Marketing & Social add-ons — ruling of record

Ruling: add-ons on the existing ladder. Not a new tier. Solo A$119 and Shop A$229 do not change.

The two add-ons:

Channel extensions (modifiers on Social, not new products):

Both add-ons require a base tier (Solo or Shop) — they are not sold standalone. They ride the CS deployment base per §6.5's phasing: after the §7 build-order step-6 eval gate passes on the CS scope.

1. Why add-on and not a new tier

1. The ladder was built to absorb exactly this. [services-pricing-model-aud](services-pricing-model-aud) §8.7 fixes the ladder at Solo A$119 / Shop A$229 / +A$75 extra location, with new offerings slotting in as add-ons A$25–79/mo and the express intent that "infra variable cost is already inside Solo/Shop". Both new products ride that rail. A new tier would reprice mid-pilot, before we have one tenant's real AI-conversation volume (pricing doc §6 lists AI conversation volume as an explicitly uncalculated item).
2. A new tier is a bundle decision, and the bundle is not decided yet. Whether a tenant wants marketing + social + CS + POS is a conversation decision, and the honest answer is we do not know the distribution yet. An add-on lets a tenant say yes to one thing now and add the other later without a migration. A tier says "you are this kind of customer" before we have met one.
3. The products have genuinely different buyers and different cost curves. Marketing (A$39) is catalogue-grounded content drafting — near-zero marginal cost. Social (A$49) carries per-channel connector and platform cost. A tier would force both into one price and hide which one is actually carrying the cost.
4. Phasing matches the build. §6.5's "after CS" holds: the CS surface is the deployment base, and the marketing/social tool sets attach to the spine (§6.1: channel + session principal, tool registry, HIL queue, operator console, feed plumbing) once the eval gate passes.

What would change this ruling: if pilot data shows ≥60% of tenants take both add-ons and their combined usage is materially above the A$79 bundle envelope, a "Growth" tier at the bundle price becomes defensible. That is a data question, and we do not have the data yet.

2. The cost model — and the one number that changes it

2.1 What these add-ons actually cost us

An add-on rides an existing tenant stack, so its marginal cost is not the whole package cost. It is the incremental AI, compute share, and platform fee.

The binding constraint is ops labour, not infra — exactly as pricing doc §8.8 states for every other offering. Onboarding a tenant's marketing assistant (brand voice, catalogue grounding, promo rules) is ~2h; social (connector auth, channel mapping, approval routing) is ~3h. At a A$20/h planning rate that is A$40 and A$60 of month-1 labour.

Month-1 economics: Mktg A$39 − A$5.55 COGS − A$40 setup = −A$6.55. Social A$49 − A$7 COGS − A$60 setup = −A$18.00. Bundle A$79 − A$12.55 − A$100 setup = −A$33.55. Decision: no separate setup-fee line. One month free as a trial covers the setup labour, and the tenant sees one number instead of two. This is also cheaper to explain than the A$199 setup fee in business-plan v1 §4 — that fee was priced for the whole package, and adding add-ons with their own setup fees would quote a tenant three separate line items for what is one product.

2.2 🔴 The platform-cost finding — X and Meta are opposite, WhatsApp is the meter

This is the part that decides the Social price, and it was not knowable from the template catalogue.

Why this decides the ladder shape:

• Meta channels are free to publish on → the Social add-on's cost is our labour, and it can carry a 78–86% margin on a flat A$49. Three Meta channels cost us A$0.00 in platform fees.
• X is not free and its cost is per-post with a 13× penalty for links. A post with a URL costs US$0.200 vs US$0.015 — 13.3×. A retailer's social posts are link posts. So X cannot ride the flat A$49 without a cap; it is +A$29 with 20 posts included, metered after, and the metering rate is cost + 20% (our existing AI-overage convention): plain post A$0.026, link post A$0.344. That is a real per-post margin on A$29 for 20 posts: A$3.06 of cost against A$29, 79% margin floor.
• WhatsApp is the only genuinely variable line — and §4 prices it.

2.3 The CS side is free — and that is a selling point, not a footnote

Meta gives 1,000 free service messages per business phone number per month (resets monthly, does not roll over; from 1 Oct 2026 the 1,001st service message is charged regular rather than free_customer_service). A typical AU SMB doing ~1 conversation/day × 2 messages × 30 days is at ~60 messages/month — A$0.00. EVE's customer-service conversations on WhatsApp cost the tenant nothing in Meta fees up to ~16 conversations/day. Above 1,000/month it is A$0.0148/msg (A$14.80 per 1,000).

This materially improves the CS product's economics versus the 2026-10-13 business plan, which listed AI model calls at US$3–8/tenant/mo without a platform-fee line at all. The WhatsApp conversation cost was missing from the business plan. It is A$0.00 for a typical tenant — the plan was right to omit it, but only by luck, not by analysis.

3. The competitive anchor — and an honest correction to it

All figures live-fetched 2026-10-20, ex-GST for like-for-like.

⚠️ Correction to the vault's existing anchor. business-plan-v2-aud-allowance §3 and business-plan v1 §2 use "Shopify AU Basic ~A$60 + POS Pro ~A$137 ≈ A$197/mo". Live Shopify AU pricing today is Basic A$42 (annual-billed) or A$56 (monthly) + POS Pro A$129 — so the anchor is A$171 annual- billed / A$185 monthly, not A$197. Our Solo A$119 undercuts the correct anchor by A$36.45 (23%), not by A$78 (40%). The pitch weakens but does not break — and the vault number was stale. Flagging rather than silently replacing, because the same figure appears in three documents.

Capability parity — the comparison that actually matters:

Shopify cannot sell AI staff, catalogue-grounded marketing drafting, or social publishing at all. The parity stack is the honest comparison for what we now sell; the bare-anchor comparison is the honest comparison for what Shopify sells.

🔴 The honest limit, stated plainly: against the bare anchor (A$155.45 ex-GST) our Solo + bundle (A$198) is +A$42.55 (+27%). The "cheaper than Shopify" line holds at the Solo tier and for Solo + one add-on (A$158, +A$2.55 — a tie). It does not hold for Solo + both add-ons. Do not quote the add-ons as "still cheaper than Shopify." Quote them against the parity stack or against agency cost, both of which are decisive.

4. WhatsApp — the meter, and the arithmetic envelope

This is the one line that cannot be a flat price, so it is metered. Echo's §8.2 flagged this as blocking the add-on price; the numbers below are the envelope the allowance must fall inside.

Unit economics (AU, AUD card, effective 1 Oct 2026):

— mira, 2026-10-20 22:45 (card t_c2f6df57) — ⚠️ the "We bill" column above is the AU rate only. A$0.1314 = A$0.1095 (Rest of Asia Pacific) × 1.20, and Meta charges by the recipient's country calling code — a tenant's audience is not necessarily Australian. At a single AU rate we lose A$0.0763/msg on Netherlands recipients (A$76.30 per 1,000) and overcharge up to 8× an India or Turkey audience tenant. The billing rule is overage_rate(recipient_country) = Meta rate(country) × 1.20. Full correction and the two new open items: [whatsapp-metering-verification-2026-10-20](whatsapp-metering-verification-2026-10-20). The A$49 line, the 150 allowance and the hybrid model are unchanged.

The A$49 WhatsApp platform line:

Arithmetic envelope for the sibling ruling (t_1d1257c3):

• A$49 line breaks even at 447 marketing templates/mo (cost recovery only).
• A$49 line holds a 60% margin at 179 templates/mo.
• The price is A$49 if the allowance sits at or below 179 templates/mo. Above that the price must rise, or the line must go to pass-through.

— mira, 2026-10-20 — t_1d1257c3 is CLOSED. The ruling:[whatsapp-cost-metering-model-2026-10-20](whatsapp-cost-metering-model-2026-10-20). Hybrid: the A$49 WhatsApp platform line includes 150 delivered marketing messages/tenant/month, then metered overage at cost + 20% (A$0.1314 marketing, A$0.0178 utility/service). The line is sold separately from the Social add-on, never bundled free. 150 sits inside the ≤179 envelope, and on the true (Meta-cost) basis it holds 66.5% margin on the line, not 60% — see the cost-basis note below.

🔴 Correction to the table above (cost basis). The table's "Cost at cost+20%" column is what we would charge, and the "Margin on A$49" column subtracts that figure — so the margins shown are understated and the 447 row is mislabelled. On the true basis (what Meta charges us): 447 msgs = A$48.95 = cost recovery (not A$58.74 / −20%); 179 msgs = A$19.60 = 60% margin; 150 msgs = A$16.43 = 66.5% margin. The stated envelope numbers (447 break-even, 179 for 60%) are on the Meta-cost basis and are correct; only the table's arithmetic mixed the two bases. The ruling in t_1d1257c3 uses the Meta-cost basis throughout.

Structural finding that constrains the billing model (also from t_1d1257c3): as a Solution Partner we become "Bill To Party" and are liable for every tenant's Meta spend under Meta's credit-line terms. Default is therefore the Tech Provider shape — the tenant owns the WABA and pays Meta directly — and we do not resell the platform (Meta's terms prohibit distribute/sell/resell/rent of it). The A$49 line price is unchanged by this ruling.

What I am NOT deciding here: the allowance-vs-passthrough question. That is card t_1d1257c3, a separate Mira run, and it owns the final allowance and the billing model. This doc fixes the envelope so that card's answer is constrained, not free-floating. If that card rules pass-through, the A$49 line becomes a platform access fee with messaging billed at cost + 20% per message above the free 1,000 — and the A$49 line price does not change.

One more Meta fact that shapes this: Meta's own pricing page lists a 1-month minimum advance notice for a rate-card update, and "Rest of Asia Pacific | Marketing | Increase" effective 1 Oct 2026. The A$0.1095 marketing rate is not permanent. Any WhatsApp add-on must carry a rate-review clause — our price and the customer's allowance are both measured against a rate that Meta changes with one month's notice. This is the same class of risk as the Oracle allowance finding in business-plan v2 §4.

5. The published ladder

Both add-ons require Solo or Shop. They are not sold standalone.

The pitch line: "Your store, your POS, your AI staff — and now the AI that writes your campaigns from your own catalogue and posts them to social, with you approving every post. A$198 all-in versus A$369 for the same capabilities bought separately."

6. Open items this ruling leaves

7. Honest limits

• Nothing here is built or metered. These are prices on a design. The Media pipeline, the channel connectors, the metering hooks — none exist yet.
• The AI conversation volume behind A$1.55–3.10 is uncalculated (pricing doc §6). It is a shared-GO estimate with one tenant's worth of data behind it at most. A marketing agent drafting campaigns is not a price-and-lookup chatbot; it is a long-context, multi-tool workload and may cost more per conversation than the CS agent.
• The 1,000-free-service-message allowance is per business phone number, per month. A tenant on 3 phone numbers gets 3,000 free — and one tenant consolidating all its messaging onto one number gets 1,000. We have not modelled whether that changes behaviour.
• X's pay-per-use model is new (Feb 2026) and X has changed its API repeatedly. The US$0.015/US$0.200 rates are current as of 2026-10-20 from docs.x.com, not a contract. A rate change hits our overage price.
• The agency comparison (A$1,500–8,000/mo) is indicative, from secondary AU marketing sources, not a quote we have requested.
• The FX rate moved. I used 1 USD = A$1.4319 (RBA, 20 Oct 2026), not the vault's planning rate of A$1.55 (Oct 2026) — a +8.2% difference that makes USD-denominated costs (X, Jasper) cheaper in AUD than the earlier plans assumed. Meta's WhatsApp card is already in AUD, so it is unaffected by FX. Any figure in this doc that used 1.55 should be re-checked.
• I did not model Zalo, TikTok or GBP. All three remain gated by platform approval (design doc §8.1, §9.4); pricing a channel we cannot access is inventing a product.

Sources

All live-fetched 2026-10-20 unless stated.

Primary (platform):

• Meta WhatsApp Business Platform pricing + AUD list-rates CSV (effective 1 Oct 2026) — developers.facebook.com/docs/whatsapp/pricing; evidence file meta_whatsapp_aud_rate_card_2026-10-01.csv (47 markets), sha256 0fd348d611694227d0bae40ba0d8e1fce322b8aa39ba6221910de8608210fb9f, live 2026-10-20. 47 markets = Meta's own published count (39 named countries + 8 regions) — the card parses complete, no rows dropped. Upstream source file as downloaded (Meta's xlsx): sha256 d874292c9719cb0ae75065a03b654456c6d475dbea9c7e85a6c10fc1a3fcb1a2
• Meta — free-message rules: 1,000 free service msgs/phone-number/mo; free entry point window; 1-month minimum notice on rate-card updates
• Meta — country calling code → market mapping (Australia +61 → Rest of Asia Pacific, 16 markets)
• Meta — rate-change table: 1 Oct 2026, Rest of Asia Pacific, Marketing: Increase
• Meta — Instagram Content Publishing: 100 API-published posts / 24h (not the older 50); Meta charges no publishing fee
• Meta — Facebook Pages API (/page_id/feed): free, no per-post charge
• Meta — Threads API: no published paid tier
• X API pay-per-usage pricing — docs.x.com/x-api/getting-started/pricing: post create US$0.015, with-URL US$0.200, read US$0.005, 3M read cap, no free tier for new developers
• RBA daily exchange rates, 20 Oct 2026: 1 USD = A$1.4319

Primary (competitors):

• Shopify AU pricing — shopify.com/au/pricing: Basic A$42 annual / A$56 monthly; POS Pro +A$129
• Hootsuite plans — hootsuite.com/plans: Standard A$129/user/mo (AUD, ex-tax, annual)
• Buffer pricing — buffer.com/pricing: Essentials US$5/channel/mo annual; Team US$10/channel/mo
• Jasper pricing — jasper.ai/pricing: Pro US$59/seat/mo annual, US$69 monthly

Internal:

• [eve-customer-service-design](eve-customer-service-design) §6.5 (the call), §6.6 item 2 (the prices), §7 item 3, §8.1–8.4 (platform gates + cost), §9.4 (revised platform plan)
• [services-pricing-model-aud](services-pricing-model-aud) §3c–3e (cost model), §8.7 (the add-on rail), §8.8 (labour-priced), §8.9 (Hostinger positioning)
• [business-plan-v2-aud-allowance](business-plan-v2-aud-allowance) §3 (AUD pricing), §4 (allowance finding)
• [business-plan-hosted-commerce-pos-ai-staff](business-plan-hosted-commerce-pos-ai-staff) §4 (setup fee), §5 (COGS)