Type to search pages, articles and docs
Product
Collabur handles the recruitment, tracking, fraud screening and payout preparation for your partner and affiliate programme. Your team approves the terms and releases every payout.
The concept
Software to track a partner programme is inexpensive. The recurring cost is the operating labour. Collabur operates that layer as a governed service and keeps commercial control with your team.
Sourcing and qualifying partners, onboarding outreach, tracking referred activity, screening for fraud and computing payout amounts — the repetitive, audit-friendly work of running a programme.
Your owners approve partner terms, commission structures, advertising-disclosure requirements and every payout release. AI colleagues cannot alter commercial terms or move funds.
The service reduces the operational burden of running a programme. It does not reduce your accountability for it.
The modules
Each module carries out a defined part of the operating work and writes to a shared approval-and-audit record.
AI colleagues identify and qualify prospective partners and conduct onboarding outreach, with same-day applicant handling to avoid losing partners. Partner terms and acceptance thresholds are set and approved by a human; borderline applicants are routed for review.
Tracks referred activity against your agreed attribution rules, designed to remain workable as cookie-based methods weaken. Produces the activity record that underpins each payout.
Screens referral patterns for suspicious behaviour — cookie-stuffing, attribution theft and invalid traffic — beyond static rules. Flagged patterns escalate to a human reviewer before they can affect a payout.
Computes payout amounts, validates conversions and reconciles against the activity record, addressing the familiar cluster of manual validation, duplicate payments and missed commissions. No payout executes without explicit human sign-off.
Monitors partner content for advertising-disclosure requirements to support your standing obligation under the FTC's 2023 Endorsement Guides. Accountability stays with your humans; suggested remediation is surfaced for approval.
Records the underlying activity and the approval step for each referral and payout, supporting audit trails and dispute resolution. This governance layer is where human control points are configured and enforced.
How a programme runs
The same operating cycle, from recruitment to payout, with human approval at defined control points.
AI colleagues source and qualify candidates and run onboarding. Applicants within approved thresholds are accepted; others are routed to a human, preserving the same-day cadence practitioners say is needed to avoid losing partners.
Referred activity is tracked against agreed attribution rules and written to the activity record. Anomalies feed screening.
Referral patterns are screened for fraud signals. Anything exceeding configured thresholds or triggering a fraud signal is escalated to a human reviewer before it can affect a payout.
Payout amounts are computed and reconciled against validated conversions, then presented for explicit human sign-off. Funds move only after approval.
Partner disclosures are monitored against configured requirements. Suspected gaps are surfaced with recommended remediation for human approval, supporting the customer's FTC oversight duty.
Agentic architecture
The line between what runs automatically and what requires a person is explicit and enforced.
Sourcing, onboarding, tracking, screening and payout computation are performed by AI colleagues across the programme.
Decisions that exceed a configured threshold or trigger a fraud signal are routed to a human reviewer. No payout executes without explicit sign-off.
Humans hold approval authority over partner terms, commission structures, advertising-disclosure requirements and payout release. AI colleagues cannot change commercial terms or move funds.
Each referral and payout is written to an approval-and-audit ledger recording the underlying activity and the approval step, supporting audit and dispute resolution.
The context
Each referral and payout carries a record of the underlying activity and the approval step, supporting audit trails and dispute resolution. Payouts move only on explicit human sign-off.
Payout handling is regulated and trust-sensitive. Procurement includes trust and compliance diligence under human fiduciary approval.
Revenue is a subscription plus a fee scaled to programme volume, deliberately independent of your partner payouts. Collabur is paid for running a sound programme, not for larger payouts.
Disclosure monitoring supports, but does not assume, your standing obligation under the FTC's 2023 Endorsement Guides. Cross-border tax and KYC on payouts are handled within the reconciliation workflow.
Compliance evidence gathered to date is US-only. UK, EU and GCC regimes require dedicated research before we operate against them, and we state that plainly.
The following describes our intended build sequence. It is subject to what we learn from operating real programmes.
First operated programmes live for a small cohort of design-partner brands. Sourcing, onboarding, tracking and payout preparation run through AI colleagues with human approval at every payout, and the audit ledger captures activity and approvals from day one. Fraud screening runs on an initial signal set, with early evidence gathered on hours displaced and real marginal cost per programme.
Operated across a larger cohort with a maturing fraud and anomaly model informed by cross-programme referral patterns, tighter attribution handling under cookie deprecation, and disclosure monitoring moved from assistive to routine. A referenceable trust and dispute-resolution track record begins to form, and the subscription-plus-volume-fee model is validated in practice.
A repeatable operated service for English-language SMB and mid-market brands, with accrued audit history, adopted by FTLAB portfolio ventures as their partner and affiliate operating capability. Fraud screening and attribution improve with accumulated data, governance controls harden, and we prepare for the MEA and DIFC catchment pending non-US disclosure research.