiny (subject lines and body must not be misleading).
- **Deliverability:** fake urgency drives spam complaints → Gmail/Apple throttle the whole domain →
  even our *true* and transactional emails stop landing.
- **Trust:** sellers who catch one false “blocked” claim distrust every future message.

**We don’t need to fake it.** Kitse has *real* gates that genuinely block orders. So the rule is:

| Gate type | Examples | Email framing |
|---|---|---|
| **HARD (legal)** | permit, food-handler card, **payout account**, **state registration** | **Truthful urgency** — “you literally can’t be paid / go live until X.” ✓ |
| **SOFT (quality)** | profile photo, story/bio, menu photos, hours | **Value framing** — “chefs with a photo get 2× orders.” Never “blocked.” |

The dashboard and emails are already built this way. The `hard-gate` email *is* the honest version of
the user’s idea — and it’s persuasive precisely because every word is true.

---

## 5. The lifecycle email system

All six are in `emails/` and previewable in `emails.html`.

| # | Trigger | Email | Type |
|---|---|---|---|
| 1 | Application approved | **Welcome & what’s next** | onboarding |
| 2 | Lead flagged priority (e.g. Canné) | **Concierge intro from Maya** | 1:1, soft |
| 3 | Stalled 48h on permit | **Permit help** — which permit, pre-filled, human review | hard, helpful |
| 4 | Live-ready but payout/state undone | **“We can’t pay you yet”** — truthful block | **hard urgency** |
| 5 | Live but no profile photo | **“Photo = 2× orders”** — explicitly says nothing is blocked | **soft value** |
| 6 | All steps complete | **“You’re live” + first-order playbook** | activation |

Design principles baked in: one primary action per email, a named human where it counts, real numbers
not vague hype, and a working unsubscribe/preferences footer on every send.

---

## 6. The redesigned onboarding (what to build)

The demo `onboarding.html` is the spec. Key elements:

1. **Progress ring + activation %** — momentum is visible at every login.
2. **Live status banner** with three honest states:
   - 🔒 *Blocked* — a required legal gate is open (truthful).
   - 🎉 *Live but leaving money on the table* — only soft items remain.
   - ✓ *Fully optimized.*
3. **Two-group checklist** — Required to sell vs Boosts orders, each task showing the *right* kind of
   nudge and a link to the email it would trigger.
4. **Assigned specialist** sidebar — a human on every account; white-glove for priority leads.

---

## 7. “Getting a rate” — how we measure and move it

“A rate” = the **apply → first-order activation rate** (15% today → 30%+ target). You can’t lift what
you don’t measure per stage.

**Instrument (north star + funnel):**
- North-star metric: **activation rate** (apply → first order).
- Per-stage conversion + **median time-in-stage** + the exact task each chef is stuck on (the dashboard
  already holds this state).
- Guardrail metrics: unsubscribe rate, spam-complaint rate, support tickets. *Honesty protects these.*

**Experiments to run (each measured on “advanced a stage within 72h”):**
1. **Send timing** — permit-help at 24h vs 48h vs 72h after stall.
2. **Framing A/B** — hard-gate truthful urgency vs neutral reminder (expect urgency wins *and* keeps
   complaints flat because it’s true).
3. **Concierge vs self-serve** — does a human specialist on every account beat automation enough to
   justify the cost? Roll out to priority leads first.
4. **Subject lines** — “only thing between you and your first order” vs plain “finish your permit.”
5. **Photo nudge value vs vanity** — “2× orders” (data) vs “customers love seeing you” (emotion).

**Cadence:** weekly funnel review; ship one experiment at a time; promote winners into the default flow.

---

## 8. Build order (suggested)

1. **Week 1–2:** onboarding dashboard (required vs boost split, progress ring, live banner).
2. **Week 2–3:** stage instrumentation + the 6 lifecycle emails wired to real triggers.
3. **Week 3:** concierge track + priority-lead flagging (land Canné on it first).
4. **Week 4+:** run the experiments in §7, review weekly, promote winners.

**Net:** same legal rigor, far less silent drop-off, honest nudges that lift the activation rate
without risking the domain or seller trust.