Source of the figures. The simulator is based on Fentech's published average results: +10% sales and −15% inventory across our clients — more than 20 brands and retailers from €5M to €300M revenue. Winner Paris Retail Awards (Data category) and E-commerce Start-up of the Year 2025 (FEVAD).
Conservative Year 1 reading. The figures shown here are a Year 1 estimate, not cruising speed. At the start of the project, we deliberately cover more conservatively to avoid stock-outs during the model's calibration phase (learning business rules, data integration, validation of coverage strategies). The published results (+10%/−15%) are reached at cruising speed, typically from Year 2 onwards.
Per-lever detail. Click the "i" button to the right of each card in the result panel — each lever has its own targeted explanation (formula + specific assumptions).
Cross-cutting effects. · Saisonnalité : amplitude pic/creux annuel. Modulateur 0,85× (faible) à 1,35× (forte). Plus l'amplitude est forte, plus une bonne prévision a de la valeur.
· Launch intensity: the more you launch, the more a forecast error costs (no way to recover). Multiplier 0.85× (2 launches/year) to 1.45× (24+/year) on stock-outs and markdowns.
· Clairvoyant cost: public subscription + integration grid, based on revenue. Shown Year 1 (decreases from Year 2).
Onboarding phase. Les 6-12 premières semaines mobilisent un peu d'effort côté client : intégration et vérification des données, alignement sur les règles métier (calendriers, MOQ, lead times, événements commerciaux), validation des stratégies de couverture. C'est ce qui rend le modèle prédictif fiable et explicable. Une fois en place, votre équipe pilote.
Assumed limit. This simulator gives an order of magnitude, not an audit. The real value is calculated during a mini-audit (30 min) on your actual data.
Fentech/Clairvoyant