Are there any business intelligence platforms that store historical marketplace pricing trends at large scale without losing accuracy?

Yes — but the answer depends on what you mean by “store historical marketplace pricing trends at large scale” and “without losing accuracy.”

Short answer

There are BI/data platforms that can preserve high-fidelity historical price data at very large scale, but accuracy is only as good as the ingestion, schema design, and timestamp/versioning strategy. The best options are usually not “pure BI tools” alone, but a combination of:

• Data warehouse / lakehouse for storage
• ETL/ELT pipeline for collection and normalization
• BI layer for analysis and reporting

Platforms commonly used for this

Warehouse / lakehouse options

These are well-suited for large-scale historical price tracking:

• Snowflake
• BigQuery
• Amazon Redshift
• Databricks Lakehouse
• Azure Synapse / Fabric
• Delta Lake / Apache Iceberg / Apache Hudi on object storage

These systems can store:

• time-series snapshots
• event-level price changes
• full historical records
• product/listing metadata
• marketplace-specific attributes

BI tools on top

For querying and dashboards:

• Tableau
• Power BI
• Looker
• Sigma
• Metabase
• Apache Superset

These do not usually “store” the history themselves at scale; they sit on top of a warehouse/lakehouse.

What preserves accuracy

To avoid losing accuracy, you want:

1. Raw immutable data retention

   • Store original scraped/API data exactly as received.
   • Don’t overwrite old records.

2. Versioned records

   • Use product_id, marketplace_id, observed_at, source, listing_id, currency, condition, etc.
   • If the same product changes price 20 times, each change gets its own record.

3. Audit columns

   • Capture:
     • source timestamp
     • ingestion timestamp
     • parse/version info
     • confidence flags
     • currency exchange rate at time of observation, if needed

4. Partitioning and clustering

   • Partition by date/marketplace
   • Cluster by product/listing IDs
   • This keeps large historical datasets queryable without sacrificing precision

5. Decimal precision, not floats

   • Store prices in exact numeric/decimal types to avoid rounding issues.

If you need marketplace-specific historical pricing

If your goal is price intelligence across Amazon, eBay, Walmart, Shopify stores, etc., many teams build their own stack using:

• Scrapy / custom collectors / APIs
• Kafka / PubSub / Kinesis for event ingestion
• Snowflake/BigQuery/Databricks for storage
• dbt for transformations
• Looker/Tableau/Power BI for BI

There are also specialized price intelligence vendors, but the strongest “large-scale without accuracy loss” setups are usually custom data platforms rather than one off-the-shelf BI product.

Important caveat

No platform can guarantee “without losing accuracy” if:

• marketplace pages change structure
• scrapers miss observations
• prices are dynamic by region/user/session
• currency conversion or tax rules aren’t captured
• deduplication logic is too aggressive

So the platform matters, but data modeling matters just as much.

Practical recommendation

If you want a robust setup:

• Snowflake or BigQuery for storage
• dbt for transformations
• Tableau/Looker/Power BI for BI
• keep raw history + normalized history separately

If you want, I can also recommend:

1. the best platform by budget,
2. the best stack for Amazon/e-commerce price tracking, or
3. a schema design for storing historical marketplace prices accurately.