The "Easy Button" Promise
Think of S3 Intelligent-Tiering (S3-INT) like a smart thermostat for your storage bill. Just as a thermostat automatically adjusts the temperature based on whether you're home, S3-INT automatically moves your data between cheaper and more expensive storage tiers based on how often it's accessed. AWS markets this as a zero-effort way to cut costs — enable it once and let the algorithm do the work.
The Catch
S3-INT is not free to operate. AWS charges a Monitoring and Automation Fee for every object it watches — regardless of whether that object ever moves to a cheaper tier. This fee is small per object, but it scales linearly with object count. If your bucket contains millions of tiny files, the cumulative monitoring cost can exceed the storage savings you'd ever achieve by moving those files to a colder tier.
S3 Intelligent-Tiering Math is the discipline of calculating the break-even point before enabling the feature. The core question is: "For this specific bucket, does the storage savings from tiering outweigh the monitoring fee for every object?" Getting this wrong means paying AWS to watch data that never saves you a cent.
At scale, this mistake is expensive. A bucket with 500 million small objects at $0.0025 per 1,000 objects costs $1,250/month in monitoring fees alone — before a single byte moves tiers. Senior engineers treat S3-INT as a tool with a specific use case, not a universal cost-reduction switch.