Skip to content

RFC-0008: Monetisation + pricing model

Decision gate. This RFC ratifies-with-modifications the original aggregator design spec §4.1 shape and codifies the decision-flip conditions that would force a re-read. Once MASTER picks, the choice lands as one ADR; the implementation work is already routed (S2S postback in #17, B2B surfaces via #16, display rules via #28).

Summary

Primary: CPS (cost-per-sale, 1.5% of basket) for retailers willing to integrate S2S postback or accept coupon-code reconciliation. Fallback: CPC ($0.10-0.15 per qualified outbound click) for retailers who can't or won't track conversions but want the traffic. Optional secondary: paid featured-placement at $50-200/mo per category, clearly labelled Sponsored. Free organic listing for every Lebanese retailer regardless of cooperation tier — catalog breadth is the moat. Buyer-side stays free in M1/M2; subscription tiers and B2B/SMB premium tooling are deferred but routed through #14 and #16. Currency is USD only per ADR-0004. The decision-flip conditions and a per-stream unit-economics table are in §Proposal.

Motivation

Three concrete pressures push the decision now:

  1. The original spec (§4.1) names a model but doesn't ratify it. "1.5% CPS + $0.15 CPC + $50-200 featured" was a Stream 4 brainstorm output, not a decided model. M2 work (vendor backoffice surface in #16; page-design surfaces for retailer rows in #28; subscription consideration in #14 alerts) starts soon — they need a settled commercial frame to design against.
  2. The retailer audit (#30) confirmed Lebanese retailer affiliate willingness is Unknown for every retailer profiled. Lebanese tech retail is bilateral and relationship-led; commercial deals are not platform-listed. We cannot assume the spec's CPS path is straightforwardly available — the RFC has to pick a model that survives "no retailer signs a postback for 6 months."
  3. Persona research (#36) sharpened the buyer-side constraint. Primary personas (Karim, First-time builder, Casual customer) all sit in the $600-1,800 price band. Per the §41 ticket constraint, monetisation that taxes the buyer is a non-starter. This rules out paid-tier-for-buyers in M1/M2 but leaves the retailer side wide open. Personas §6.3 also flagged Office IT and Diaspora as high-LTV outliers — different sales motions worth surfacing but not anchoring on.

The genre-level evidence (competitive-landscape §3.5) already converges: PCPartPicker runs ~0.5-1% effective Amazon Associates; Geizhals runs pure CPC; Skroutz ran pure-aggregator-CPC for 11 years before pivoting to marketplace commission; Idealo runs CPC + an own-checkout integration. Pricena's exact rates are not public. No global aggregator anchors on featured-placement alone, and none anchors on a buyer subscription. The viable shapes are CPS, CPC, hybrid, or marketplace-commission-with-checkout; everything else is a layer.

Proposal

The recommendation in one table

Stream Role Default rate Buyer pays? Retailer cooperation M1/M2 in scope?
CPS (rev-share) Primary 1.5% of basket on confirmed conversion No High (S2S postback OR monthly CSV reconciliation) Yes — ship as primary path; live behind a feature flag until the first retailer signs
CPC Fallback $0.10-0.15 per qualified outbound click No None — we set the rate, retailer pays or we drop their Sponsored-eligible boost Yes — ship as the default for retailers without postback capability; cap per-session per-retailer to prevent click-fraud farming
Featured placement Secondary, opt-in $50-200/month per category top slot, clearly Sponsored No Low — pay-to-play directly Yes — surface designed in #28; first sale possible from M5+ onwards
Free organic listing Universal floor $0 No None Yes — every Lebanese retailer with a scrapeable catalog gets indexed free, with Source: <domain>, updated <timestamp> attribution and a 48-hour takedown SLA
B2B / SMB tier Deferred TBD via #16 Possibly (the SMB IT buyer, not the consumer) Different surface (admin-tools / quote-export) No — flagged for #16 admin tools; RFC scopes the opportunity, doesn't price it
Aggregate market data Deferred TBD post-launch No (sold to retailers / suppliers) None — derived from our scrape data No — flagged as post-launch revenue stream; no data-product work in M1/M2
Buyer subscription Rejected for M1/M2 n/a Would tax buyer n/a No — explicitly out of scope per ticket constraint; price-history + alerts (#14) stay free in M1/M2

Comparison matrix on the four ticket-named axes

Axis CPS 1.5% CPC $0.10-0.15 Hybrid (CPS where signed, CPC default) Pure featured-placement
Beirut friction High — needs retailer postback or signed reconciliation cadence; bilateral-deal reality (#30 §7) means Wave-1 outreach is 60-90 days per signed retailer (per spec §6.5) Low — we set the rate, retailer pays or we drop the boost; works without retailer cooperation; fits the "every retailer scraped free, commercial layer on top" reality Medium — we run two reconciliation cadences in parallel; cost is operational, not negotiation Medium-low — the retailer either pays for the slot or doesn't; no reconciliation; but pitch is "pay $X for visibility" which is a harder sell than "pay $X per conversion" in a price-sensitive market
Scaling Strong — revenue scales with GMV; per spec §4.3 model, 250-500 conversions × $30 average commission ≈ $8-15k/mo by Month 11 Bounded — caps at outbound-click volume × per-click rate; harder to scale past low single-thousands per month at Lebanese traffic levels Strong on the CPS retailers, bounded on the CPC tail. The CPS cohort is the growth engine Bounded by the count of category top-slots × monthly rate × % of slots filled; spec models $50-500/month from this stream alone
Retailer cooperation requirements High — postback installation OR signed monthly reconciliation; coupon-code attribution as the third-tier fallback per spec §6.1 None — outbound-click-counted-by-us, retailer either accepts or we drop the listing's commercial layer Mixed — high on the CPS subset, none on the CPC tail Low — the retailer pays for visibility; no integration
Unit economics (range, estimates) \(5-50 per signed retailer per month at v1 conversion volumes; **\)300-2,000/month per Office IT outlier per persona §6.3**; the long tail is small but stacks $20-200/month per active retailer at typical Lebanese outbound-click volumes; ceiling visible Combines both; long-run dominated by CPS as Wave-1 retailers sign; CPC carries the early months $50-200/mo per filled slot × ~6-10 categories × % filled ≈ $200-1,200/month at maturity (estimate)

Sources for genre-level peer rates are in reference/monetisation.md. Lebanese-specific numbers above are estimates keyed to the spec's §4.3 revenue model; nothing in this section is a forecast, and they are calibrated against the persona §6.3 LTV table.

Why CPS as primary, CPC as fallback, hybrid in practice

The recommendation is structurally a hybrid, but the primary-vs-fallback labelling matters because it shapes the M2 vendor-onboarding ladder (spec §6.1) and the #28 page design "View on retailer" CTA logic:

  1. CPS rewards the retailer's actual revenue, not their patience for our traffic. Lebanese retailers respond to "pay only when we deliver" framing better than "pay for clicks I can't audit" — coffee-three-times-before-signing culture (spec §6.5) tolerates a high-friction ask if the asymmetry is favourable to the retailer.
  2. CPC works on retailers who can't or won't audit click-to-conversion. Many Lebanese retailers run on WhatsApp + cash-on-delivery; their order-confirmation system isn't a postback-firing event, it's an Instagram DM and a delivery moto. CPC is the answer for them. We should expect a meaningful share of long-tail retailers to never sign a CPS deal.
  3. Pricing CPS at 1.5% holds against verified peer rates. Amazon.ae pays affiliates 2.00% on Electronics / PC store / Wireless [6] — the cleanest 2026 MENA-region electronics-category comparable. Pricena's only public rate disclosure (2014 Wamda founder interview) was 2-10% CPA depending on category [5]. Skroutz's marketplace commission is materially higher — public sources put the band at 7-15% per category + €469 setup fee as of 2024-2025 (€248 setup in 2022; pricing has escalated) [1] — but justifies it with aggregated-checkout value-add we won't have until M2/M3 (spec §3.1). PCPartPicker's effective Amazon Associates rate is widely estimated at 0.5-1% but has no public primary source [3] — flag as (estimate). 1.5% sits below Amazon.ae's electronics affiliate payout (2.00%), inside the lower half of Pricena's 2014 disclosed band (2-10%), and well below Skroutz's marketplace tier we can't justify until #15 multi-merchant cart aggregator ships. Persona research (§6.3) explicitly endorses this rate as right-shaped across all six personas. The strategic framing is "a Lebanese retailer paying 961tech 1.5% pays less than Amazon's regional ecosystem cost to acquire the same shopper in the UAE."
  4. Pricing CPC at $0.10-0.15 is the conservative end of the industry pattern. Idealo's published per-country CPC is DE €0.51 standard, UK £0.35, AT €0.36, FR €0.40, IT €0.25, ES €0.35 with a +€0.02/year automatic escalator [2] — a public, primary-source rate card. Geizhals charges €0.38 per click with logo integration / €0.41 without [4]. Both are Western European tech-CPC anchors. Lebanese CPM-equivalent ad rates are universally below Western Europe; setting CPC at $0.10-0.15 (~€0.09-0.14) keeps our pitch credible to retailers comparing 961tech outbound to Meta / Google PMax cost-per-click on the same SKU and stays well below the European floor. Correction to original spec: the spec claimed Idealo rates are "not public" and "~£0.24/click"; both are wrong (rates are public; UK rate has inflated to £0.35 since the figure the spec inherited).
  5. Decision-flip conditions (the things that would change the primary):
  6. Flip CPS → CPC default: if Wave-1 retailer outreach (Tech Titan, PC Station LB, PCBuildingLeb, Sbeity, Macrotronics per spec §6.2) lands zero signed CPS deals by Month 9 (end of spec's first soft-launch quarter), demote CPS to "available on request" and make CPC the public default rate-card.
  7. Flip toward Skroutz-style commission: if the multi-merchant cart aggregator (#15) ships and shows >5% conversion rate from our cart to a completed Lebanese-side checkout, the value-add starts justifying a higher take rate. Revisit at the M3 boundary.
  8. Add buyer subscription: if #14 price-drop alerts telemetry shows a power-user cohort (>20 alerts/month per user) at >5% of total active accounts, revisit a freemium-tier paid alerts/saved-build-history option. Default still free.

Pricing model for retailers (cost side from THEIR view)

Adopt the spec's 4-stage onboarding ladder (§6.1) verbatim. Restated for the cost side:

Stage What the retailer pays What they get
1. Scraped + Attributed (no contact) $0 Free organic listing with Source: <domain>, updated <timestamp>; outbound-click traffic free; 48h takedown on request
2. Notified (cold-emailed / visited) $0 Same as Stage 1 + opt-in to be contacted with anonymized traffic data
3. Light Integration (verbal / email agreement) 1.5% CPS on reconciled sales "Verified Partner" badge; coupon-code or monthly CSV reconciliation; analytics dashboard preview
4. Full Integration (signed terms, postback) 1.5% CPS on instant-attributed sales "Premium Partner" badge; analytics dashboard; first-pick on featured-placement slots; published affiliate code-of-conduct page commits us to non-discriminatory ranking

Pure-pay model is rejected. "Only paid retailers indexed" drops the catalog from ~16 audited retailers to maybe 3 — the catalog dies, and the personas' "no multi-retailer price comparison" pain (§5.6 universal pain) goes unrelieved. Catalog breadth is the moat in a market where Pricena deliberately skips (competitive §4.1).

Tiered "free organic + paid premium" is what we ship. Same shape as Geizhals' organic-CPC + non-bidding ranking (competitive §2.2) plus a Skroutz-style premium tier for retailers who want analytics dashboards and feature-eligible placement.

Cross-cutting

Currency

USD only. ADR-0004 ratified English-only UI; the same logic applies to invoicing — Lebanese retailers price in USD (retailer audit §2-§3 — every audited retailer displays USD), banking-crisis lollar/fresh-USD bifurcation makes LBP invoicing actively bad, and our settlement rails (Whish corporate $30k cap, USDT TRC20, OMT — per spec §7.1) are all USD-denominated. LBP is a non-starter. Currency-display per competitive §5.3 is a separate UI concern and stays buyer-facing.

VAT and accounting (defer detail to #40)

  • Below the LBP 5B (~$56k) VAT threshold in Year 1, no VAT registration required (spec §7.3). RFC assumes Year-1 revenue stays under threshold; if revenue overshoots, #40 handles the upgrade.
  • Income classification: affiliate / commission income added as a line on MASTER's existing R10 personal income tax declaration in Stage v0; SARL incorporation deferred until $1.5k+/mo MRR per spec §7.1.
  • Retailer-side withholding: Lebanese retailer-paid commissions are accounted at the retailer's end; we issue invoices, retailer pays Net 7. We do not withhold for them.

Settlement rails

Rail Stage Cap / cost
Whish Money personal v0 (informal) Personal wallet limits
Whish Money corporate v1 (post-SARL or once $1k+/mo) $30k/month per spec §7.1
USDT TRC20 All stages, fallback No cap, ~$1 fee per transfer
OMT All stages, in-country fallback Per-transfer fee, lower cap
Bank transfer v1+ (SARL business account) Standard banking rails

Retailers pick their preferred rail at onboarding (spec §6.4).

Trust patterns

  • Sponsored content always labelled. Featured placements display Sponsored badge + visually distinct treatment (competitive §3.4 #3) per the Idealo / Google self-preferencing precedent: Landgericht Berlin II, Case No. 16 O 195/19 Kart(2), judgment 14 Nov 2025 — Idealo €374.1M damages + €91.1M interest = €465M; Producto GmbH (operator of Testberichte.de) €103.7M; combined ~€572M / ~$665.6M total awarded across plaintiffs. Both parties announced appeal [2].
  • Public affiliate code-of-conduct page modelled on PCPartPicker's /code-of-conduct/industry-affiliate/ — commits 961tech to non-discriminatory ranking and per-source price-time attribution. Strong differentiator (competitive §3.6).
  • Per-listing attribution with Source: <domain>, updated <timestamp> is universal regardless of retailer cooperation tier. Free Stage-1 retailers are visibly attributed; premium Stage-4 partners get the badge but the same attribution discipline.

Trade-offs

Trade-off Cost Mitigation
Running CPS + CPC reconciliation in parallel doubles operational surface area Two billing flows, two dispute-resolution paths, two retailer-onboarding scripts Coupon-code attribution as a third third-tier fallback per spec §5.3 © keeps the retailer-side simple even when our side runs both flows; vendor backoffice UI (#16) hides the duality from retailers
Featured placement risks trust erosion if not labelled cleanly Idealo lawsuit precedent shows the legal exposure when "Sponsored" lookalike treatment misleads users Sponsored label + visual differentiation + organic ranking that ignores featured-tier per competitive §3.4; public code-of-conduct page commits us; Lebanese norms are tolerant of clearly-labelled paid placement per ticket framing
B2B/SMB tier is a different sales motion than B2C aggregator Office IT buyers go through retailer reps, not aggregator UIs (persona §4.1) — selling to suppliers serving SMB IT is a separate channel Defer to #16; RFC flags the opportunity ($300-2,000/mo per Office-IT-shaped account per persona §6.3) without committing to a sales motion
CPS friction in a bilateral-dealing market means Wave-1 outreach is 60-90 days per retailer (spec §6.5) First $5k+ month not realistic before Month 11-12 per spec §4.3 model Decision-flip condition #1 above lets us pivot to CPC default at Month 9 if the CPS path stalls; not a one-way door
No buyer subscription means no recurring buyer-side revenue in M1/M2 If retailer-side revenue stalls, no offsetting buyer revenue Persona constraint is hard — primary cohort is price-sensitive ($600-1,800 builds per persona §3); buyer-side monetisation is a v3+ revisit, not a M2 lever
Anchoring on USD invoicing excludes the small set of LBP-only retailers None observed in retailer audit §2-§3 — every audited retailer prices in USD, so the cost is theoretical Revisit if a Tier-2 LBP-only retailer surfaces in #21 LLM extraction onboarding

Alternatives

Alternative A: Pure CPC default, no CPS at all

Run only CPC; never integrate retailer postbacks. Geizhals' shape (competitive §2.2).

Why rejected. Caps the revenue model at outbound-click volume × per-click rate. At Lebanese outbound traffic levels, this ceilings the project at ~$3-5k/mo per spec §4.3's implied CPC tail. Closes off the path to MASTER's $5-10k/mo Phase 2 target (spec §4.6). Worse: pure-CPC pitches retailers on "pay per click" which is the harder sell to a Lebanese retailer skeptical of audit transparency. Correction to original spec: the spec described Geizhals as "pure CPC, no bidding"; primary-source check shows Geizhals actually runs a dual model — Pay-by-click or Pay-by-order — at the merchant's choice [4]. Geizhals can run that dual model because they're a 27-year incumbent with Heise (German tech publisher) ownership since June 2021 and brand trust; we cannot.

Alternative B: Pure CPS, no CPC

Don't index any retailer who can't sign a CPS deal. Drop the long tail.

Why rejected. Catalog dies, same as the pure-pay model rejected above. The personas' "no multi-retailer price comparison" pain (§5.6) requires breadth; locking out long-tail retailers undermines the core value prop. Also forces Wave-1 outreach to land in the first 6 months or the project has zero revenue — operational risk too high.

Alternative C: Marketplace-commission shape (Skroutz path)

Take 7-12% commission on aggregated checkout, run the cart end-to-end ourselves.

Why rejected (for now). Skroutz took 11 years to make this pivot (competitive §2.3). It requires multi-merchant cart aggregator (#15) shipped + working + trusted. M1/M2 surface area doesn't support it. Decision-flip condition #2 above explicitly defers this to the M3 boundary review; this RFC is the current shape, not the forever shape.

Alternative D: No affiliate, run on direct retailer fees only

Featured placements + sponsorships only. Don't bother with click/conversion attribution.

Why rejected. Doesn't scale (per ticket framing) — caps at the count of paying retailers × monthly fee × % filled slots. No global aggregator anchors here and the ones that do stay tiny. Risky.

Alternative E: Buyer-side subscription as primary

Charge buyers $5/mo for advanced features (price history, alerts, build saving).

Why rejected. Violates the §41 ticket constraint: "monetisation that taxes the buyer is a non-starter" in a $600-1,800-build price-sensitive market. PCPartPicker is fully free; Pricena is freemium without paywalling the comparison itself. Subscription on top of free aggregation is a v3+ option once a power-user cohort is provable.

Open questions

These are the items that cross into "MASTER hasn't expressed a preference" territory. Per §41 ticket workflow, surfaced explicitly with options ranked recommend-first. Each option carries the consequence in one line.

The RFC recommends featured placement at $50-200/month per category, labelled Sponsored with visual differentiation. Three sub-questions MASTER may want to weigh in on:

  1. Recommend (default). Allow up to 1 featured slot per category page (max ~10 slots site-wide), Sponsored label, organic ranking ignores them. Trust is preserved; revenue ceiling is bounded.
  2. More aggressive. Allow 2-3 featured slots per category, label still Sponsored. More revenue; trust risk increases especially for first-time builders (persona §6.3 marks featured listings as Low receptivity for first-timers, anti-pattern for trust-seeking cohorts).
  3. Reject featured placement entirely. PCPartPicker shape — no paid placement. Cleanest trust posture; loses the $200-1,200/mo featured-stream estimate. Closes off a small but non-trivial revenue lever.

Recommend (1). The Lebanese norm tolerates clearly labelled placement per ticket framing, and persona-§6.3 marks featured as strong fit for Casual + Office IT specifically — exactly the surfaces (laptop / prebuilt category pages) where featured slots make sense.

OQ-2: B2B/SMB tier — opportunity scoping vs immediate pricing

Office IT buyers are a high-LTV outlier ($5-30k batches; $300-2,000/mo affiliate per Office-IT-shaped account per persona §6.3). Two paths:

  1. Recommend. Defer pricing to #16 admin tools / vendor backoffice; the RFC scopes the opportunity but doesn't price it. Quote-mode UX ships in #28 (persona §6.1.6); commercial layer waits for actual Office-IT buyer telemetry.
  2. Price now. Set a B2B-tier listing fee ($200-500/mo per supplier) targeting Lenovo distributors, Dell partners, etc. Higher revenue if it lands; risks pricing wrong without telemetry; introduces a different sales motion before we're sure it pays.

Recommend (1). Pricing without B2B-buyer telemetry is guessing; defer.

OQ-3: Aggregate market data product (post-launch)

Scraper data → aggregate price index, market reports, retailer-comparable benchmarks. Sell to retailers, suppliers, or financial / FMCG-adjacent analytics buyers. Three sub-questions:

  1. Recommend. Flag as a post-launch revenue stream, no design or pricing work in M1/M2. Revisit at M3 boundary.
  2. Price now. Spec a "961tech market intelligence" subscription at $500-2,000/mo for retailer-side analytics dashboards. High potential if it lands.
  3. Don't pursue. Some aggregators avoid productizing scrape data due to retailer pushback risk.

Recommend (1). Most of the M1/M2 effort is catalog and matcher; data-product work is a distraction.

OQ-4: Wave-1 retailer outreach plan ownership

The retailer audit (§7 #1) flagged that Lebanese retailer affiliate willingness is Unknown and outreach is the unblocker. Two paths:

  1. Recommend. RFC commits to an outreach plan as part of §Implementation plan below, with a Month-9 decision-flip checkpoint. MASTER owns outreach (per spec §6 hardest practical truth — Lebanese B2B is relationship-led).
  2. Defer to a separate ticket. Keep this RFC strictly to the rate-card; create a follow-up ticket for outreach planning + scripts.

Recommend (1). Outreach is the implementation; pulling it into a separate ticket fragments the decision.

Implementation plan

Routes to existing tickets. No new tickets needed unless OQ-2 or OQ-3 flips against the recommendation.

  • CPS payment + reconciliation → #17 affiliate reconciliation S2S postback. RFC sets the contract and rate; #17 implements it. Coupon-code attribution as third-tier fallback per spec §5.3 ©.
  • CPC tracking + per-retailer rate-card configuration → no existing ticket. Surface in #16 admin tools / vendor backoffice (rate config) + click-attribution work in #17 (the same click-token storage already covers CPC-counted clicks); if ungrouped, file a follow-up.
  • Featured placement surfaces → #28 page design. RFC sets Sponsored labelling and category-top-slot constraint; #28 designs the UI affordance.
  • Vendor backoffice billing + invoice flow → #16 admin tools. Whish corporate + USDT TRC20 + OMT settlement options surfaced per spec §6.4.
  • Subscription gate consideration for #14 → #14 price-drop alerts. RFC says: stays free in M1/M2; revisit if power-user telemetry surfaces.
  • Legal / VAT detail → #40. RFC flags VAT below threshold in Year 1; PIT additional line on R10; SARL incorporation gated to $1.5k+/mo MRR per spec §7.1.
  • Public affiliate code-of-conduct page → file as a new content-shaped ticket post-acceptance. Modelled on PCPartPicker's; commits 961tech to non-discriminatory ranking and per-source attribution.
  • Wave-1 retailer outreach plan → MASTER-owned per OQ-4 recommendation. Targets per spec §6.2: Tech Titan, PC Station LB, PCBuildingLeb, Sbeity, Macrotronics. Month-9 checkpoint for decision-flip evaluation.
  • Glossary updates → add CPS, CPC, Featured placement, Stage-1/2/3/4 partner definitions to glossary. Done as part of this commit.

Out of scope

  • Implementation code — all S2S postback / billing / dashboard work belongs in #17, #16, #28.
  • Tax / income classification mechanics — flagged at the stream level; deferred to #40.
  • Retailer outreach scripts — operational work, post-acceptance.
  • Forecasting — RFC uses ranges and labels them as estimates per §41 ticket constraint; a forecast requires GMV telemetry we don't yet have.
  • Pricing detail for B2B/SMB tier — deferred per OQ-2.
  • Pricing detail for aggregate market data — deferred per OQ-3.
  • Buyer-side payments / subscription — explicitly rejected per Alternative E + §41 ticket constraint.

Sources

Inline [N] citations refer to entries here. Where a claim has no verifiable public source, the RFC tags it (estimate) or (industry pattern, no public source) per §41 spot-check methodology. Detailed per-stream analysis with full citation blocks lives in reference/monetisation.md § Sources; this section is the RFC's own evidence anchor.

[1] Skroutz — 7-15% commission per category + €469 setup fee (2024-2025). Verified partially. Per-category structure confirmed; band publicly cited as 7-15% (not the spec's 7-12%); €469 setup confirmed for 2024-2025 (was €248 in 2022 — pricing has escalated). Skroutz's own merchant commission page was 403-blocked on the verification fetch; range comes from integration-vendor coverage. Sources: e-tailize integration guide, ChannelEngine Skroutz marketplace guide.

[2] Idealo — published per-country CPC + €465M Berlin Regional Court ruling. Spec was wrong on "rates not public" — Idealo publishes per-country, per-category CPC rates openly, with an automatic +€0.02/year escalator. Current rates: DE €0.51, UK £0.35, AT €0.36, FR €0.40, IT €0.25, ES €0.35 (effective 2025-04-01). Monthly minimum revenue €20 on idealo.de. Source: partner.idealo.com/de/preise, partner.idealo.com/uk/pricing. Berlin court ruling: Landgericht Berlin II, Case 16 O 195/19 Kart(2), judgment 14 Nov 2025. Idealo €374.1M damages + €91.1M interest = €465M; Producto GmbH (Testberichte.de) €103.7M; combined ~€572M. Sources: SCiDA project legal commentary, TechCrunch coverage.

[3] PCPartPicker — disclosure + code-of-conduct pages confirmed; effective rate not publicly disclosed. Both /disclosure/ and /code-of-conduct/industry-affiliate/ exist and are search-index confirmed (direct WebFetch returned 403 anti-bot). The "0.5-1% effective rate" widely repeated in industry coverage has no public primary source — flag as (estimate). Source for page existence + content snippets: PCPartPicker disclosure, PCPartPicker industry-affiliate code of conduct, Wikipedia: PCPartPicker.

[4] Geizhals — dual Pay-by-click OR Pay-by-order model with public CPC rates; Heise ownership since June 2021. Spec was wrong on "pure CPC, no bidding." Wikipedia confirms dual model. Public CPC rates: €0.38 per click with Geizhals logo on 90% of subpages, €0.41 per click without. Sort-tiebreaker chain (price → shipping → stock → rating) claimed in spec is unverifiable from primary sources within scope. Sources: unternehmen.geizhals.at/haendler/, Wikipedia: Geizhals.

[5] Pricena — 2-10% CPA per category, per founder Wamda interview, 2014. ⚠ Date warning: 2014 source, twelve years old. Only public, attributable MENA aggregator rate disclosure found in research. No newer disclosure in primary sources. Pricena explicitly skips Lebanon. Source: Wamda — Pricena is growing and fast (2014).

[6] Amazon.ae Associates — 2.00% CPS on Electronics / PC store / Wireless category. Cleanest 2026 MENA-region electronics-category comparable (primary, verbatim from commission table). Other category rates: Apparel 9%, Books 8%, Beauty/Health 7%, Furniture 5%, Grocery 4%. Source: affiliate-program.amazon.ae commission table. Noon Affiliate publishes "up to 10%" headline (electronics typically 2-4%), with an alternate $1.50/order CPO via CPX network and 80% confirmation rate on GCC orders. Source: affiliates.noon.com, CPX Affiliate Noon program detail. eXtra (KSA) runs no public self-serve affiliate program (curated bilateral only) — confirms regional norm of opacity.

[7] Daraz Bangladesh affiliate — "up to 18%"; voucher-code attribution alongside hyperlink clicks; 7-day cookie window; outsourced to DCM Network (MENA). Primary-source confirmation that coupon-code attribution is a parallel attribution mechanism alongside click-tracking in markets with thin postback adoption — direct support for the spec's third-tier coupon-code fallback. Daraz also disclosed seller commission band 0-17.2% by category (electronics/computer accessories ~3.4-5.1%; fashion 12-17%); minimum payout BDT 2,500 (~$21). Sources: daraz.com.bd affiliate program, daraz.pk affiliate program, Daraz seller commission rates summary.

[8] MySmartPrice revenue mix — 80% affiliate / 18% advertising-and-online-directory / 2% other. ⚠ Date not crisp on the StartupTalky source — treat as directional, not 2026-current. Useful single-data-point ceiling: even India's most mature affiliate-led aggregator can't push display advertising past ~18% of revenue. Sources: StartupTalky MySmartPrice profile. Smartprix headcount per Tracxn (Aug 2025): 2 employees, 50% YoY decline — comparison-site survival ≠ scale. Sources: Tracxn: Smartprix. 91mobiles annual revenue ~₹49.8 Cr (~USD 6M) FY-ending Mar 2025; same-shape comparator with brand advertising (Samsung, Nokia, Airtel cited) layered on top of CPS. Source: Inc42 — 91mobiles acquires Pricebaba (2018).

[9] MercadoLibre — public seller commission 11-17% per category; commerce blended take rate ~22-24% (commerce revenue / GMV); Mercado Ads 2.1% of GMV in Q4 2024. 10-K confirms structure: "Final value fees represent a percentage of the sale value… charged to the seller once an item is successfully sold." Naive arithmetic from Q4 2024: $12.2B commerce revenue / $51.5B GMV = 23.7% blended (includes shipping fees + ads + commissions). Critical observation for the RFC: no LatAm peer found denominates seller fees in USD — even with severe FX volatility (Argentine hyperinflation triggers IAS 29 accounting in MELI filings), the practice is local-currency fees plus platform-side FX hedging. Lebanon's case differs because Lebanese retailers themselves price in USD per retailer audit §2-§3; the LatAm precedent doesn't directly apply. Sources: MELI Q4 2024 press release, MELI 10-K filed Feb 2025 — SEC EDGAR, Mercado Ads %-GMV trend, global-selling.mercadolibre.com selling fee.

[10] Buscapé / Zoom — historical hybrid CPC + CPS model, post-2019 collapse; PriceCheck (South Africa) — prepaid CPC bidding model with R0.10 increment, parallel marketplace commission, "FREE to list, NO monthly subscription fees." Buscapé as standalone comparator largely a cautionary case — Naspers sold to Zoom 2019, Tracxn lists 2 employees as of Jul 2024. The unbundled price-comparator collapsed under Google Shopping pressure. PriceCheck pattern is directly portable: prepaid credits + bid-up ranking on the comparison surface, separate marketplace commission post-sale. Sources: NeoFeed: Buscapé — quase fim de um ícone (2019) ⚠ pre-2022, Tracxn: Buscapé, PriceCheck merchant CPC help, PriceCheck merchant FAQ portal.

See reference/monetisation.md § 5 Sources for full citation blocks per source.