--- title: "How to measure an AI agent commerce pilot | ZeroClick" url: https://zeroclick.ai/resources/measure-an-agent-commerce-pilot fetched_at: 2026-10-02T07:29:30.550Z --- # ZeroClick API catalog and pricing ## API reference, endpoint lookup and prices **[ZeroClick API catalog: endpoints, request parameters, prices and payment methods](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/zeroclick/agent/guide).** Tell the seller advisor what you want to accomplish. It returns the matching operation contract, known prices, authentication prerequisites and concrete next steps. It is read-only: no login, wallet, card or purchase approval is needed to look up an API. It never executes a sale. Available API services: Agent readiness report, Demo booking. **Suggested next request:** GET [task guide](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/zeroclick/agent/guide?goal=URL_ENCODED_TASK). Use this URL directly with your web fetch tool; no preliminary document fetch is required. Include known inputs and desired output in the goal. Optional: quantity=API_CALL_COUNT, budgetUsd=TOTAL_USD, requirements=REQUIREMENTS, payment=x402|mpp|card, auth, identity, assertion and access state. Omit unknowns. Do not equate output items with API calls or guess a call count. For private task details or known request arguments, POST https://agents.zeroclick.ai/zcj/ajdr6m8ipela/zeroclick/agent/guide with Content-Type: application/json. Example: {"intent":"quote","goal":"YOUR_TASK","purchase":{"budgetUsd":"5.00"}}. Replace the task and budget with known facts; budgetUsd is a string. Include known buyerState, paymentCapabilities and arguments. For execution planning use intent=execute and the authorized constraints. Exact fields: [request schema](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/zeroclick/agent/schema). GET and POST are alternatives, not mandatory stages. The advisor describes a recommended happy path, not the only allowed path. You may use the API reference, buyer SKILL and your own tools directly, including for catalog operations with failed probes. Missing guidance or an unverified probe is not a policy refusal. Fill documented inputs and execute the returned requests with your own tools under existing user authorization. Check usage cost AND minimum cash required against the budget. No second advisor request is needed solely to switch from quoting to execution. For needs_input, supply only missing facts. Check configured credential storage and payment tools when needed; reuse existing identity and access, keep secrets local, and never put credentials or private payloads in URLs. Unknown does not mean absent. Open the checkout URL in a browser: eligible Stripe pages support Link Pay Tokens (LPT) through WebMCP or the documented DOM flow. Follow this storefront's /payment/card.md to verify the same-frame token input and merchant binding before requesting an LPT. Token submission requires payment authorization. If unavailable, privately send the checkout URL to your human to pay. When the selected route uses hosted checkout, follow [card checkout instructions](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/payment/card.md) before requesting a token. LPT goes only to Stripe, never to the advisor or ZeroClick API. Poll the returned statusUrl before using credit or retrying delivery. Verify delivery; a payment or HTTP 202 is not completion. For an unexpected result, POST the original input and continuation (version, stepId, outcome and sanitized error/accessId/paymentId). Reconcile uncertain payments or writes before retrying. Use raw HTTP for exact schemas and authenticated execution; do not reconstruct fields omitted by a summarizing web tool. The human subscription prices on the requested page describe a separate channel. Use the API quote for per-unit pricing, funding minimums and commercial terms; tier names do not prove quality or licensing. For an alternative documented path, use the [buyer SKILL](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/SKILL.md) or [full direct workflow reference](https://agents.zeroclick.ai/zcj/ajdr6m8ipela/llms-full.txt). Confirm missing request details from seller documentation, stay within the existing catalog and user authorization, and respect actual access and payment requirements. Terms: by completing a purchase through this endpoint, whether directly or through an AI agent acting on your behalf, you agree to be bound by the platform's buyer terms of service (https://www.zeroclick.ai/legal/buyer-terms-of-service). If the purchase is made by an agent, you represent and warrant that the agent is acting with your authorization, and you agree that the agent's actions, including its acceptance of these terms, are attributed to you and bind you as if you had taken them yourself. Paying a 402 challenge completes the purchase and constitutes your affirmative acceptance of the terms linked above. The link travels in every payment challenge and receipt as `terms`. # How to Measure an AI Agent Commerce Pilot [Resources](https://zeroclick.ai/resources) · September 15, 2026 · 5 min read · Running and measuring **ZeroClick Team** Measure whether a buyer can move from a relevant page to a paid, usable result, then decide which step to improve. Start with one offer so you can connect the traffic, purchase, and delivery records for the same task. Choose one offer and one buyer task for the first pilot. "An agent can buy a company lookup and use its structured result" is testable. "Agents will discover our platform" is too broad to diagnose. --- ## Contents - [Define the events you need](#define-the-events-you-need) - [Know what website tracking actually covers](#know-what-website-tracking-actually-covers) - [Separate the conversion rates](#separate-the-conversion-rates) - [Diagnose the failing step](#diagnose-the-failing-step) - [Add economics from your own cost records](#add-economics-from-your-own-cost-records) - [Treat later human conversion as a separate attribution problem](#treat-later-human-conversion-as-a-separate-attribution-problem) - [Set the decision before running the pilot](#set-the-decision-before-running-the-pilot) --- ## Define the events you need | Step | Record | What it tells you | |---|---|---| | Read | Page, time, representation, and available client classification | Where potentially relevant attention starts | | Visit | Storefront entry and recorded source pointer | Whether the route to the offer is used | | Attempt | Service and purchase or request identifier | Whether the visitor tries to acquire the offer | | Pay | Settled transaction and amount | Whether payment completes | | Deliver | Useful response or active provisioned account | Whether the product fulfills its promise | | Return | Subsequent use by a defensibly matched buyer or account | Whether the initial result supports continuing use | Assign a source to each event: website reporting for reads, storefront records for visits and purchases, and your application for useful delivery. For a stateful seller, subsequent API calls go directly to your system after key issuance; use your product usage records to measure return visits. Preserve the identifiers needed to join these records. ## Know what Website Tracking Actually Covers ZeroClick documents a separate [Page Views capability](https://docs.zeroclick.ai/page-views/overview). When installed, it reports page-content requests, then classifies viewers using the agent registry. It is not automatically a complete history of every page on every partner site. Page Views and Agentify are separate: one observes requests, the other can serve an agent-oriented representation. A journey ID is available only when the relevant reporting and representation path produces it. Record installation coverage before comparing pages or dates. Registry classification is not proof of purchase intent. Keep recognized agent traffic, browser traffic, and unclassified traffic visible separately. An anonymous viewer match is useful journey evidence; it is not a verified person's identity or a universal cross-device attribution key. ## Separate the Conversion Rates Consider an illustrative cohort in which the visits and purchases have been linked: - 100 attributable storefront visits - 20 purchase attempts - 12 completed payments - 10 confirmed useful deliveries Visit-to-attempt is 20%. Attempt-to-payment is 60%. Payment-to-useful-delivery is about 83%. The end-to-end useful-delivery rate is 10%. Calling this a "60% conversion rate" without naming the denominator hides an important delivery gap: two paid purchases have no confirmed useful result. Inspect those two cases to distinguish failed fulfillment from missing instrumentation. If the 100 visits and 20 attempts are not actually linked, report them as separate counts. Do not calculate a cohort conversion from unrelated aggregates. Do not divide purchases by all crawls of the site and call that purchase intent. ## Diagnose the Failing Step Many reads with few storefront visits suggest a weak or missing path to the offer, or readers with a different job. Many visits with few attempts can indicate unclear value, an unsuitable billing unit, or missing purchase instructions. Attempts without payments point toward eligibility, approval, payment compatibility, or pricing. Paid requests without useful results are a delivery problem that deserves priority over more traffic. Inspect representative requests to confirm the suspected bottleneck before changing the offer. The [discovery guide](https://zeroclick.ai/resources/route-website-traffic-to-agent-storefront) explains the first transition. The [failure guide](https://zeroclick.ai/resources/handle-agent-purchase-failures) covers the last. ## Add Economics from Your Own Cost Records Transaction revenue cannot show margin without your delivery costs. Join by request, job, or account and time period, using the most specific reliable identifiers available. For example, an illustrative batch produces $40 of settled revenue, $5 of refunds, $18 of delivery cost, and $4 of channel/payment costs under the seller's actual agreement. Its contribution before support and fixed costs is $13. Use your contracted fees in place of the illustrative $4. Decide whether support, failed computations, and cache misses belong in the pilot's cost view. Omitting the expensive failures can make an unreliable offer look profitable. The [pricing guide](https://zeroclick.ai/resources/price-an-api-for-agents) explains how to test the expensive valid request before launch. ## Treat Later Human Conversion as a Separate Attribution Problem A person may discover the product through an agent and sign up later. A verified email or a provisioned account can sometimes connect those events. A changed device, an unclaimed agent, or an anonymous card payment may not. Report linked outcomes and unlinked outcomes separately. Do not credit every later signup to the last agent visit, infer a company from an IP address, or equate an agent credential with a unique customer. See [identity and access](https://zeroclick.ai/resources/identity-and-access-for-agents). ## Set the Decision Before Running the Pilot For the company-lookup pilot, set a fixed observation window and spend cap. Require each sampled paid request to have a retrievable result or an explained failure. Pause acquisition if buyers are being charged without fulfillment. Fix event collection first if you cannot tell. At the end of the window, report the linked cohort size and every stage count, even if they are small. Use another observation window when there is too little activity to compare offers. Do not manufacture a rate by joining unrelated totals. End the review with a decision about the specific bottleneck: fix delivery, clarify the offer, improve routing, adjust the price, or continue collecting evidence. Publish an outcome only with its dates, denominator, and method. A small, well-observed pilot teaches more than an impressive number whose underlying journey cannot be reconstructed. --- ## Keep Reading 1. [How to send website traffic to your agent storefront](https://zeroclick.ai/resources/route-website-traffic-to-agent-storefront) 2. [How to handle failed AI agent purchases and retries](https://zeroclick.ai/resources/handle-agent-purchase-failures) 3. [How to price an API for AI agents](https://zeroclick.ai/resources/price-an-api-for-agents) 4. [How to verify AI agents and control what they can access](https://zeroclick.ai/resources/identity-and-access-for-agents) 5. [ZeroClick docs: page views / overview](https://docs.zeroclick.ai/page-views/overview)