Skip to main content

Working with historical data

Tables preserve every extraction your robot performs, creating a comprehensive historical database. Understanding how to work with this historical data helps you track changes over time, analyze trends, and manage your data effectively.

M
Written by Melissa Shires
Updated today

✅ In order to automatically create a historical database, enable monitoring for your robot.

Historical data includes all previous extractions with identical parameters:

  • Same Origin URL

  • Same input parameters (search terms, filters, etc.)

  • Multiple extraction dates

For example, if you monitor Amazon prices daily for "laptop", you'll have one row for each day's extraction - that's your historical data.

Show Historical Data

In Tables, you have the ability to show or hide historical data using the checkmark in the top right corner.

☑ Show Historical Data (checked)

Displays:

  • All extraction runs ever performed.

  • Multiple versions of the same data.

  • Complete timeline of changes.

  • Every attempt, including duplicates.

Perfect for:

  • Tracking price changes over time.

  • Monitoring inventory fluctuations.

  • Analyzing content evolution.

  • Building trend reports.

☐ Show Historical Data (unchecked)

Shows only:

  • Most recent extraction for each unique parameter set.

  • Current state of your data.

  • Latest successful run per URL.

  • Deduplicated view without repetition.

Perfect for:

  • Current data snapshots.

  • Clean exports without duplicates.

  • Status dashboards.

  • Quick data reviews.

Example - product page monitoring

Imagine monitoring a product page daily for a week.

With historical data ON:

  • 7 rows (one per day)

  • See price progression: $99 → $95 → $89 → $92 → $87 → $85 → $79

  • Identify trends and patterns

With historical data OFF:

  • 1 row (latest extraction)

  • See current price: $79

  • Clean, current view

Working with historical data effectively

Identifying extraction versions

Each historical entry includes:

  • Extract Date - When the data was captured

  • Status - Success or failure of that run

  • Origin URL - What page was scraped

  • Input parameters - What variations were used

Common historical data patterns

Daily monitoring

  • Same URL extracted every day

  • Historical view shows progression

  • Hide historical for current snapshot

Bulk operations

  • Multiple URLs extracted once

  • Historical data rarely needed

  • Each URL typically has one entry

A/B testing

  • Same URL with different parameters

  • Historical tracks all variations

  • Essential for comparison

Combining historical toggle with filters

View this week's price changes

  1. ☑ Show Historical Data

  2. Filter: Extract Date = This week

  3. Filter: Status = Successful

  4. Result: Complete price history for the week

Latest data from all sources

  1. ☐ Show Historical Data

  2. No filters needed

  3. Result: Current state across all URLs

Failed extractions over time

  1. ☑ Show Historical Data

  2. Filter: Status = Failed

  3. Result: Pattern of failures to investigate

Historical data and exports

💡 Enabling the historical checkbox applies when you export data.

Exporting with historical ON:

  • Includes all versions

  • Larger file sizes

  • Complete audit trail

  • Time-series analysis ready

Exporting with historical OFF:

  • Only latest data

  • Smaller, cleaner files

  • Current state snapshot

  • No duplicates

Use cases for historical data

Price monitoring

Track product prices over time:

  • Daily extractions build price history

  • Identify best times to buy

  • Spot pricing patterns

  • Alert on significant changes

Content tracking

Monitor website updates:

  • Track article revisions

  • Document policy changes

  • Archive terms of service

  • Prove content timing

Inventory analysis

Watch stock levels:

  • Build availability patterns

  • Predict restocking cycles

  • Identify fast-moving items

  • Plan purchasing decisions

Competitive intelligence

Monitor competitor changes:

  • Track pricing strategies

  • Document feature updates

  • Analyze marketing messages

  • Build competitive timeline

Managing large historical datasets

Use filters strategically

  • Date ranges to focus on recent data

  • Status filters to exclude failures

  • Origin URL filters for specific sources

Export in chunks

  • Monthly exports for archiving

  • Filtered exports for analysis

  • Recent-only for reporting

Plan extraction frequency

  • Daily for volatile data

  • Weekly for stable content

  • Monthly for archival purposes

Did this answer your question?