← All reports · published 2026-04-19

Legacy Soil & Stone — Cement Phase 1 Launch Readiness

*Compiled 2026-04-18 (G16 Claude). Research only — no bench testing assumed or required. Path 1 (cement) is the Phase 1 SKU set; novelty paths 3/5/6 stage as Phase 2.*

Executive summary

• Cement-line Phase 1 can take its first paying order in 3 to 4 weeks. Not 1 weekend. Not 3 months. The blocker is not chemistry, recipe, pricing, or website — those are green. The blocker is the inquiry-to-payment-to-shipping operational seam, plus 4 first-customer dry runs of the 28-day cement cure that nobody can shortcut.
• The architect work to close the gap is ~2 working sessions of mine (intake form deploy, payment account wired, email templates, shipping protocol doc). The calendar drag is curing time and 3 BOSS decisions Mark has to actually make (Stripe vs alternatives, return-shipping carrier, refund policy).
• The honest first-order ship date is the cure-cycle floor of the slowest SKU the customer orders. Worry stones: ~2 weeks intake-to-ship. Garden stones / planters: ~5-6 weeks intake-to-ship (28-day outdoor cure is non-negotiable for the planter; 21-day for garden stones). Setting customer expectation up-front IS the launch — not after the order.
• Phase 2 conversion gate is research-completed AND 25 paid Phase 1 orders OR $7,500 cumulative revenue, whichever comes first. Not bench tests. Research lands when the Path 3/5/6 docs are signed off; orders are the demand signal that says "yes, expand the catalog." Both gates have to fire.
• The biggest risk is not chemistry. It's the 4-6 week cure window combined with grief-state customers who expected Amazon-speed. The mitigation is brutally honest timing on the order page, the intake email, and the website. Not in fine print.

Green: what's already done

That is a real list. Most launches don't have a quarter of this assembled. The hard work of the last six weeks is paid for.

Gaps: what's actually missing

Architect-closeable (I do these — no Mark labor)

G1. Order-intake page on legacy.thebarnetts.info. The public site today is an *investor handoff package*. There is no "begin an order" button. Customer Intake Form lives in Drive as a markdown doc, not a fillable web form.

• *What I do:* port Customer_Intake_Form.md to a static HTML form on the existing site (new page /order.html linked from the homepage). Form posts to a Cloudflare Worker that emails the submission to Mark + writes a row to a Cloudflare D1 table for tracking. Uses the existing deploy pipeline. ~3 hours of architect work.
• *Cost:* $0 (Cloudflare free tier covers Workers + D1 at this volume).

G2. Customer communication templates. First-touch acknowledgment email, "your cremains were received" email, "production has begun" email, "shipping today" email, "share a photo of the finished piece if you want" email.

• *What I do:* draft all five templates in Mark's voice (Bob Ross standard, no funeral language, integrity pillar wording). Stored as a single markdown file in Drive Operations/Customer_Email_Templates.md. Mark reads + tweaks language; I wire them into the Worker so they fire automatically on intake submission and on order-status changes.
• *Time:* 2 hours architect + 30 min Mark review.

G3. Order tracking spine. Where does "order #1" live? Right now: nowhere.

• *What I do:* a 5-column Cloudflare D1 table (order_id, customer name, intake date, SKU, status enum) plus a one-page private dashboard at legacy.thebarnetts.info/admin/ (basic auth, same pattern as bidetai). Status enum: intake_received, cremains_received, in_production, curing, shipped, delivered, closed. Mark can flip statuses from the dashboard; status flips trigger the matching email from G2.
• *Time:* 4 hours architect.

G4. Shipping protocol document. What box, what padding, what insurance, what label, what tracking, what does the unboxing look like, what's in the package besides the stone (memorial cloth, certificate, seed-paper card per Line_One_Process step 7).

• *What I do:* research USPS Priority Mail vs UPS Ground for fragile-cement-up-to-7lbs at insured value $200-$500, draft a one-page shipping SOP, source the boxes/cloth/cards needed (a Uline order, a Moo card order). I bring the cost-estimate sheet back; Mark approves the spend.
• *Time:* 3 hours research + a BOSS decision (G-B2 below).

G5. Legal disclaimers and refund policy text. Two pages: Terms (what we promise, what we don't, force majeure on cremains shipping) and Refunds (the inevitable "I changed my mind" / "it broke in shipping" / "the color isn't what I expected" cases). Cement memorial work has a specific failure mode the customer needs to understand: *cement is permanent; once your cremains are in the slurry, we cannot un-make the piece.* That belongs in writing.

• *What I do:* draft both. Use small-business memorial-craft norms as the baseline (no lawyer needed at $300 ASPs; revisit with counsel before $50K cumulative revenue). Mark reads and ships, or kicks back and I revise.
• *Time:* 3 hours research + draft.

G6. Product photography. The website currently has zero photos of finished pieces. Customers ordering a $395 planter sight-unseen need to see something.

• *Reality check:* without bench testing, Mark has no finished pieces to photograph. This is the one gap research can't close, because the asset doesn't exist yet.
• *What I do near-term:* commission AI-generated *concept renderings* labeled clearly as "design concept — actual hand-finished piece will vary" on the order page. Use the same Gemini imaging stack already in flight for the Grind project. Diversity rule applies (feedback_diversity_in_generated_images.md) to any human/hand imagery.
• *What I do longer-term:* once first paid order ships, photograph it before it leaves and replace concept renders with real product photos. Photo-taking is a 10-min Mark task per order, not a project.
• *Time:* 2 hours of image generation + 2 hours of layout.

G7. Admin daily-digest. Mark needs to know "an order came in" without checking a dashboard. Tie order-intake events into the existing morning-digest pipeline + a Slack/Telegram ping.

• *What I do:* one Worker route adds new orders to today_agenda.md and pings Slack #legacy-soil channel. Reuses existing notification infrastructure.
• *Time:* 1 hour.

BOSS decisions (these need Mark, framed as questions)

G-B1. Payment processor — Stripe, PayPal, or Square?

*Question for BOSS:* "Want me to set up Stripe (best API, 2.9%+30c, 2-day payout, requires SSN/EIN to open the account)? Or PayPal (familiar to older customers, 3.5%+, can open with email only)? Or Square (good if you want in-person sales at any future craft show, similar fees to Stripe)?" My recommendation: Stripe. Cleanest API for the intake form, familiar to grief-state customers, the API support means I can wire it into the order page directly. Mark gives go-ahead + provides SSN/EIN at signup time; I never see the credentials.

G-B2. Shipping carrier and insurance level.

*Question for BOSS:* "USPS Priority Mail Cubic up to 7lb runs ~$12-18 with $100 free insurance, additional insurance ~$3 per $100. UPS Ground similar weight ~$14-22 with $100 free, additional ~$2 per $100. UPS handles fragile better but costs more. Want me to default Priority Mail with $300 declared value (~$25/order baked into intake fee tier already)?" My recommendation: USPS Priority Mail Cubic, $300 declared value. Already inside the existing intake fee math.

G-B3. Refund policy — three options, each with tradeoffs.

*Question for BOSS:*

• *Option A — No refunds once cremains are received.* Cleanest, hardest to defend in a grief context, may cost reviews. Industry-standard for memorial work though.
• *Option B — Full refund before cremains arrive; replacement-only after.* If the piece breaks in shipping, we re-cast (can't, no more cremains) or refund the difference between purchase price and a small "consolation" SKU. Honest tradeoff.
• *Option C — Insurance-backed: declared value covers shipping breakage; production defects covered by re-cast from any pearls remaining; intake fee non-refundable once cremains arrive.* Most professional, most paperwork.

My recommendation: Option C. Matches the integrity pillar of the brand; gives the customer a real answer; defensible.

G-B4. Domain for transactional emails. Right now legacy.thebarnetts.info is a subdomain of thebarnetts.info, which has SPF/DKIM tied to Mark's grandfathered Workspace. *Question for BOSS:* "Want order emails coming from orders@thebarnetts.info (uses your existing Workspace mailbox; you'll see customer replies in Gmail) or orders@legacy.thebarnetts.info (separate identity; cleaner brand split; needs a small SPF/DKIM addition)?" My recommendation: orders@thebarnetts.info. Zero extra config, replies land where Mark already reads mail.

G-B5. The first-order safety net. *Question for BOSS:* "Do you want me to flag the very first paid order with a 'PAUSE — manual review' status before any cremains shipping label is sent to the customer? Catches a misfire on the new system before the first grief-state customer is involved." My recommendation: yes, for orders 1-3, then auto-flow takes over.

Timeline to first paying order

Calendar dates from today, Saturday 2026-04-18.

Realistic first-paid-order date: 2026-05-09 to 2026-05-23 (2-4 weeks after site opens). First *shipped* order: add the cure window (worry stones +2 weeks, planter +5-6 weeks).

If Mark hits the BOSS decisions today instead of Sunday, every downstream date pulls left by 24 hours.

If a single BOSS decision stalls (e.g., Stripe verification gets a doc request), each day of stall = 1 day of slip on the open-for-orders date but not on first-customer date (the customer pipeline is still warming).

Phase 1 order workflow

Inquiry → Intake → Production → Cure → Ship. Four touch points, three of them automated.

Step 1 — Inquiry (automated)

Customer arrives at legacy.thebarnetts.info. Hero: "Begin Your Memorial." Click → /order.html. Reads Bob Ross standard + integrity promise + the honest cure-time disclosure above the form. Fills the form (the existing intake doc rendered as web form). Selects SKU(s), intake fee tier, address. Stripe checkout for product price + intake fee + shipping. On payment success: order #N is created in D1, status intake_received, "thank you / here's how to ship cremains" email auto-fires, Slack pings #legacy-soil, Mark sees it in tomorrow's morning digest.

*Mark labor: zero. He sees a notification.*

Step 2 — Intake / cremains receipt (semi-automated)

Customer ships cremains to Mark via the prepaid USPS label included in the email. When the package arrives, Mark opens admin dashboard → Mark received button on order #N → status flips to cremains_received → "we have them, production begins" email auto-fires.

*Mark labor: 1 click + opening the package.*

Step 3 — Production (Mark's actual craft)

Per Line_One_Process.md. Sieve → granulate to pearls → mix per SKU recipe → cast → vibrate. End of session, Mark clicks In production → status in_production → "your piece is being made" email auto-fires.

*Mark labor: this is the work. Filmable per the Bob Ross standard if Mark wants. 20-30 min active per piece.*

Step 4 — Cure (zero labor, calendar runs)

Worry stones 2 days + 5 days air-dry = 7 days. Candle holders 2 days + 14 days = 16 days. Garden stones 7 + 21 = 28 days. Planter 7 + 28 = 35 days (60 days before plant — but ship at 35). Mark clicks Curing once cast is in the rack.

*Mark labor: zero. Just calendar.*

Step 5 — Finishing + ship

Microcement overlay, polish, embed candle cup if applicable, sealer, sign certificate, wrap in memorial cloth, photograph (replaces concept renders on the website over time), pack, USPS Priority Mail label printed from the dashboard, drop at post office. Click Shipped → tracking number entered → "your memorial ships today" email + tracking auto-fires.

*Mark labor: 30-45 min finishing + 10 min packing + drive to post office.*

Where automation lives vs where Mark lives

• Automated: all customer-facing emails, all status notifications, all payment processing, all order tracking, all "what's coming up" reminders to Mark, all post-shipping followup.
• Mark's hands: receiving the package, the actual production work, the finishing + photo + pack-and-drop.
• Mark's screen: ~5 button clicks across the lifetime of an order. No copy-paste anywhere.

Phase 2 conversion gate

Phase 2 = activating Path 3 (cold-cast polyester), Path 5 (crystal geode + resin hybrid), Path 6 (pearl-accretion rolled) as paid SKUs alongside the cement line.

Gate: BOTH conditions must fire before any Phase 2 SKU goes live.

Condition A — Research completion gate

• Each Phase 2 path has a documented research dossier covering: materials list with vetted suppliers, finished-piece reference photos from comparable artisans, durability claims with citations, failure-mode analysis, COGS calculation, suggested catalog price within the existing $175-$395 ladder (no premium tier — feedback_no_upselling.md), and a written answer to "what does the customer see when this goes wrong."
• The dossier doesn't require a Mark bench test. It requires *me* finding existing comparable work, supplier specs, and chemistry references that say "this works." Per feedback_no_bench_testing_research_only.md.
• Estimated research effort: ~6 hours per path. Phase 2 research load is ~18 hours of architect work, parallelizable with Phase 1 operations.

Condition B — Phase 1 commercial signal

• 25 paid orders cumulative OR $7,500 cumulative revenue (whichever fires first).
• 25 orders is the smallest sample where a 4-SKU mix shows real preference data. Less than that, you're guessing. More than that, you're delaying.
• $7,500 is roughly the working-capital floor for adding a second material chemistry (resin + pigments + pressure pot per Path 3, or alum/MAP + grow shelf per Path 5) without crowding out Phase 1 cash needs.
• Either bar firing means demand is real and the cement line isn't a fluke.

When both fire: Phase 2 SKUs go live one at a time, 2-week stagger between each, with a "research-only" launch posture (assume the production process works as designed; first paid order is the validation).

The gate is intentionally conservative on the demand side. This protects Mark from launching a polyester resin SKU and then having to support 3 unfamiliar material chemistries simultaneously when Phase 1 is still finding its footing.

Risks + mitigations

R1. Cure-time customer expectation mismatch (HIGH likelihood, HIGH impact)

A grieving customer pays $395 for a planter, expects it in 2 weeks, gets it in 6. Result: angry email, refund request, 1-star review.

• Mitigation: Cure time is *the first thing the order page shows*. Above the SKU list. Bold. Plain language: "Your piece will ship 2-6 weeks after we receive your loved one's cremains, depending on which product you choose. This is concrete — it cures slowly because slow cement is strong cement, and we don't rush the work that holds your loved one." Repeat in the intake confirmation email. Repeat in the production-begins email. The honesty IS the brand vs Parting Stone (6-10 weeks per their site, openly published).

R2. Cremains lost or damaged in customer-shipped intake (MEDIUM likelihood, EXTREME impact)

USPS loses 1 in 1,500 Priority Mail packages. If it's the cremains, there is no second chance.

• Mitigation: Prepaid USPS Priority Mail with signature confirmation + $100 declared value (free) included in the intake instructions. Customer instructed in the email to *photograph the cremains and the sealed shipping container* before drop-off. If a package is lost, we have proof of contents. Refund the intake fee; offer a 50% discount on a future order (no guarantee of restoring what was lost — be honest). This is a published policy, not a surprise.

R3. Shipping breakage on the outbound piece (MEDIUM likelihood, HIGH impact)

$395 planter, 7 lbs, drops 4 ft off a porch. Cement breaks. Customer gets shards.

• Mitigation: USPS Priority Mail with $300 declared insurance per G-B2. Foam-in-place packaging or molded pulp insert (researched + sourced in G4). Customer's refund policy per G-B3 Option C: insurance covers cost of the piece; no re-cast possible because no more cremains. Customer hates this answer but the policy is on the order page before they buy.

R4. Cement intake fee miscalibrated against actual cremains volume (MEDIUM likelihood, MEDIUM impact)

Intake fees ($25-$150) are a per-volume labor proxy. If the average customer ships heavier than estimated, margin compresses on the bottom tier. If lighter, customers feel overcharged.

• Mitigation: First 5 orders: I record actual cremains weight against intake tier paid. After 5 orders, recalibrate if delta > 20% in either direction. Per project_legacy_soil_pricing.md "to revisit" note this was already planned. I'll surface the comparison in a 5-order recap automatically.

R5. Concept renderings vs reality gap (MEDIUM likelihood, LOW-MEDIUM impact)

AI renders look prettier than the first hand-cast pieces will. Customer sees gap, feels deceived.

• Mitigation: Renders labeled clearly. First-customer policy: if the finished piece is *visually significantly worse* than the concept render in the customer's judgment (subjective — that's intentional), they get the intake fee back as goodwill. Once a real photo of a real piece is on the site, this risk closes. Ship 5 orders, replace renders.

R6. Mark personally overwhelmed by first 3 orders happening simultaneously (MEDIUM likelihood, HIGH impact on Mark, LOW impact on business)

School, mini-bus duty Tue/Thu, evening crash after 6th period (user_typical_day.md), introvert recovery time. Three intake packages arriving in the same week could land in flow-state-blocking territory.

• Mitigation: The order tracking dashboard surfaces "you have N pieces in production, M curing, K shipping this week" in the morning digest so Mark sees it without having to think. Per user_work_rhythms.md the antagonist is overwhelm, not procrastination — make the work load visible so it's never a surprise. If a 3-order week shows up early: I can hold new order intake at the order page (toggle a flag in D1) for 7 days while Mark catches up. Customers see "we are at capacity this week, please leave your email and we'll notify you when we re-open" — honest, doesn't break trust, common in artisan craft sites.

R7. State Vet / Stream B ambiguity bleeds onto Stream A perception (LOW likelihood, MEDIUM impact)

Customer reads about NOR/composting on the site, gets confused about whether their cremains will be composted (they won't — Stream A is concrete only). Refund or chargeback before any work happens.

• Mitigation: Order page is Stream A only for Phase 1. Stream B (NOR, composting) is documented in the investor materials but not orderable. Reduces blast radius until Stream B regulatory work clears.

R8. The investor-handoff site doesn't convert grief-state customers (HIGH likelihood, MEDIUM impact)

The current site is written for investors. Grief-state customers need shorter sentences, more whitespace, fewer bullet points, more emotional resonance, less COGS analysis.

• Mitigation: New /order.html page is a customer-facing rewrite — Bob Ross standard, story-led, short. The investor docs stay where they are (linked from a footer "About / Investor Materials"). Two audiences, two surfaces, one domain. Done in G1.

What this plan deliberately does NOT include

• A premium tier or upsell at any point. Per feedback_no_upselling.md. The cement-line catalog stays at $175 / $225 / $295 / $395 forever.
• Any "Mark try X" task. Per feedback_no_bench_testing_research_only.md.
• Any copy-paste shell commands aimed at Mark. Per feedback_no_control_no_copypaste.md.
• A sales channel via funeral homes or vet clinics. D2C only for Phase 1 per feedback_legacy_soil_closed_topics.md — the vet channel debate is parked.
• Any commitment to the State Vet approval timeline. Stream B stays parked. Stream A stands alone.
• Reopening the efflorescence/ASR debate. Closed by binder reformulation. Per feedback_legacy_soil_closed_topics.md.

Single next-action

If Mark wants to move on this: the only thing standing between today and "site is open for orders on 2026-04-25" is the 30-minute BOSS-decision pass on G-B1 through G-B5. Everything downstream is mine to execute.

If Mark wants to stall on this: that's also fine. The plan keeps. The website keeps serving the investor handoff package as it does today. No money has been spent. No customers are waiting.

The cement line is real, the math is real, the chemistry is solved. The only question is whether Mark wants the order door open in 7 days.