Koatji City Capture Architecture

# Koatji City Capture Architecture
Version 1.0 | Forged 2026-03-19

---

The One Invariant

A city is captured when 10 accounts in a single zip code are on Odeko auto-reorder and two of them have referred at least one other account.

That is the exit criterion. Everything below is the path to get there.

---

Vocabulary

---

Phase 1: PRIME — Core Truth

The specialty coffee market has one real power structure: baristas.

Owners sign contracts. Baristas make the daily buying decision. The owner who never touches the espresso machine buys what their head barista asks for. The head barista uses what their staff enjoys making. The staff enjoys what tastes good and has a story behind it.

This means:

1. The buying decision is social before it is economic. A shop does not switch oat milk because the price-per-liter improved. They switch because a barista at a nearby shop they respect started using it.

2. Capture is not about persuading one shop. It is about creating a local preference cluster. When 3-4 shops in a neighborhood all use Koatji, the holdouts start asking questions.

3. The minimum viable capture unit is the zip code, not the city. A city is just a collection of zip codes. Win one zip, then bleed into adjacent zips. Never try to capture Miami. Capture Wynwood first.

4. Speed of reorder is the real metric. A won account that reorders in 30 days is a kept account. A won account that never reorders is a sampling event that didn't convert. Track reorder velocity, not sign-up counts.

---

Phase 2: EXPLODE — 6 Radical Approaches

### A. Physical Presence: The Residency Model
Rather than one-off sampling visits, run a "barista residency" — 3 consecutive Tuesday mornings at the same shop, Mohamed behind the bar making drinks for the morning rush. By week 3 the staff knows him, the regulars know him, and two neighboring owners have walked in and asked what that oat milk is.

Key insight: Frequency beats reach in physical sales. One person visiting 10 shops once is weaker than one person visiting 3 shops four times.

### B. Digital Outreach: The Insider List
A private weekly email called "The Shot Log" — 200-word practical notes for specialty coffee operators: what drinks are trending in New York, a wholesale cost comparison on espresso beans, one product recommendation. Koatji is the sponsor, not the subject. The subject is things coffee shop owners actually want to know.

Key insight: Content that is useful to receive builds more trust than content that is persuasive.

### C. Distribution Leverage: The Odeko Warm Introduction
Koatji is already on Odeko. Most specialty shops in target cities already have Odeko accounts. A new account is one click from their first order, not a new vendor relationship. The pitch is not "switch your oat milk." The pitch is "add it to your existing Odeko cart."

Key insight: Odeko transforms the conversion from a new relationship to an incremental action. Close to Odeko, not to Koatji directly.

### D. Community: The Barista Competition Circuit
Sponsor a latte art competition in the target city. Not as a brand presence but as the practice milk. Every barista who trains for the competition for 3 weeks has poured thousands of shots with Koatji. By competition day the product is muscle memory.

Key insight: Competitions create concentrated skill-building events. The barrier to trying a new product disappears when winning requires using it.

### E. Content: The 24-Hour City Capture Vlog
Every field day is filmed. Mohamed walks into a shop, tastes their current oat milk, offers a comparison shot with Koatji, films the barista's reaction. Published to TikTok same day. The shop's handle is tagged. Their followers see it. Three shops in that neighborhood DM asking for a sample.

Key insight: Every sales visit is a content asset. The person watching the video is the next sales call that doesn't require showing up.

### F. Data: Gravity-Guided Drop Sequencing
Build the route before the visit. Score every prospect by Gravity Score, cluster by geography, sequence the drop to hit the highest-score shops in the tightest radius. A field day covering 8 shops in a 1.2 mile radius beats a field day covering 14 shops across 12 miles.

Key insight: Route density is a force multiplier. When shops are close together, a deal at one creates ambient awareness at the next.

---

Phase 3: FORGE — The City Capture Formula

### Stage 0: Market Qualification
Entry: A city or metro on the expansion list.
Exit: Go / No-Go decision + identified zip code target(s).

Actions:
- Pull Odeko availability confirmation for the city.
- Count specialty coffee shops per zip code via Google Maps scrape + Yelp filter.
- Identify 3 candidate capture units (zip codes with 15+ specialty shops, walkable density).
- Score candidate units: existing Koatji awareness, barista competition scene, any warm contacts.
- Select 1 primary capture unit and 1 backup.

Metrics:
- Shops per sq mile in candidate zip.
- Odeko confirmed: Y/N.
- Any existing warm contacts in market: count.

Time box: 3 days max. This is desk work.

---

### Stage 1: Pre-Capture (Market Sweep)
Entry: Capture unit selected.
Exit: 150+ qualified prospects with Gravity Scores assigned, top 30 flagged as priority.

Actions:
- Scrape all shops in capture unit and 2 adjacent zips (buffer zone).
- Enrich each with: Instagram handle, follower count, posting frequency, specialty indicator (third-wave, single-origin, latte art content).
- Assign Gravity Score to each account.
- Flag all Drop-Ready accounts (within 0.5 miles of each other or existing routes).
- Build the outreach queue ordered by Gravity Score desc.

Metrics:
- Total prospects in pipeline.
- Gravity Score distribution (count above 60, 40-60, below 40).
- Drop-Ready count.

Tools: CityScorecard, prospect list CSVs, IG profile scanner.

Time box: 1 week.

---

### Stage 2: Wave 1 — Digital Outreach
Entry: Qualified prospect list ready.
Exit: 15

Actions:
- Touch 1: IG DM to all priority-30 accounts. AI-generated, personalized to their feed (reference a specific drink they posted). No template feel.
- Wait 48 hours.
- Touch 2: Email to any account with a public email. Subject line references the IG message ("followed up on Instagram — here's more detail").
- Wait 72 hours.
- Touch 3: IG DM follow-up for non-replies. Shorter. Offer a specific action ("can I drop by Thursday around 10am?").
- Reply routing: Classify all replies (interested / not interested / already have it / need more info / referral).
- Book sampling appointments from interested replies.

Metrics:
- Reply rate per touch.
- Appointment conversion rate (replies to booked).
- Reply type distribution.

Tools: Mail bridge, IG DM automation, reply classifier (GPT-4o-mini), AccountsTracker.

Rate limits: 100 DMs/day hard cap. 200 emails/day. Pause if account health flags.

Time box: 2 weeks.

---

### Stage 3: Wave 2 — Field Blitz
Entry: 3+ sampling appointments booked OR end of Wave 1 window.
Exit: 10+ shops sampled, 3+ in stage `closing` or `won`.

Actions:
- Build the weekly drop route from Drop-Ready accounts + booked appointments.
- Target 6-8 shops per field day, max 1.5 mile radius.
- At each visit: bring 4 sample cartons, a comparison shot setup (their current oat milk vs Koatji), a leave-behind card with QR to the Odeko product page.
- Film the visit if the owner/barista is willing. 60 second reaction clip.
- For any shop that gives a strong reaction: ask for a referral on the spot ("anyone else nearby you think would want to try this?").
- Log each visit in AccountsTracker same day: outcome, stage, next action.

Metrics:
- Shops visited per week.
- Sampling-to-trial conversion rate.
- Referrals collected per field day.
- Content clips captured.

Field day target: 3 days/week during blitz. Each day = 6-8 shops.

Time box: 3 weeks.

---

### Stage 4: Wave 3 — Conversion
Entry: Account in stage `sampling` for 7+ days.
Exit: Account placed first Odeko order.

Actions:
- Follow up with every sampled account on day 7 (IG DM or text if number was exchanged).
- Message: direct, no fluff. "How'd the oat milk do this week? Ready to add it to your Odeko order?"
- For hesitant accounts: offer a 2-week trial case (cost of goods + shipping, no markup). Remove all risk.
- Walk them through the Odeko add process if needed. Literally send the Odeko product URL.
- For any account that won't use Odeko (some smaller shops buy direct): offer direct delivery until they grow into Odeko.

Metrics:
- Days from sampling to first order (target: <14).
- Conversion rate (sampling to order).
- Odeko onboarding rate vs direct delivery rate.

Time box: 2 weeks per account before moving to nurture or dead.

---

### Stage 5: Wave 4 — Retention and Social Proof
Entry: First Odeko order placed.
Exit: 3 consecutive reorders OR referral given.

Actions:
- Send a handwritten note or voice memo on first order day. Not a template. 30 seconds.
- Week 2: check in. Ask if the baristas like it. Offer to do a free barista training session (30 min, latte art + milk science).
- Post the account on Koatji's Instagram story. Tag them. They repost. Their followers see it.
- Week 4: ask for a referral. "Is there another shop nearby you think would be into this?"
- Monitor reorder date. If no reorder by day 28: follow up. Something went wrong. Find out what.

Metrics:
- Reorder rate (orders 2 and 3).
- Average days between reorder 1 and 2.
- Referrals generated per won account.
- Barista training sessions run.

Time box: Ongoing. This stage never ends.

---

### Stage 6: Wave 5 — Zip Expansion
Entry: Beachhead zip has 5+ active accounts on Odeko.
Exit: Adjacent zip has its own beachhead.

Actions:
- Use referrals from Stage 5 accounts to identify the first priority account in the adjacent zip.
- Use content from existing accounts ("3 Wynwood cafes are using it — here's what they say") as social proof in outreach.
- Run a lighter version of Stages 2-4 in the new zip. The heavy pre-capture work is already done for the metro.
- Repeat until the full metro has a capture unit in each walkable sub-territory.

Metrics:
- Active zips per metro.
- Accounts per zip.
- Referral chain depth (how many hops from beachhead to new account).

---

Phase 4: SYNTHESIZE — Scale Validation

Micro (Single Shop)

What works: The sampling + Odeko handoff flow is clean. One visit, one comparison shot, one QR code. Low friction.

What breaks: If the owner is not present during the visit and the barista likes it but can't make the buying decision. Fix: always ask "are you the owner?" before the comparison shot. If not, ask when the owner is in and come back.

What breaks: If the shop is already on a competing oat milk with a contractual exclusivity clause (rare but exists). Fix: identify this in Stage 1 enrichment by looking for any brand partnerships in their IG bio or posts.

---

Meso (Neighborhood / Zip Code)

What works: The referral flywheel. Once 3+ shops in a zip are won, the remaining shops start asking about Koatji before being contacted.

What breaks: Mohamed running all 6 stages simultaneously across 20+ accounts in one zip. The cognitive load is too high for one person. Fix: time-box each stage strictly. Only 30 accounts in active outreach at any one time. Everything else is queued.

What breaks: The content pipeline dying when field days stop. Fix: batch-edit and schedule clips from field days. One field day should produce 5-7 posts scheduled across 2 weeks.

---

Macro (City / Metro)

What works: The zip-by-zip expansion model. It compounds. Each won zip reduces the CAC in the adjacent zip.

What breaks: Odeko availability gaps. If Odeko doesn't deliver to a specific zip yet, the conversion flow breaks. Fix: identify Odeko dead zones in Stage 0 and route around them with direct delivery until Odeko catches up.

What breaks: Trying to blitz two cities at once as a solo operator. The field work is local. You cannot blitz Miami and New York simultaneously. Fix: one city at a time until a city is at 30+ active accounts, then hire a city rep and hand off the playbook.

---

Phase 5: CREATE — SCAMPER Analysis

Substitute: Replace cold DMs with warm introductions via Odeko reps. Odeko has a sales team that already knows these shops. An introduction from an existing vendor rep is worth 10 cold DMs.

Combine: Combine the barista training session with a micro-event. Invite 3-4 baristas from nearby shops to one location. One training session, 3 new sampling opportunities, content that shows community not just product.

Adapt: The latte art competition model used by espresso brands in Australia (common there) is nearly absent in US specialty markets. First mover advantage exists in most US metros.

Modify: Shorten the Wave 1 window for high-Gravity accounts. If a shop scores above 80, skip touch 2 (email) and go straight to a phone call or an in-person drop. High Gravity means high readiness. Don't slow it down with an email sequence.

Put to other use: The 1,200 SFL prospect list is a distribution list for "The Shot Log" newsletter even for shops that never convert to customers. They remain aware of Koatji and may convert in 6 months.

Eliminate: The three-email sequence for accounts scoring below 30. Low-Gravity accounts should get one IG DM and then move to field-only if route-proximate. Don't spend sequence budget on low-signal accounts.

Reverse: Instead of Mohamed going to shops, run a quarterly "Koatji at Origin" event — Mohamed hosts coffee shop owners for a 90-minute product and sourcing deep dive. Owners come to him. Positions Koatji as education-forward, not just a supplier.

---

Phase 6: EVOLVE — Stress Tests

Stress Test 1: Mohamed has 3 salespeople

What changes: Each salesperson owns one capture unit. Mohamed owns the playbook, the brand voice, and the content. Salespeople execute Stages 2-4. Mohamed handles Stage 5 (retention) for VIP accounts and Stage 6 (expansion strategy).

New requirements:
- AccountsTracker must support multiple salesperson views.
- Reply routing must route hot leads to the right rep, not just to Mohamed.
- Weekly sync: each rep reviews their capture unit metrics with Mohamed. 30 minutes.
- Compensation: base + per-account-activated commission. Reorder bonus after account hits 3 consecutive orders.

Risk: Voice inconsistency. Each rep will have a different style. Fix: a 2-page "How We Talk to Coffee Shops" guide plus the first 10 outreach messages for each rep are reviewed by Mohamed before sending.

---

Stress Test 2: Odeko is not available in a city

What changes: The conversion flow requires direct delivery until Odeko expands.

Fallback stack:
- Stage 4 becomes: "Place an order directly at koatji.com" with a custom discount code for the account.
- Mohamed or a driver makes weekly drops to active accounts.
- Set a threshold: if direct delivery accounts in a city exceed 15, the unit economics likely justify a city-specific delivery run once weekly.
- Actively lobby Odeko to expand: provide them the account list with Gravity Scores. They want this data.

Risk: Direct delivery burns Mohamed's time at scale. Fix: cap direct delivery accounts at 20 per city until Odeko arrives. Any account above that cap goes on a waitlist with a 30-day re-ping.

---

Stress Test 3: Reply rate drops to 5

**5

**At 5
- Immediately audit the last 50 DMs and 50 emails. Is the personalization actually personal or has it drifted toward template-feel?
- Check if a competing brand ran a promotion in the same period. External suppression is different from message failure.
- A/B test 3 new hooks on the next 30 outreach messages.
- Increase Drop-Ready field visits as the primary channel while digital recovers.

**At 3
- Pause digital outreach in the affected channel for 2 weeks.
- Run pure field blitz: no DMs, just walk-ins with samples.
- After 2 weeks, restart digital with fully rewritten copy based on what worked in field conversations.

The underlying principle: Field visits have a floor reply rate of roughly 40

---

The City Capture Formula (Summary)

Stage 0: Market Qualification       3 days      Desk work only
Stage 1: Pre-Capture (Sweep)        1 week      CityScorecard + enrichment
Stage 2: Wave 1 — Digital           2 weeks     DMs + email, 30 priority accounts
Stage 3: Wave 2 — Field Blitz       3 weeks     6-8 shops/day, 3 days/week
Stage 4: Wave 3 — Conversion        2 weeks     Odeko close, reduce all friction
Stage 5: Wave 4 — Retention         Ongoing     Reorders + referrals + content
Stage 6: Wave 5 — Zip Expansion     When ready  5+ accounts in zip 1 = open zip 2

Total time to beachhead: 8-10 weeks from market qualification to first won account.
Total time to Hot Zone: 16-20 weeks (5+ accounts in one zip, referrals self-generating).
Total time to captured city: 12-18 months (city-wide presence via zip-by-zip expansion).

---

Metrics and KPIs by Stage

---

Resource Requirements

Solo (Mo, current state)

• Weekly capacity: 3 field days (18-24 shop visits), 100 DMs, 200 emails, 1 content batch.
• Accounts in active pipeline: 30 max. Everything else queued.
• Cities in play: 1 active. 1 in pre-capture prep.
• Bottleneck: Field visit time. The system can produce more leads than Mo can convert in person.

Small Team (Mo + 2 city reps)

• Weekly capacity: 6 field days per city, 300 DMs, 500 emails.
• Accounts in active pipeline per rep: 30. Total 90 across team.
• Cities in play: 2-3 active.
• Bottleneck shifts to: Reply routing and rep training, not field time.

Scaled Team (Mo + 5+ city reps)

• Mo's role: Playbook, content, brand, strategic accounts only.
• Reps own: Full stages 2-4 in their assigned city.
• New requirement: A CRM that is not spreadsheet-based. Stage 5 becomes a dedicated account success function.
• Bottleneck shifts to: Mo's bandwidth for content and culture, not pipeline volume.

---

Scaling Rules

Hire a city rep when:
- A city has 15+ active accounts and Mohamed is the only person servicing them.
- Mohamed is making 10+ field visits/week in a single city (unsustainable with other cities in queue).
- A city has a Hot Zone and spillover is happening faster than Mohamed can convert it.

Open a new city when:
- Current primary city has 30+ active accounts, 2+ Hot Zones.
- Referral rate in current city is self-sustaining (minimal cold outreach needed to maintain pipeline).
- Odeko confirmed in new city.
- A warm contact exists in new city (lowers ramp time significantly).

Do not open a new city when:
- Current city conversion rate is below 30
- Reorder rate is below 50

---

Anti-Patterns

The City-Wide Launch Trap. Announcing "we're now in New York" and cold-blitzing 500 shops simultaneously. Volume without density. You get 10 accounts spread across 50 square miles who never cross-pollinate. None of them ever see each other using Koatji. The social proof flywheel never starts. Win one neighborhood before announcing a city.

The Email-First Habit. Email feels scalable. DMs feel manual. The data says DMs convert 3-5x better for this category. If you default to email because it's easier to batch, you're optimizing for your workflow, not for results. DMs first. Always.

The Sampling Without Follow-Up. Leaving a case of product and never following up is not a sales strategy. It's a donation. Every sample has a 7-day follow-up on the calendar before it leaves Mohamed's hands.

The Conversion-Without-Reorder Metric. Counting won accounts without tracking reorders. A shop that ordered once and stopped is not a captured account. It is a failed test. The only number that matters is reorder velocity.

The Overbuilt Tech Trap. Building more pipeline tooling before using the current tooling at capacity. The CityScorecard, AccountsTracker, and mail bridge are enough to capture 3 cities. Do not build more software until the process is proven.

Chasing the Whole City. Miami is not the target. Wynwood is the target. When Wynwood has 8 active accounts, Brickell becomes the target. Never name a whole city as the goal. The goal is always the next zip code.

Selling to Owners Who Don't Touch the Product. Some coffee shop owners are investors, not operators. They will say yes and then the baristas will never use it. Always get confirmation from the person who actually makes the drinks. The handshake that matters is with the head barista, not the owner on a laptop in the corner.

---

Appendix: The Beachhead Playbook (First Won Account in a New City)

The first win is the hardest and the most important. It sets the proof point for every subsequent outreach.

1. Identify the most respected specialty shop in the target zip. Not the biggest. The most respected by other baristas.
2. Go in person, no prior DM. Order a cortado. Watch how they work.
3. Come back the next day. Same thing.
4. Third visit, introduce yourself to the barista or owner. Talk about the drink, not about Koatji. Ask what they're using. Show genuine interest.
5. On the fourth or fifth visit, offer the comparison shot. By now you're not a cold salesperson. You're the guy who comes in every week and appreciates what they do.
6. When they agree to try it, bring the case personally. Not via Odeko yet. In person. Make it feel like a gift from one coffee person to another.
7. The first reorder is on Odeko. By then the relationship is warm enough that they'll set it up in 5 minutes.

The beachhead account is not just a customer. It is the proof of concept you cite in every outreach in that city for the next 6 months.

---

Document generated by Koatji Content Engine — 2026-03-19
Stage 3: FORGE output from the 6-phase City Capture creative pipeline