The cloud cost guide for small teams
Most overspending comes from four line items. Finding them takes an afternoon.
Small teams rarely overspend because of one bad architectural decision. They overspend because a handful of line items grow quietly and nobody owns the bill.
The usual four
- Data egress, especially traffic crossing regions or leaving to a third-party service on every request.
- Idle compute — staging environments, oversized instances and forgotten workers running around the clock.
- Storage tiers, where objects that are never read remain on the most expensive class available.
- Observability volume, which for many teams is the second largest bill after compute and grows with every added log line.
A one-afternoon process
Export the last three months of billing at the line-item level. Sort by cost. Take the top ten and for each one write down who asked for it and what breaks if it is halved. Most teams find that at least three of the ten have no current owner.
What not to do
Do not begin with a migration. Re-platforming to cut costs is a multi-quarter project that typically underdelivers against its own estimate. Exhaust configuration and rightsizing first; they carry a fraction of the risk.
About the author
Senior Writer, Software & Startups
Daniel Okoye covers software business models, developer tooling and early-stage company building. He has spent a decade analysing SaaS pricing and go-to-market strategy.
