Popular searches
Computer & gaming
Audio & video
Home & kitchen
Mobile & office
Personal care & style
Currently Analyzing
Recent searches
How the reliability score (S‑Score) is built
S‑Score compresses price/performance, durability, and “would I trust this buy?” signals into one 0–100 number. The cheapest listing does not automatically win; review patterns, price stability, how well the product matches your query, and brand/support signals are combined.
What inputs does the model use?
- Themes extracted from user reviews across sources (repeat failures, praise, comfort, etc.)
- Price level and volatility (flash drops, noisy promos, TR vs global offer consistency)
- Brand, warranty, and support signals (firmware updates, service access — category dependent)
- Overlap between what you typed and the product’s specs (AI query–product match)
- Seller and fulfilment signals (clear returns, trusted channels, stock consistency)
Score bands
Example: one product
With a live API, each row is generated from real data; for now this is illustrative copy.
Keychron K8 Pro
Wireless + wired, TKL, hot-swap (sample)
Legendary
Review synthesis
PBT keycaps and dual connectivity are often praised; switch failure complaints are rare and scattered.
Price stability
Reasonable recent variance; TR and global offers sit in similar clusters, no suspicious one-off pricing.
Query match
Requests like “mechanical, wireless, TKL” align strongly with the spec sheet.
Brand & support
Firmware and community support read positive for keyboards; documentation and update signal is strong.
Summary: strong reviews, high query–feature match, stable price band — hence sample S‑Score 96 (Legendary band).