Methodology
How TradeInFinder decides what to show.
TradeInFinder is built around simple direct value, resale comparisons, and clean upgrade paths. We show source, freshness, and confidence because a useful number is only useful if you can trust what it means.
How values are normalized
Every value record includes a source URL, raw ingest reference, retrieval time, staleness threshold, confidence score, and verification status.
Value types stay separate: instant credit, store credit, gift card, purchase credit, and resale estimate.
User-facing conditions are simplified to Good, Damaged, and Poor / not accepted.
Values without source or freshness metadata are not allowed into the normalized layer.
Confidence and freshness
Confidence rises when the match is exact, recent, parser quality is high, and the merchant itself is trustworthy.
Verified values can still go stale if they age past the record's freshness threshold.
Low-confidence or family-level estimates show as ranges instead of fake exact amounts.
Manual overrides can improve trust when an admin has reviewed the value directly.
Fallback hierarchy
TradeInFinder uses the cleanest available match first, then steps down in a visible and honest order.
1. Exact verified value
2. Exact estimated value
3. Same model, different storage adjusted estimate
4. Similar family estimate
5. Unavailable
What the ranking rewards
A slightly lower but verified value can rank above a higher stale estimate because trust matters.
The ranking blends usable value, freshness, confidence, merchant trust, and exactness of match.
Low-confidence values are visible, but they should not dominate the page just because the raw number is high.
Resale remains clearly labeled as an estimate with more uncertainty than direct store value.