Real-Time vs Historical Data Real-Time vs Historical Data
  1. Zype
  2. Analytics
  3. Analytics 3

Real-Time vs Historical Data

Overview

Analytics 3 API introduces a flexible and powerful framework for querying both real-time and historical data, with control over fidelity intervals (granularity). Understanding the difference helps you select the right data for your use case.

 

Real-Time Data

  • What it is: Ingested and queryable almost immediately (delay < 1 min)

  • Use Cases:

    • Monitoring live viewer engagement

    • Operational dashboards

    • Near-real-time alerting

  • Caveats:

    • Higher likelihood of minor accuracy fluctuations

    • Typically available with limited fidelity (e.g., 1-minute or 1-hour buckets)

    • Best for short time windows (e.g., last 30 mins to 24 hours)

Historical Data

  • What it is: Aggregated, validated, and stored for long-term analysis

  • Use Cases:

    • Business intelligence reporting

    • Viewing trends over weeks or months

    • Comparing performance across content, time periods, or user cohorts

  • Caveats:

    • More accurate and consistent than real-time

    • Supports broader time windows and higher fidelity intervals

Fidelity Intervals

Analytics 3 supports various fidelity intervals (granularity levels), such as:

  • 1min, 5min, 15min, hour, day, week, month

  • Applies to both real-time and historical queries (based on availability)

Best Practices

Time WindowRecommended Interval
< 24 hours1min, 5min, 15min
1–7 dayshour, day
7–30+ daysday, week, month

Note: Not all intervals are supported on all endpoints. Always refer to the endpoint documentation.
 

Additional Query Controls

group_by

Use group_by to segment results by one or more dimensions. You can combine multiple values with commas (e.g., group_by=platform,video_id).

Available group_by values:

  • video_id
  • video_title
  • platform
  • subscription_id
  • transaction_id
  • monetization_type
  • playlist_id
  • playlist_name
  • app_id
  • app_name
  • app_version
  • player_name
  • player_version
  • content_type
  • playout_channel_id
  • playout_channel_name
  • playout_destination_id
  • playout_destination_name
  • platform_source
  • external_user_id
  • domain
  • country

sort_by

Sort your results by a field + direction:

sort_by=start_date_asc
sort_by=video_id_desc

limit and offset

Control result pagination:

limit=10       # Return only 10 records  
offset=20      # Skip the first 20 records

 

Summary

Data TypeLatencyBest ForInterval Options
Real-Time< 1 minDashboards, live event tracking1min, 5min, 15min, hour
HistoricalImmediateReporting, trend analysishour, day, week, month

Was this article helpful?

0 out of 0 found this helpful

Articles in this section