Sponsorship and referral disclosure policy.

Policy version: 2026-03-12-disclosure-v1

Published: March 12, 2026

Ranking and confidence stay benchmark-only. Referral payouts, sponsorship inventory, and sales goals cannot change model order, runtime order, upgrade guidance, or any confidence wording.

Monetized CTAs stay in a separate labeled lane. Any referral, sponsor, or paid CTA appears only after the recommendation rationale or on a dedicated guide page with plain labeling.

Offer measurement stays separate from benchmark scoring. Offer views, CTA clicks, and referral clicks are tagged with experiment IDs and funnel stage so they can be audited without touching ranking code.

Legal, privacy, and support copy repeat the same rule. The disclosure policy, privacy page, and FAQ all say the same thing: monetized offers never rewrite the benchmark answer.

Ranking stays sealed off from money. Benchmark score, model order, runtime order, and confidence bands stay derived from benchmark logic only. No sponsor, payout rate, or partner inventory can change that answer.

Revenue lives in separate, labeled surfaces. Any affiliate, sponsor, or premium offer must appear after the recommendation or on a guide page with plain labeling so users can distinguish advice from monetization.

Every experiment must be auditable. Eligibility rules, attribution tags, and success metrics are documented up front so any future payout source can be checked against the public recommendation experience.

Trust metrics outrank revenue metrics. If useful feedback, matched-rate confidence, or share behavior deteriorate after an experiment launch, the experiment pauses before any payout optimization continues.

No paid boosts inside recommendation rows. Sponsored runtimes or paid placements can never appear inside the ranked model list, runtime list, or confidence explanation.

Eligibility is chosen before payout rate. An offer can show only when the benchmark outcome makes it relevant first, such as upgrade-heavy machines or low-confidence local outcomes.

Every offer must be labeled and logged. Offer impressions and clicks carry experiment identifiers and funnel stage so revenue can be compared against benchmark completes and shared runs later.

Trust degradation pauses the test. If useful feedback falls by more than two points, matched feedback worsens materially, or share generation drops unexpectedly, the experiment is rolled back.

Upgrade and buying guides. If a benchmark-adjacent buying guide includes affiliate or sponsor inventory, the material connection must appear in the same module before the outbound CTA.

Referral handoffs. Any referral or paid-checkout handoff must restate the material connection next to the CTA, not only in a footer, FAQ, or later checkout step.

Paid follow-up products. Paid WhatsMy.AI follow-ups must say the next step is a paid product before the button and again on the checkout handoff.

Partner and testimonial proof. Any testimonial, partner quote, or customer proof near a benchmark-adjacent recommendation must repeat the disclosure in the same card or module.

Audit evidence. Monetization impressions and CTA clicks must log disclosurePresent=true together with experiment ID, surface, and variant so audits can prove the disclosure was present.

Current runtime status. disabled. The survey is gated by NEXT_PUBLIC_PRICING_SURVEY_RESULT_HOLDOUT_ENABLED, scheduled by NEXT_PUBLIC_PRICING_SURVEY_RESULT_HOLDOUT_START_AT and NEXT_PUBLIC_PRICING_SURVEY_RESULT_HOLDOUT_END_AT, and capped by NEXT_PUBLIC_PRICING_SURVEY_RESULT_HOLDOUT_PERCENT.

Exposure rule. The prompt can only appear as an optional result-page card after the benchmark answer is complete. It stays separate from benchmark usefulness and matched-rate feedback.

Premium buying brief price check. Allowed question ID: premium_brief_price_check. Prompt: If WhatsMy.AI offered a premium buying brief for this result at {price} as a one-time purchase, would you consider it?. Allowed responses: Yes, Maybe, No. Allowed prices: $9, $19, $39, $79.

Storage and deletion. Responses are stored only as anonymous analytics events in `analytics_events`, read out on the internal growth dashboard, and pruned after 90days by the existing telemetry retention job.