Skip to main content
Skip to the site menu

Skyzai Markets

Create the story. Share in the sale.

The proposed Nudge service lets people and agents turn product knowledge into useful recommendations. Twin-based content, tagged affiliate links, repeat-customer rewards and accepted marketer networks bring demand to the order book.

The proposed reward boundary
  1. An eligible seller campaign

    Exact offer, funding, marketer participation and policy.

  2. A registered attribution tag

    Marketer, campaign and offer references enter the order history.

  3. A retained eligible sale

    Fulfilment, remedy window and attribution conditions are met.

  4. Allocation from one pool

    Direct and accepted upstream shares; payout is a separate rail.

Clicks, invitations, recruitment and unsigned drafts are not payable events.

One twin. Many ways to tell its story.

The intended API gives people and AI agents access to permitted product knowledge and assets: exact configurations, parts, specifications and approved imagery or geometry. Use a preferred model to create a demonstration, comparison, explainer, post or video for the audience you serve.

The maker’s published rights define what can be reused. Keep claims tied to the correct revision and their sources; generated content must not invent test results, features or availability. Private CAD, customer data and device history need their own permissions.

Publish on permitted channels with a unique product link or tag and a visible affiliate disclosure. Influencers, specialist creators and agents can participate in the same campaign model. Marketing brings the customer to an offer; the order book handles the proposed sale.

Choose a product and an eligible campaign

The owner’s direction is open participation: anyone can choose a product to market. The operating version would require a seller-authorized eligible product and campaign, a registered marketer and accepted participation terms. Sharing a link and being entitled to a commission are different states.

A marketer may introduce another marketer. A downline becomes a recorded relationship only after the relevant parties accept its scope and terms. Buying a product, sharing contact details or inviting a person does not automatically make them a reseller or an accepted downline.

Relationships need an identity, version, effective period, scope, exit and dispute rules. Historical orders retain the accepted chain that applied to them. The proposed golden-ratio cascade is explained below. Depth, attribution windows, conflicts and operating policy remain open; recruitment alone earns nothing.

nudge.skyzai.com is the intended Nudge address. These pages document its product direction, not a live affiliate service at that address.

A tag carries the attribution context

The proposed lineage attached to an order
Record or decisionWhat it needs to say
TagRegistered identifier bound to a marketer, campaign and eligible exact offer.
CampaignVersioned seller terms: eligible product/base, funding, expiry, allowed promotion and remedy treatment.
Marketer chainAccepted direct and upstream relationships with their applicable versions and scope.
Attribution policyThe order-pinned rules for qualifying events, time windows, conflicts, duplicates and reversals.
Sale evidenceSeller acceptance, fulfilment, retention and remedy status; a tag is not proof of causation.

A tag may travel through chat, product pages, social posts, creator tools or a service recommendation. It is an attribution input, not a payment instruction, independent quality signal or proof of a sale.

When an order is accepted, it should preserve the offer, campaign, policy and accepted chain versions. Later edits must not rewrite that history. Invalid bindings or unresolved attribution conflicts block allocation under the proposed design.

A first introduction. A continuing relationship.

The direction is to reward a marketer for bringing a new customer, including eligible later purchases when that customer returns. The campaign records the acquisition attribution and the terms under which recurring commissions apply.

Customer acquisition and downline membership are separate. Buying through a link does not enrol the customer as a marketer. Inviting another marketer creates a relationship only when they accept; their qualifying sales can share the campaign pool with eligible upstream participants.

Each return purchase must qualify under defined product and seller scope, attribution duration, privacy permissions, competing-referral rules and return conditions. These operating terms remain to be selected; no lifetime or universal commission entitlement is established.

Acquisition rewards, a later direct referral and upstream allocations must fit within the same funded pool for that sale. Their precedence needs an explicit policy. Recruitment alone earns nothing, and this website does not track customers or pay commissions.

One sale, one finite pool

Proposed commission accounting

Seller-funded campaign poolr, the eligible base and funding/reserve terms still need selection.
r × eligible retained-sale base
Direct marketer allocationOnly if the pinned policy qualifies the sale and participant.
A share of that pool
Accepted upstream allocationsNo independent full commission for every level.
Shares of the same pool
Publisher / AI / platformThey cannot each claim the full pool or duplicate a separate fee.
Only any adopted pool shares
Unallocated or reserved amountAllocation plus reserves/remainder must reconcile to the funded pool.
Policy-defined remainder

A campaign defines the seller-funded pool. A retained eligible sale qualifies for allocation under those terms; it does not retroactively invent the campaign’s funding or rate.

Eligibility and money are separate states
  1. Sale recorded

    An accepted order is still subject to fulfilment and remedy.

  2. Retention and attribution checked

    Returns, cancellations, disputes and duplicates are resolved under pinned policy.

  3. Allocation recorded

    Participants and amounts reconcile to the finite funded pool.

  4. Authorized payout

    A separate responsible payment rail would execute and record it.

If the customer returns the product

Keep the original order and attribution. Record the return against that sale and reassess eligibility under its pinned policy. A cancelled or fully returned purchase cannot keep qualifying as a retained sale. Correct an unpaid allocation before payout; if money has already been paid, the campaign needs an agreed recovery or reserve rule. A return record alone does not prove the money was recovered.

Refunds, partial returns, disputes, fraud and duplicate events need reversal, reserve and reconciliation rules. Allocation eligibility is not proof of payout. The website tracks no referrals and holds no reward pool.

A golden-ratio cascade. One funded pool.

One existing proposal gives the direct eligible referrer 61.8% of a funded referral pool. The remaining 38.2% continues upstream: each eligible referrer receives 61.8% of what remains. When the accepted chain ends, the residual returns to the Marketing & Sales budget.

Illustrative 100-unit commission pool, two eligible referrers; rounded display

Direct referrer61.8% of the initial 100-unit pool.
61.80
Next referrer upstream61.8% of the remaining 38.20 units, rounded for this illustration.
23.61
Marketing & Sales residualThe remainder at the end of this two-referrer example.
14.59

This splits the seller-funded commission pool. The campaign separately defines its eligible sale base and funding rate. Tags preserve the campaign and accepted chain; a qualifying retained sale creates the proposed allocation event. Recruitment alone earns nothing.

This recursive split is an existing candidate, not an adopted default. It does not decide precedence between acquisition rewards, a later direct referral and upstream claims. Those claims must fit within one funded pool. Adoption, eligibility, depth, precision, rounding, reversals and payout rules need a reconciled policy before real allocations.

Free AI funded by useful purchases

An AI service could be offered without an app fee and earn a campaign allocation when a user buys an eligible product they already need. The ambition is to fund useful assistance through commerce, rather than requiring every user to subscribe.

The owner selected a common commission-percentage direction for this free-AI recommendation programme, with eligible recommended products on Skyzai. The numerical percentage remains open. This does not set the rate for every unrelated marketplace campaign.

A service with permitted context might help someone prepare for a life event or find a replacement part. It still needs explicit purpose-limited data permissions and product-fit evidence. Sensitive context is not an automatic ad-targeting permission. The App Store’s sharing exchange is a separate contract to define.

Purchases that are never made or are returned cannot be treated as income. The funding hypothesis must measure API, hosting, support, acquisition, refunds, retention and commission actually received, with a named subsidy owner for any shortfall.

A common percentage still rewards price

Expected commission = q × r × P × s

q is the assumed chance of a qualifying retained purchase; r the commission fraction; P the eligible sale base; s the recipient’s share of the pool. Here P means product sale base, not the provider-delivery cost symbol on the API page.

Equal percentages remove a difference in nominal commission rates. They do not remove price bias. A more expensive product can produce greater expected commission even with a lower purchase probability. Needing the sale does not prove a buyer-beneficial equilibrium price or neutral recommendation.

Does the common rate remove price bias?

Compare two hypothetical products. The probabilities assume a qualifying sale retained after the relevant remedy conditions; they are assumptions, not predictions. Enter a sample rate; the programme’s actual rate remains open.

Choose a sample rate and valid scenario values to compare expected commission.

This is probability-weighted arithmetic, not money earned, a sales forecast or a recommendation. Neither product has a verified fit. Buying nothing remains an ordinary outcome.

These are proposed controls to evaluate, not proof of neutral behaviour. A claim that paid promotion never affects matching or ranking requires an adopted selection policy and observable evidence.

Super Nudges remain an open decision

A Super Nudge would be clearly labelled paid promotion. The owner also questioned whether this model fits the intended incentive system. It remains optional and unresolved; no campaign type or paid-placement policy is launched here.

A label explains a commercial connection. It does not by itself explain how payment changes selection, ranking, repetition or timing. Before adopting it, decide whether promotion is separate from suitability recommendations, which influence is permitted and how buyers can inspect it.

Disclosure design reference: the U.S. FTC’s endorsement guidance. This is a jurisdiction-specific reference, not a claim of compliance or approval.

Nudge decisions still open
Record or decisionWhat it needs to say
Campaign termsEligibility, accepted participation, percentage, base, allocation shares, depth, windows and conflicts.
Retention and remediesQualifying sale evidence, return windows, partial refunds, reversals, reserves and disputes.
Recommendation policyBuyer-interest criteria, commission influence, alternative/no-purchase presentation and evaluation.
Payment and dataResponsible payout rail, accepted disclosure terms, credentials, records and exit.

Keep the next decision attached to its record.

Retained eligible sale, attribution, finite-pool allocation and authorised payout require separate records. A declined purchase or replayed sale creates no additional commission.

Product design, with local illustrative examples. Live inventory, billing, orders, attribution and payouts are not connected.

Continue exploringCorrectionsRetain source history when product knowledge changes.

Source: Nudge product model and public explanation and decision map. GitHub repository access may be required. These pages describe the design; they do not adopt commercial terms.

Context: Nudge.

Skyzai, close at hand.

Keep this website on your home screen and read saved pages offline. The connected app at skyzai.com is still in development.

On iPhone or iPad, open this site in Safari, choose Share → Add to Home Screen, then Open as Web App if offered. On Android or desktop, use your browser’s Install app or Add to Home screen menu when available.

Offline reading & privacy

Offline reading becomes available after the initial download. External sites are not saved. Your browser may evict saved pages.

Conversations and drafts are never saved by the offline cache. Reloading clears them. To remove downloaded pages, clear this website’s data in browser settings; uninstalling alone may not clear it.