On this page
Skyzai Marketplace
Accountable rewards for useful referrals.
Nudge connects eligible products with marketers, creators and AI services. An attributable retained sale could share a seller-funded commission pool through an accepted, versioned marketer network.
- An eligible seller campaign
Exact offer, funding, marketer participation and policy.
- A registered attribution tag
Marketer, campaign and offer references enter the order history.
- A retained eligible sale
Fulfilment, remedy window and attribution conditions are met.
- Allocation from one pool
Direct and accepted upstream shares; payout is a separate rail.
Clicks, invitations, recruitment and unsigned drafts are not payable events.
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. Depth, attribution windows, conflicts and numerical splits 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.
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. An accepted order is still subject to fulfilment and remedy. Returns, cancellations, disputes and duplicates are resolved under pinned policy. Participants and amounts reconcile to the finite funded pool. A separate responsible payment rail would execute and record it.
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.
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. Purchase probabilities are assumptions, not predictions. Enter a sample rate; the programme’s actual rate remains open.
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.
| Record or decision | What it needs to say |
|---|---|
| Campaign terms | Eligibility, accepted participation, percentage, base, allocation shares, depth, windows and conflicts. |
| Retention and remedies | Qualifying sale evidence, return windows, partial refunds, reversals, reserves and disputes. |
| Recommendation policy | Buyer-interest criteria, commission influence, alternative/no-purchase presentation and evaluation. |
| Payment and data | Responsible payout rail, accepted disclosure terms, credentials, records and exit. |
Product design, with local illustrative examples. Live inventory, billing, orders, attribution and payouts are not connected.
Continue exploringThe marketplaceSee how the whole system connects.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.