What does an AWS infrastructure audit include for a small SaaS?
By the Keel team · Last checked · 5 min read
Short answer
An AWS infrastructure audit for a small SaaS should check how code reaches production, whether backups actually restore, who can access what, how the network and data are protected, what is monitored, and where money is wasted. A useful audit ends with a short, ranked list of fixes rather than a long report, and it should need only read-only access to your account.
6
Pillars of the AWS Well-Architected Framework
Source: AWS, Oct 2026
£0
Extra charge for the AWS Well-Architected Tool
Source: AWS Well-Architected Tool FAQs
28%
Organisations hit by ransomware that fully restored their data
Source: Veeam, 2026
What is an AWS infrastructure audit?
An AWS infrastructure audit is a structured review of how your AWS account and delivery pipeline are set up, measured against good practice, so you know what is risky and what to fix first.
For a small SaaS it is a practical check, not a compliance exercise.
It is different from a formal compliance audit such as SOC 2, which an independent auditor performs against a framework. An infrastructure audit is practical: it looks at your actual setup and tells you what could hurt you. If you are facing a security questionnaire, how a small SaaS should answer one covers that side.
What areas should an AWS infrastructure audit cover?
An AWS infrastructure audit for a small SaaS should cover six areas: releases, backups and recovery, access, network and data protection, monitoring, and cost.
These map closely to the pillars of the AWS Well-Architected Framework, which are operational excellence, security, reliability, performance efficiency, cost optimisation and sustainability.
| Area | What gets checked | A common finding |
|---|---|---|
| Releases | Pipeline, tests, health checks, rollback | Deploys are rolled back by hand, if at all |
| Backups and recovery | Backup settings, retention, restore tests | Backups exist but no restore has ever been tested |
| Access | Root user, MFA, IAM roles, long-lived keys | CI uses stored access keys with admin rights |
| Network and data | Security groups, public resources, encryption | A database or bucket reachable from the internet |
| Monitoring | Uptime, errors, logs, alert routing | Alerts go to one person's inbox |
| Cost | Idle and oversized resources, commitments | Non-production running around the clock |
What should you get at the end of an AWS infrastructure audit?
At the end of an AWS infrastructure audit you should get a short ranked list of fixes, each with the risk it removes and rough effort, plus a call to walk through it.
A long report with no priorities is hard to act on for a team that also has to ship features.
- A summary score or rating per area, so you can track progress.
- The top five or so fixes, in order of risk.
- Evidence for each finding, such as the resource or setting involved.
- What is already good, so you do not change things that work.
- A walkthrough call with the person who did the review.
How much does an AWS infrastructure audit cost?
An AWS infrastructure audit costs nothing if you run the AWS Well-Architected Tool yourself, since AWS charges nothing extra for the tool, while a review by an outside team costs whatever that team charges, ideally a fixed price agreed up front.
Formal compliance audits are a different product and cost more.
The free route works if someone on your team has the time and experience to answer the questions honestly and judge the results. A paid review adds an outside view and a fix list ranked by someone who has seen many setups. Keel, for example, publishes a fixed price of £1,250 for its read-only assessment, credited in full if you sign up within 30 days.
What access does an AWS infrastructure audit need?
An AWS infrastructure audit should need only read-only access to your AWS account and read access to the repositories that define your infrastructure and pipeline.
Anyone asking for admin rights to review your setup is asking for more than the job needs.
- A read-only IAM role the reviewer assumes, which you can remove afterwards.
- Read access to the infrastructure code and CI configuration.
- A short conversation about how releases, backups and on-call work in practice.
- No access to customer data. The review looks at settings, not records.
When should a small SaaS get an AWS infrastructure audit?
A small SaaS should get an AWS infrastructure audit before a moment when weaknesses become expensive: a big launch, a fundraise, an enterprise deal, or right after an outage.
It is also worth doing once when the person who built the setup is leaving.
Backups deserve particular attention in any audit. Veeam's 2026 research found that only 28% of organisations hit by ransomware fully restored their data, and an audit is often the first time anyone checks whether a restore works. See how to prove your AWS backups restore.
Can you audit your own AWS infrastructure?
Yes, you can audit your own AWS infrastructure using the free AWS Well-Architected Tool and the self-checks in guides like this one, as long as you are honest about the answers.
The usual weakness of a self-audit is not knowledge but blind spots: people rarely flag the shortcuts they took themselves.
A practical approach is to run the self-check below, fix the obvious gaps, and get an outside review when a launch, deal or investor makes the stakes higher. The guide on whether DevOps as a service is worth it covers what to do with the findings if you lack time to fix them.
Would your AWS setup pass a quick infrastructure audit?
Tick each item that is true for your account today.
- Every deploy runs tests and rolls back automatically on failure.
- A database restore test passed in the last 90 days.
- The root user has MFA and no access keys.
- No database or bucket is reachable from the public internet by accident.
- Alerts for downtime, errors and spend reach more than one person.
- Someone reviewed the bill by service in the last month.
Decision point: Five or six ticks: an audit will mostly confirm what you know. Three or fewer: an audit, self-run or paid, will likely find risks worth fixing this quarter.
Sources
- AWS, AWS Well-Architected Tool FAQs (checked October 2026)
- AWS, Shared Responsibility Model (checked October 2026)
- AWS Well-Architected, Perform data backup validation and recovery testing (checked October 2026)
- AWS docs, Analyzing your costs with AWS Cost Explorer (checked October 2026)
- Veeam, Data Trust and Resilience Report 2026 (2026)
Keep reading
Keel is an AWS release and recovery operations service for SaaS teams of 3 to 15 developers, provided by Bargainn Services Ltd, a UK company registered at Companies House (16497746). Registered office: 3rd Floor, 86-90 Paul Street, London, EC2A 4NE, United Kingdom.