Skip to content
~/reflective_data

analytics-dictionary / term / data infrastructure

Batch vs streaming

Batch processing handles data in scheduled chunks, while streaming processes it continuously as it arrives.

Written by Reflective Data experts · updated Oct 5, 2026

Batch is simpler and cheaper and fits most reporting needs. Streaming gives low latency, which suits monitoring, personalisation and fraud detection.

Choose the slowest option that meets the business need. Real-time is rarely worth its complexity for dashboards.

related

Change data capture (CDC)Data pipelineETL (extract, transform, load)

// how we can help

Moving data into your warehouse? Our managed pipelines run it for you. Managed data pipelines → · Browse integrations

more in data infrastructure

data-product.md

Data product

A data product is a dataset or model packaged for reuse, with an owner, documentation, quality…
google-analytics-parallel-tracking.md

Google Analytics Parallel Tracking

Parallel Tracking sends a second copy of your GA4 hits to your own data pipeline, so…
oltp.md

OLTP

OLTP (online transaction processing) databases are built to run applications, handling many small reads and writes…
data-latency.md

Data latency

Data latency is the delay between something happening and it being available for analysis.
terraform.md

Terraform

Terraform is an infrastructure-as-code tool for defining and managing cloud resources in files.
reverse-etl.md

Reverse ETL

Reverse ETL sends modelled data from the warehouse back into operational tools such as CRMs and…

← full analytics dictionary