Methodology 2026.07

1. Category benchmark

Each assessable category has a versioned benchmark with price classes, mandatory gates, expected class-median features, above-class bonus features and negative characteristics. It is calibrated so a typical product in its class scores around 70–76.

2. Knock-out gates

A score is calculated only after all four gates pass:

At least one documented European price in euros; manufacturer-only distribution is never penalised.
Unambiguous brand, model and variant identity.
Every mandatory category feature is present or transparently marked as a legally standard assumption.
No active safety recall.

Unverified price or identity means not assessable. A missing mandatory feature or active recall means rejected at the gate. We do not guess through uncertainty.

3. Price class and boundary rule

The lowest documented European offer determines the price class. Within ±10% of a class boundary, both neighbouring classes are calculated and the better result counts. If the cheaper class is borrowed, a −3 adjustment replaces any price-position bonus.

4. Deterministic score

70 − 6 × missing expectations + bonus features (max. +15) + negative characteristics + warranty adjustment + price-position bonus

Warranty above 24 months adds 3; 24 months or unknown adds 0; below 24 months subtracts 3. Price position adds 5 in the lowest third, 2 in the middle third and 0 in the upper third. The result is clamped to 0–100.

5. Evidence rule

Only verified claims count. Every benchmark-relevant feature carries a source, retrieval context and excerpt. T1 submitted evidence takes precedence over T2 manufacturer sources, T3 retailers or price portals, and T4 other web sources. Conflicts are resolved conservatively.

T1Manufacturer or team submissionT2Manufacturer publicationT3Retailer or price portalT4Other documented web source

6. One award at 70

There is exactly one European Best Value Award. It is earned at 70/100 or more after every gate has passed. There are no tiers or grades. Payment never changes a score or a gate result.

7. Version and change control

Changes to a benchmark, threshold, formula or gate create a new version. Published results permanently retain the versions used. Calibration can affect only a future benchmark version, never an existing decision.

{ "changelog": "2026.07 \u2014 Messlatten-Methodik: Bewertung gegen versionierte Kategorie-Benchmarks (Preisklassen, Pflicht/Erwartet/Plus/Negativ). K.o.-Gates, Grenzfall-Regel \u00b110 %, Garantie- und Preispositions-Bonus, Plus-Cap +15. Nur belegte Werte z\u00e4hlen (Beleg + Tier; ausdr\u00fccklich eingereichte T1-Fakten haben Vorrang). Pilot 2026-07-16: 5/5 Validierungsf\u00e4lle korrekt. Ersetzt das fr\u00fchere 9-Bereiche-Regelwerk (archiviert). 2026.07-2 \u2014 Eine unbelegte Klassen-Erwartung kostet 3 statt 6 Punkte; widerlegte Erwartungen bleiben bei 6. Grund: nur der widerlegte Fall ist eine Aussage \u00fcber das Produkt, der unbelegte kann auch heissen, dass der Hersteller wenig ver\u00f6ffentlicht.", "engine": "benchmark", "evidence": { "assumed_declared_by": "benchmark", "tiers": [ "T1", "T2", "T3", "T4" ] }, "gates": { "eu_price_verified": true, "identity_confirmed": true, "mandatory_features": true, "max_price_age_days": 30, "no_active_recall": true }, "precheck": { "base_spread": 4.0, "coverage_spread": 12.0, "min_outreach_coverage": 0.35, "min_outreach_mid": 68.0 }, "scoring": { "basis": 70.0, "garantie_bonus": 3.0, "garantie_malus": -3.0, "grenzfall_leihe_malus": -3.0, "grenzfall_toleranz": 0.1, "malus_erwartung": 6.0, "malus_erwartung_unbelegt": 3.0, "plus_cap": 15.0, "preisbonus_mittleres_drittel": 2.0, "preisbonus_unteres_drittel": 5.0, "schwelle": 70.0 }, "thresholds": { "award": 70.0 } }