Option C — 15-year TCO + bridging cost for 10,000 deliveries/yr
Two figures requested: (1) total cost of build + ownership of Option C over the 15-year SHA term; (2) separately, the bridging/cross-chain cost for 10,000 stream deliveries per year. Unit basis: a stream delivery = one claim/session; the diagrams assume ~10 deliveries per campaign, and each delivery's payout is one bridge-out crossing (~$1) on the 0%-fee Etherlink bridge (token + LayerZero message in a single crossing), paid by Verifluence.
Part 2 (first, because it's the clean one) — bridging cost for 10,000 deliveries/yr
In Option C the TVL moves to Etherlink (deposit/topup in; withdrawal/refund/release out — see 7. BC-satisfying solution). The rail is the Etherlink official LayerZero bridge (OriginalTokenBridge on Base), which takes 0% on the token — a value crossing costs only the LayerZero message fee (~$0.25–0.85 on-chain; ~$1 used here as a conservative round number) + EtherLink claim gas (~$0.05). Token-bridging therefore costs the same as pure messaging. Per delivery: one release crossing (~$1). 10,000 deliveries ≈ 1,000 campaigns (≈10 each).
| Item | Unit | Volume | Cost / yr |
|---|---|---|---|
releaseFunds LayerZero message (per delivery) | ~$1.00 | 10,000 | ~$10,000 |
dealRegister LayerZero message (per campaign) | ~$1.00 | ~1,000 | ~$1,000 |
| EtherLink claim gas (VF-paid) | ~$0.05 | 10,000 | ~$500 |
| Bridging total — 10,000 deliveries/yr | ≈ $11,500 / yr (~$1.15 / delivery) |
The old "expensive bridging" fear is retired by on-chain evidence. Earlier drafts worried that bridging TVL per-movement could run ~$65–170K/yr at 10K deliveries (assuming ~$5–15 per token bridge + slippage). The live Etherlink bridge disproves that: it takes 0% on the token and only the fixed LZ message fee (~$0.25–0.85, no scaling with amount), so per-event TVL bridging stays at the messaging level (~$11–15K/yr incl. claim gas) — not 6–15× higher. The buffer/batching workaround (DeFi expert #1), whose whole purpose was to cut that cost, is therefore no longer cost-motivated; VF bridges 1:1 per event — cheap and clean per-campaign segregation (see 7. BC-satisfying solution).
The residual concern is solvency, not fees. The escrow holds wrapped LZUSDC (lock-and-mint, backed by the bridge's shared Base lock), so the real risk surface is the cross-chain solvency invariant + the bridge's own security (the post-$292M-exploit class) — not the per-crossing price. The relayer that drives releases, plus a stuck-message recovery queue, remain the #1 audit focus, detailed in 7. BC-satisfying solution.
Part 1 — Total cost of ownership (build + ownership), 15 years
TCO = one-off build + fixed ownership/maintenance + variable cross-chain fees (scale with volume).
A. Build (one-off)
The coding is small: an EscrowGateway (Base) + EscrowLogic (EtherLink), the LayerZero OApp integration, and relayer wiring come to roughly 8 weeks full-time for one engineer (~$20–40K). The dominant build line is the initial cross-chain security audit ($50–100K), not the dev work — for a build subtotal of ~$70–140K. So the expense of Option C is not the build; it's the ownership (maintenance + per-delivery fees) below.
⚠️ Risk to the 8-week estimate: the cross-chain solvency accounting + payout-queue (DeFi expert #1 — Base pays on EtherLink's command from a buffer) is the part most likely to overrun and is the #1 audit focus. 8 weeks fits a basic version; the solvency invariant + queue + buffer-rebalance is where it can stretch.
B. Ownership / maintenance (recurring, ~volume-independent)
This is the cost of keeping a cross-chain system alive for 15 years — roughly 0.2–0.4 of a senior engineer's time plus episodic re-audits. It exists because Option C adds a second chain, a second contract as the source of truth, a cross-chain messenger on the critical path of every payout, and a relayer spanning two chains — each a new thing to fund, watch, and re-audit.
The drivers, with rough annual cost: LayerZero migrations / upgrades — LZ has had major protocol changes (V1→V2, endpoint/DVN/executor config), each forcing re-integration + re-test of the OApp (~$8–15K). Re-audits on contract changes — any change to the cross-chain contracts (bug fix, LZ migration, feature) needs a delta security audit, and cross-chain audits aren't cheap, 1–2×/yr (~$10–25K). Relayer ops — VF is the relayer: keep hot wallets funded on both Base and EtherLink (XTZ gas), plus key rotation, infra, alerting, on-call (~$8–15K). Monitoring & reconciliation — detect stalled/failed LZ messages and keep Base↔EtherLink deal state consistent (funds on Base, truth on EtherLink), with manual remediation of edge cases (~$5–10K). Incident response — more failure modes (messenger outage, reorg, partial delivery → funds stuck mid-settlement) mean reserved engineering time (~$3–8K). Dependency tracking — LayerZero / EtherLink / USDSM / RPC pricing & API changes to adapt to (~$2–5K). Together: ~$30–60K / yr → ~$450–900K over 15 years.
With the buffer design (expert #1), Relayer ops and Monitoring & reconciliation above expand to include buffer rebalancing + cross-chain solvency monitoring — pushing maintenance toward the upper end of the range. It's also the operationally riskiest piece: a solvency mismatch or an empty buffer stalls payouts until rebalance.
Why it isn't near-zero (like today): the current Base setup is one chain, one contract, no message-critical-path — payouts are a direct claim() on Base, so there's almost nothing extra to keep running. Option C's maintenance is the price of standing complexity, and it's owed every year regardless of how cheap the 8-week build was. (For contrast, Option B ≈ $10K/yr: EtherLink is a passive registry written from the backend — no funds cross chains, nothing on the payout critical path.)
- (15 years is long for a cross-chain stack — a LayerZero/EtherLink major change could force a partial rebuild, not captured here.)
C. Variable cross-chain fees (scale with volume)
These dominate at scale because they are per-delivery:
| Volume | Cross-chain fees / yr | × 15 yr |
|---|---|---|
| 10,000 deliveries (~1,000 campaigns) | ~$11.5K | ~$173K |
| 200,000 deliveries (10K campaigns — diagram baseline) | ~$250K | ~$3.75M |
| 1,000,000 deliveries (50K campaigns) | ~$1.25M | ~$18.75M |
TCO (15 years) = A + B + C
| Scenario (steady volume) | Build (A) | Maintenance (B) | Fees (C) | TCO 15y |
|---|---|---|---|---|
| 10,000 deliveries/yr | ~$105K | ~$675K | ~$165K | ≈ $0.95M |
| 10K campaigns/yr (200K deliveries) | ~$105K | ~$675K | ~$3.75M | ≈ $4.5M |
(Build/maintenance shown at midpoints; real TCO scales with growth because fees are per-delivery. Note: build ≈ $105K is ~2–11% of TCO — the cost is ownership, not coding.)
For contrast — Option B 15-yr TCO: simple registry build ~$20–40K, maintenance ~$10K/yr (~$150K), fees ~$45K (at 10K campaigns) → ≈ $0.2–0.25M total — roughly 15–25× cheaper than Option C, for the same on-chain EtherLink presence.
Bottom line
- Bridging, 10,000 deliveries/yr: ~$11–15K/yr. Option C bridges TVL to Etherlink, but the official bridge is 0% fee (verified on-chain), so token-bridging costs the same as messaging — the old ~$65–170K token-bridge fear is retired (see Part 2). Fees are not the Option-C cost driver; ownership is.
- Option C 15-yr TCO: ~$0.95M at low volume (10K deliveries/yr) up to ~$4.5M at the diagram's 10K-campaigns baseline, and it keeps climbing with growth — vs ~$0.2M for Option B.
- The build is the cheap part: coding is 8 weeks of one engineer (~$20–40K); with the audit it's ~$70–140K — only a few % of the 15-yr TCO. Don't let "it's only ~2 months to build" obscure that the cost is the per-delivery fees + ownership that run for 15 years.
Estimates for internal negotiation — not a quote; build/maintenance are engineering judgement, fees are the diagram's per-message basis.