Measure the real task first. There is no universal number of pages per GB.
Measure traffic before a large collection job
Run a representative sample of pages with the same browser settings as the intended job. Record the sample size and transferred data for the same period. Images, video, retries and duplicate requests can change the result substantially.
Use the sample to estimate your traffic, allow for variation and monitor usage during the full run. USM rotating plans last 30 days; unused traffic does not roll over, and exhausted allowances stop connections without automatic top-ups.
Calculate traffic from your sample
Estimated traffic = sample traffic ÷ completed sample tasks × planned tasks. Use the same measurement source and unit throughout. Actual usage may differ from this estimate.
For example, a sample using 0.2 GB for 100 completed checks implies about 2 GB for 1,000 comparable checks. This assumes the next run behaves like the sample. The figures do not describe any particular website.
Allow for differences between runs
- Images and other page resources required by the task.
- Retries, redirects and incomplete attempts that still transfer data.
- The difference between a rendered browser and a simple HTTP request.
- How often you repeat checks and whether earlier results are still useful.
Review usage while the run is active
Start with a limited batch and compare actual consumption with your estimate. If usage rises sharply, pause and check why. Extra traffic may not be producing useful results.
GB allowances limit traffic, while plans sold by IP count limit the number of IPs. Check the other limits and the term of each plan; neither billing model means unlimited use.