Degraded data quality
A yellow or red data-quality badge on a search result means one or more sub-scores fell back from their preferred data source. The SellerScore is still computed, but trust it less.
This page is the diagnostic.
Read the warnings on the row
Hover or tap the data-quality badge. You’ll see a list of warning codes with short explanations.
→ Data quality and warnings for the full reference of every warning code.
Common warnings and what to do
demand_sfr_unavailable
SFR (Brand Analytics) couldn’t be fetched. Demand sub-score uses BSR + Google Trends fallback.
What to do:
- For very common keywords (top 10k by search volume), the fallback is reasonably accurate; trust the demand sub-score within ±10.
- For niche keywords, fallback is noisier; cross-check against the actual BSR shown on the row.
- Re-run the search later; SFR availability fluctuates.
margin_cogs_estimated
COGS estimate used the benchmark table because Crawlee/AliExpress data wasn’t available. Today, this fires on every search (Crawlee is Coming-soon).
What to do:
- Treat the margin sub-score as a floor — your real supplier price is likely better, sometimes much better.
- For high-conviction products, do an AliExpress / Alibaba lookup to validate the benchmark.
- When Cross-platform pricing ships, this warning will mostly disappear.
trend_pytrends_outage
Google Trends was rate-limited or unreachable. Trend and seasonality sub-scores fell back to a neutral 50.
What to do:
- Wait and re-run — pytrends outages are usually short.
- If it’s persistent (over a few hours), file a ticket. Hilal monitors pytrends availability and may have already noticed.
- For decision-making, trust the BSR and review counts more than the trend sub-score until it refreshes.
cross_platform_disabled
Cross-platform sub-score is the placeholder. Today, this fires on every search.
What to do:
- Lower the cross-platform weight to 0% for now; the placeholder isn’t informative.
- Renormalize the other weights (demand 28%, margin 22%, competition 22%, trend 17%, seasonality 11%, cross-platform 0%).
- When Crawlee ships, restore the default weights.
fees_partial
SP-API returned partial fee data. Margin sub-score is partly estimated.
What to do:
- Re-run the search; this is usually transient.
- If persistent, the product may be in a category with unusual fee structure. Open the product detail and verify the fee breakdown looks sensible.
bsr_missing
BSR data unavailable for the product. Often co-fires with demand_sfr_unavailable.
What to do:
- If the listing is very new (< 30 days), BSR data may simply not exist yet. Move on or wait.
- For older listings, BSR-missing is unusual and indicates an upstream issue. Re-run.
When data quality is yellow vs. red
| Yellow | Red | |
|---|---|---|
| Threshold | One sub-score fallback | Two+ fallbacks, or a critical (margin) fallback |
| What it means | The score is approximate but directionally correct | The score is informational; verify before acting |
| Typical action | Glance at the warnings; usually proceed | Open the product detail; read score reasoning; verify independently |
Most searches today carry at least yellow because margin_cogs_estimated and cross_platform_disabled are universal until those Coming-soon features ship. That’s expected; don’t panic at yellow.
Red is the signal worth pausing on. A red SellerScore that looks high probably is partly fabricated. Open the detail, read the reasoning, and don’t commit money until you’ve verified.
What to do for a consistently low data quality
If every search you run comes back red:
- Check the marketplace — newer or smaller marketplaces have less data. US, UK, DE are best; some smaller marketplaces are noisier.
- Check your Brand Registration status — Brand Registered sellers get richer data on their account-level queries.
- File a ticket if patterns suggest an account-level data issue.