Agentic Payments — Forward Revenue Stream¶

Created: 2026-05-06 Updated: 2026-05-08 — AWS Marketplace agentic payments pilot (fiat-settled, end-of-month) launching ~June 1, 2026 reframes priority. AWS MP is now the primary rail; x402 USDC is secondary. Status: Pre-pilot readiness work in May 2026. Full agentic ramp Q3 2026. Parent docs: PRICING_ALIGNMENT_AUDIT_2026_05_06.md, PRICING_PAGE_COPY_GTM_POSITIONING.md

TL;DR¶

Three rails are converging on the same architectural bet: AI agents pay per-request without accounts, API keys, or invoices.

AWS Marketplace agentic payments (primary) — fiat-settled (USD), end-of-month invoice from AWS to buyer, AWS pays Vellocity per the existing seller-of-record arrangement. Pilot launches ~June 1, 2026. This is the dominant rail for Vellocity because (a) Vellocity already has a seller account, (b) the metering dimensions are already seeded, (c) fiat settlement is acceptable to enterprise finance teams without crypto accounting.
x402 / USDC (secondary) — Coinbase-driven micropayment protocol, settled in USDC per call. Better fit for non-AWS-resident agents and aggregator marketplaces like agentic.market.
Stripe machine-payments protocol (tertiary) — announced but not GA. Watch.

Vellocity already has the substrate to monetize all three: a credit-per-capability system on /3pi-partners and an MCP server exposing 50+ capabilities. Same rate card from PRICING_ALIGNMENT_AUDIT_2026_05_06.md § "Usage-Based Pricing Layer" powers all three rails.

This is a third buyer persona (subscription buyers, AWS Marketplace human buyers, and AI-agent buyers) — net-new revenue that does not cannibalize the tier subscriptions, because no human prefers cent-by-cent micropayments over a flat $1,499/mo plan.

Strategic Context: AWS Internal Agent Ranking + Listing-Bloat (May 2026 field intel)¶

Two field-intel observations that make Vellocity's discovery + agentic positioning unusually defensible:

AWS's internal agent-ranking algorithm uses cross-source consistency signals. Specifically, agents weight: matches/similarities between (a) the seller's own website, (b) the AWS Marketplace listing, and (c) third-party signals like G2 reviews and customer feedback. The more aligned these surfaces are, the higher the listing ranks in agent-driven discovery.
Partners are gaming this incorrectly. Established partners are spawning 5+ near-duplicate listings differentiated only by industry vertical, believing volume helps discoverability. Under the cross-source consistency model, this actively hurts — split signals across listings dilute the consistency score, and similarity between sibling listings degrades match precision.

Why this matters for Vellocity: - The "Marketplace SEO + AI visibility" capability already pitched for Command tier becomes far more valuable when reframed as "align your signals before you spawn listings" — Vellocity is one of the few tools that could actually score and improve cross-source consistency programmatically (website crawl + listing pull + G2 fetch + similarity scoring). - The agentic-payments rail launching ~June 1 is paired with the agent-ranking algorithm. Sellers ready on Day 1 get tested against agents that prefer high-consistency listings. Vellocity customers who fix their cross-source consistency with the platform's content tooling rank better in the same pilot. - This is also a cleaner pitch to SMBs (per the May 2026 ICP note): they don't have legacy listing bloat to undo, so they can be set up consistent from Day 1.

Full treatment: AWS_MARKETPLACE_DISCOVERY_DYNAMICS.md — the load-bearing strategy artifact built from this thesis. Includes messaging arsenal (talking points, audience tiers), the failure-mode analysis of why listing-bloat backfires under cross-source consistency, the product roadmap reframe (which Vellocity engines need the catalog → 3P → web lens), and the open-work table tying the thesis back to GAP 1 in ../aws-agent-mode-ranking-audit.md.

Why AWS MP Agentic Is The Primary Rail (May 2026 reframe)¶

The earlier framing (Q3 2026, x402-first) assumed Vellocity would build agentic settlement from scratch via aggregators. The AWS MP pilot inverts this:

Factor	AWS MP agentic (primary)	x402 USDC (secondary)
Settlement currency	USD (fiat)	USDC stablecoin
Settlement cadence	End-of-month invoice	Per-call, instant
Buyer onboarding	Already on AWS account	Wallet provisioning required
Vellocity readiness	Seller-of-record + metering dimensions already seeded	Endpoint code not yet built
Buyer enterprise-friendliness	Standard SaaS line item	Crypto-account question for finance teams
Pilot window	~June 1, 2026	Already live on agentic.market
First-mover positioning	High — only ready sellers participate in pilot	Medium — open marketplace

For Vellocity, the conversion math favors AWS MP: nearly zero net-new infrastructure to participate, fiat settlement removes a buyer objection, and being a launch-day participant in the AWS pilot is an enormous credibility marker for the broader 3PI partner pitch.

x402 stays on the roadmap because: 1. Non-AWS agents still need a rail 2. Aggregator marketplaces (agentic.market) drive their own discovery flywheels 3. USDC settlement may dominate certain verticals (crypto-native ISVs, autonomous agent operators)

But AWS MP is the work to get done first.

Why this fits Vellocity specifically¶

Credits-per-capability already exists. The /3pi-partners "Capability Gap" table prices each capability in credits today (10–30 cr / 20 cr / 25 cr per call). That's a unit-priced model already; x402 is the same model with a different settlement layer (USDC stablecoin instead of subscription tokens).
MCP server is already exposed. Vellocity's MCP capabilities are addressable today, but require API keys + accounts. x402 lets the same endpoints monetize against a buyer who can't (or won't) provision an account — the AI agent itself.
No cannibalization with subscription tiers. Human buyers prefer flat $X/mo for predictability. Agents prefer per-call because they don't have a budget cycle, just a task. The buyer overlap is near-zero.
First-mover positioning in the GTM-for-Marketplaces category. None of Tackle, Suger, Clazar, Labra, or Feenix has signaled x402/agentic payments. Being first to expose marketplace-GTM capabilities to AI agents would be a defensible "agentic GTM operating system" claim ahead of category coalescence.

What "supporting x402" looks like¶

Three layers of integration, in increasing order of investment:

Layer 1 — Listing on agentic.market (or equivalent aggregator)¶

Effort: ~1-2 weeks. Lowest commitment.

Register Vellocity's most agent-friendly capabilities on agentic.market with a USDC price per call.
Reuse existing MCP endpoints; no code changes to the platform.
Each call deducts USDC from the agent's wallet, settles via the aggregator.
Treat agentic.market as a distribution channel (like AWS Marketplace, but for AI agents).

Layer 2 — Native x402 endpoint on Vellocity¶

Effort: ~3-4 weeks. Adds Vellocity-direct option.

Add HTTP 402 Payment Required support on a /api/v2/agentic/* namespace (mirrors the credits-charged endpoints).
Each capability returns 402 with payment instructions; on settlement, returns the result.
Bypasses the aggregator's cut (typically 10-30%) but requires Vellocity to manage USDC settlement directly.
Compatible with Stripe's machine-payments protocol if Stripe formalizes it as a parallel rail.

Layer 3 — Hybrid subscription + agentic overage¶

Effort: ~6-8 weeks. Highest revenue ceiling.

Subscribers (Starter/Accelerate/Command) get monthly credit allowance; overage routes to x402 micropayments at a small premium.
Avoids "you ran out of credits — please upgrade" friction for high-usage subscribers.
Captures both buyer personas in one billing surface.
Positions Vellocity as the GTM platform that doesn't punish heavy use — a pricing moat against competitors with hard credit caps.

Pricing the per-call rail¶

Anchor on the existing credit table from /3pi-partners — translate each capability's credit cost to a USDC price.

Suggested first-pass conversion: 1 credit ≈ $0.05 USDC (conservative; adjust after first 1,000 calls).

Capability (from /3pi-partners)	Credits	x402 USDC price (suggested)
AI content studio (35+ workflows)	10–30	$0.50–$1.50
Co-sell intelligence (ICP overlap, partner matching)	20	$1.00
AWS CleanRooms account-overlap analysis	20	$1.00
Marketplace listing SEO + AI visibility	25	$1.25
Performance prediction + launch readiness	8–10	$0.40–$0.50
Competitor analysis	12	$0.60
Deal influence tracking	15	$0.75
LinkedIn graph analysis	10	$0.50
Media studio (decks, video, narration)	20	$1.00

Why $0.05/credit: matches typical x402 micropayment ranges (under $2/call discourages friction; under $0.10 disappears under aggregator fees). Subscription tiers stay attractive — at this rate, 500 credits in Starter = $25 of usage at agentic prices, but $299 of subscription revenue. The subscription is the deal; agentic is the spillover.

Open questions¶

Settlement currency. USDC is the obvious default (x402's native rail), but agentic.market and Stripe's machine-payments protocol may diverge on supported stablecoins. Defer until aggregator landscape settles in Q3.
AWS Marketplace + x402 collision. AWS Marketplace has its own metering API. If agents could one day pay through AWS-MP-metered dimensions instead of USDC, that's a different settlement architecture. Keep the API surface stable so we can route either way.
MCP capability surface. Not every capability is agent-friendly (e.g., long-running workflows that require human review). Curate the agentic-exposed subset deliberately — exposing the wrong endpoints creates support load.
Brand voice / quality control. Subscription customers configure brand voice once and reuse it. Agentic callers are stateless. Decide whether agentic-rail calls use a default house voice or a per-call voice override.
Receipt + audit story. Agentic micropayments don't have invoices. Need a settlement-summary endpoint for any partner that wants a paper trail.

Pre-pilot readiness (May 8 → June 1)¶

The AWS MP agentic pilot launches ~June 1. Vellocity has ~3 weeks to be a launch-day participant. Most of these are verification tasks rather than build tasks — Vellocity's existing infrastructure (seller-of-record, MCP server, capability rate card, metering dimensions) is already most of the way there.

Week 1 — Confirm + align¶

Verify SaaS Contract listing posture. Vellocity has a SaaS Subscription product on AWS MP today (the new $299/$1,499/$3,999 monthly + yearly pricing). Confirm the parallel SaaS Contract product (with the 4 metering dimensions) is properly registered or queue a new listing.
Confirm metering dimension alignment with the Usage-Based Pricing Layer rate card. The marketplace_pricing_dimensions table seeded words $0.001/1K, images $0.05/img, audio_minutes $0.01/min, video_minutes $0.50/min. Verify these still match what AWS expects for an agentic-eligible product; pilot may require additional dimension types (e.g., agent_calls or capability_invocations).
Tag MCP capabilities as agent-payable. Add metadata to each capability indicating settlement eligibility — at minimum a boolean agentic_billable flag in the capability registry. Curate the agent-exposed subset deliberately; not every capability is agent-friendly (e.g., long-running workflows that need human review).

Week 2 — Infrastructure¶

Test the AWS MP metering API end-to-end on dev/staging. AWS provides a MeterUsage API that sellers call to report consumption. Implement a thin reporting layer that aggregates capability calls per AWS customer ID and submits via MeterUsage daily. Defer real-time metering to Q3.
Implement an agent-identity → AWS-customer-ID mapping. Agents authenticated to a Vellocity-AWS-customer's MCP endpoint should have their calls counted against that customer's AWS metering. Likely keyed by the AWS account ID extracted from the customer's resolved AWS Marketplace entitlement.
Add a settlement-summary endpoint. Agentic micropayments don't have invoices; partners and customers will want a per-period consumption report. Mirror the AWS MP usage reports format so customers can reconcile.

Week 3 — Go-to-market¶

Private Offer template for agentic buyers. Standard AWS MP private offers assume human buyers and seat-based commitments. Agent buyers may need a usage-cap structure (e.g., "$25K TCV against the metering dimensions, hard cap at $30K to prevent runaway agent loops"). Draft 2-3 templates.
Identify 1-2 beta partners willing to be early agentic buyers. Most likely a 3PI already deep in MCP integration who wants to expose Vellocity capabilities to their own agent workflows.
Customer-facing doc: "How AI agents pay for Vellocity capabilities on AWS Marketplace." Lead with fiat settlement, end-of-month invoice, no API keys for agents — these are the differentiators vs. standalone API key + Stripe.

Day-of (June 1 ± 1 week)¶

Verify pilot listing is live on AWS Marketplace.
Submit first metering report even if zero usage — confirms the integration path before real volume.
Announce participation to existing customer base + 3PI partner pipeline. First-mover positioning has a short half-life; capture it in the announcement window.

Why x402 stays Q3, not earlier¶

Q2 priorities (in order): 1. Rename + reprice (shipped 2026-05-06 → 2026-05-08). 2. AWS MP agentic pilot readiness (this section, May 8 → June 1). 3. First signed 3PI LOI (per aws-3pi-market-analysis.md §4.5 — the load-bearing risk that doesn't go away just because a new rail opens).

x402 is exciting but redundant against AWS MP for AWS-resident agents — and AWS-resident agents are most of the addressable market for Vellocity in 2026. Build the AWS MP rail first; revisit x402 in Q3 once volume justifies the second rail.

None of the top-rung valuation rungs in aws-3pi-market-analysis.md §4 require either rail to be live, but participating in the AWS MP launch as a Day-1 seller is itself a credibility marker that helps the LOI conversation.

Tracking signal — what to watch in Q2¶

agentic.market payment volume (currently $49.6M+ visible on their dashboard — track quarterly delta)
Stripe machine-payments protocol GA timing (currently announced; not GA)
Coinbase x402 SDK adoption in agent frameworks (LangChain, Vercel AI SDK, MCP runtime)
Any 3PI partner signal of agent-as-buyer use cases — if a partner is pitching "our customers want to expose to agents", x402 jumps to Q2