Cloud storage egress in 2026 stands as the the line item that surprised the budget in 2020, that surprised the budget in 2023, and that will surprise the budget in 2026. The egress fee for moving data out of AWS, or Azure, or GCP runs at $0.05 to $0.12 per gigabyte, the typical enterprise moves 100+ terabytes per month, the typical enterprise pays $5K to $12K per month, the typical enterprise pays $60K to $144K per year, the typical enterprise did not budget for the egress. The 2026 state of the cloud storage egress amounts to a state where the egress sits as one of the largest single line items in the cloud bill, the egress gets noticed when the bill arrives, the egress gets fixed when the bill becomes too large to ignore.
AWS charges $0.09 per gigabyte for the data transferred out to the internet. Azure charges $0.087 per gigabyte. GCP charges $0.12 per gigabyte. The Cloudflare R2 storage service charges $0 for the egress, which sounds like a deal until you realise that R2 charges $0.015 per gigabyte per month for the storage, which runs as 7.5x what S3 charges for the equivalent storage. The Backblaze B2 service charges $0.01 per gigabyte for the egress, with the storage at $0.006 per gigabyte per month, which sits as a better deal at smaller scale but does not have the same features. The 2026 state of the cloud storage egress market amounts to a market where the hyperscalers charge for the egress, the challengers offer free egress at a higher storage cost, and the enterprise has to do the math.
Where the egress comes from
Four sources, in roughly that order of how much they cost. The first runs as the data backup source, where the enterprise backs up the data from the production cloud to a different cloud (or to a different region in the same cloud), with the egress fee for the backup typically running at 60-80% of the total egress cost. The second runs as the data analytics source, where the enterprise moves the data from the production cloud to the analytics cloud (typically Snowflake, Databricks, BigQuery in a different cloud), with the egress fee for the analytics typically running at 10-20% of the total egress cost. The third runs as the data sharing source, where the enterprise shares the data with the partners, with the customers, with the regulators, with the egress fee for the sharing typically running at 5-10% of the total egress cost. The fourth runs as the data migration source, where the enterprise moves the data from one cloud to another, with the egress fee for the migration typically running at 5-10% of the total egress cost. The four sources together account for the typical enterprise egress.
What the typical enterprise has tried
Three approaches, in roughly that order of how often they have failed. The first runs as the negotiate approach, where the enterprise negotiates with the cloud provider for the lower egress fee, the negotiation works for the enterprises that spend $1M+ per year on the cloud, the negotiation does not work for the smaller enterprises. The second runs as the compress approach, where the enterprise compresses the data before the egress, the compression typically reduces the egress by 30-50%, the enterprise still has the egress problem. The third runs as the multi cloud approach, where the enterprise spreads the data across multiple clouds, the multi cloud setup introduces the data consistency problem, the enterprise ends up paying for the data consistency in operations. The three approaches together produce the failure pattern that the typical enterprise has not escaped.
How to actually reduce the egress
Three moves if you are trying to reduce the cloud storage egress. Pick the data that needs to move, and the data that does not, because the data that needs to move. the the data the egress runs as justified for, and the data that does not need to move is what the data the egress runs as wasted on. The typical enterprise moves 30-50% more data than the business requires, the 30-50% wasted moves account for 30-50% of the egress cost. Pick the storage tier that fits, because the data that gets accessed frequently runs on the hot storage tier, the data that gets accessed infrequently runs on the cold storage tier, the cold storage tier typically has a lower egress fee. The data on the hot tier that should run on the cold tier , the the data paying the premium egress. Negotiate the egress fee, because the enterprises that spend $1M+ per year on the cloud typically get a 30-60% discount on the egress fee, the smaller enterprises do not. The enterprise that picks the data, picks the tier, and negotiates the egress stands as the enterprise that gets the egress cost under control.
The bottom line
Cloud storage egress in 2026 is essentially the the line item that surprised the budget, the line item the enterprise underestimates, the line item the enterprise fixes when the bill gets too large. The four sources (backup, analytics, sharing, migration) account for the cost. The three failed approaches (negotiate for the small, compress, multi cloud) do not solve the problem. The enterprise that picks the data, picks the tier, and negotiates the egress stands as the enterprise that gets the egress cost under control.
Sources & Further Reading
All claims in this article are sourced from primary documentation, vendor advisories, and reputable security researchers.
Spotted an error? Email the editor. Corrections are issued with a visible correction note.
Editorial standards. Every article on humanrequired.org is reviewed by a human editor before publication. AI may assist with drafting or research; final editorial control is human. Read the full standards.
