SaaS Security Best Practices: A Checklist for the Apps You Use and Build
by GainHQ Editorial Team | Jan 19, 2026 | Software Development Insights | Editorial policy
Table of Contents
- What Is SaaS Security?
- Who Is Responsible for SaaS Security?
- Why SaaS Security Needs Attention Now
- SaaS Security Best Practices for the Apps You Use
- SaaS Security Assessment Checklist: What to Check Before You Buy
- SaaS Security Standards and Frameworks
- How to Write a SaaS Security Policy
- SaaS Security Best Practices If You Build SaaS
- SSPM vs CASB vs SaaS Discovery Tools
- How GainHQ Can Help
Quick answer: SaaS security means protecting the data, identities and settings inside the cloud applications your business uses, and, if you build SaaS, the product you ship. The provider secures the platform, but under the shared responsibility model the customer always keeps responsibility for its data, configuration and user identities (Microsoft Learn). The controls that matter most are phishing-resistant MFA, least-privilege access, hardened configuration, vetting of vendors and integrations, logging, and a tested response plan.
What Is SaaS Security?
SaaS security is the set of policies, controls and tools that keep software-as-a-service applications and the data in them safe from unauthorised access, misuse and loss. It has two sides:
- Securing SaaS you use: the CRM, email, file storage, HR and finance tools your teams log into every day. Most of the work here is identity, access, configuration and vendor risk.
- Securing SaaS you build: the product your company sells. Here the work starts in design and development, long before a customer signs in.
Most of this guide covers the first side, because that is where every organisation has to act. A separate section below covers teams that build SaaS products.
Who Is Responsible for SaaS Security?
Responsibility is shared, and in SaaS the customer still owns more than many teams assume. Microsoft's shared responsibility model sets out the split for SaaS as follows (Microsoft Learn):
| Area | Who is responsible in SaaS |
|---|---|
| Your data | Customer |
| Configuration and settings | Customer |
| Identities and user accounts | Customer |
| Client devices | Shared |
| Applications | Shared |
| Network controls and operating system | Provider |
| Physical hosts, network and data centre | Provider |
In practice, a vendor can run a secure platform and you can still lose data through a weak password, an over-privileged admin or a sharing setting left open. Those are on your side of the line.
Why SaaS Security Needs Attention Now
Attackers increasingly come in through suppliers, stolen logins and unpatched software, which are exactly the weak points in a SaaS-heavy stack.
- Verizon's 2025 Data Breach Investigations Report found that third-party involvement in breaches doubled to 30%, and that credential abuse was the most common initial attack vector at 22% (Verizon, 2025).
- The 2026 edition reports that 31% of breaches now start with exploitation of software vulnerabilities, overtaking stolen credentials as the top way in (Verizon DBIR 2026).
For SaaS users, that means vendor risk and identity matter as much as your own network. For SaaS builders, it means patching speed and secure code are now front-line defences.
SaaS Security Best Practices for the Apps You Use
Start with identity and access, because most of what you control in SaaS is who can sign in and what they can do. Then harden settings, control what connects to your apps, and make sure you can see and respond to problems.
1. Enforce phishing-resistant MFA
Turn on multi-factor authentication for every user, starting with administrators, and prefer phishing-resistant methods such as security keys or passkeys over SMS codes. CISA's Secure by Design pledge makes MFA adoption its first goal for software makers (CISA), so choose vendors that support it well.
2. Centralise sign-in with SSO
Connect SaaS apps to one identity provider with single sign-on. You get one place to enforce MFA and password policy, and one switch to cut access when someone leaves.
3. Apply least-privilege access
Give each person the smallest role that lets them do their job, keep admin roles to a few named people, and review access every quarter. Broken access control is ranked first in the OWASP Top 10:2025.
4. Offboard fast and remove dormant accounts
Disable accounts on the day people leave, and regularly remove accounts nobody has used. Unused accounts with valid credentials are easy targets, and shared logins make it impossible to know who did what.
5. Harden configuration and keep it hardened
Review sharing, guest access, public links, data retention and audit logging after setup and after every major change. Security misconfiguration is second in the OWASP Top 10:2025, and configuration is always the customer's job in SaaS.
6. Control third-party integrations and OAuth apps
Every connected app, browser extension and OAuth grant is another route to your data. Keep an approved list, review the permissions each integration requests, and revoke what is no longer used.
7. Find and manage shadow SaaS
Teams sign up for tools without telling IT. Use your identity provider, expense data or a discovery tool to keep an inventory of the SaaS in use, then bring important tools under SSO or retire them.
8. Protect and back up sensitive data
Classify the data you store in each app, use the encryption and data loss prevention features your vendors offer, and check whether you need your own backups. Being able to restore data matters as much as encrypting it.
9. Collect logs and monitor activity
Make sure audit logs are switched on, kept long enough and sent somewhere your team actually reviews them. Providing evidence of intrusions through logs is one of CISA's pledge goals (CISA), so ask vendors what logs they give you.
10. Train people and rehearse incidents
Teach staff to spot phishing and consent-grant tricks, and include SaaS accounts in your incident response plan: who revokes sessions, rotates keys and contacts the vendor.
SaaS Security Assessment Checklist: What to Check Before You Buy
A SaaS security assessment checks whether a vendor's controls, and your own setup of its product, meet your requirements before you trust it with data. Use this checklist when you approve a new tool and when you renew.
| Area | What to check or ask for | Source |
|---|---|---|
| Shared responsibility | Which controls the vendor runs, and which data, configuration and identity tasks stay with you | Microsoft Learn |
| Independent attestation | A current SOC 2 report and which trust services categories it covers | AICPA |
| Certification scope | An ISO/IEC 27001:2022 certificate, and whether its scope includes the service you are buying | ISO |
| Cloud controls | A completed CSA questionnaire (CAIQ) or a CSA STAR registry entry | CSA |
| Authentication | MFA for all users, phishing-resistant options and SSO support | CISA pledge goal 1 |
| Default credentials | No shared or default passwords on any account or integration | CISA pledge goal 2 |
| Patching | How fast the vendor fixes vulnerabilities and tells customers | CISA pledge goal 4 |
| Vulnerability disclosure | A published vulnerability disclosure policy with a reporting channel | CISA pledge goal 5 |
| Logging | Which audit logs you get, how long they are kept and how you export them | CISA pledge goal 7 |
| Third parties | Sub-processors and integrations that touch your data, and how they are vetted | Verizon 2025 (third-party involvement at 30%) |
Also ask about data location, breach notification timelines, backup and restore, and how you get your data back when you leave. If you are assessing a software company or product more deeply, see our guide to technical due diligence.
SaaS Security Standards and Frameworks
No single law or standard defines SaaS security, so most organisations use a recognised framework to structure their controls and ask vendors for evidence against one.
| Framework | What it is | How it helps with SaaS |
|---|---|---|
| SOC 2 (AICPA) | An auditor's report on a service organisation's controls across five categories: security, availability, processing integrity, confidentiality and privacy | The most common evidence SaaS vendors provide to customers |
| ISO/IEC 27001:2022 | Requirements for an information security management system | Shows a vendor runs security as a managed, audited system |
| NIST CSF 2.0 | Six functions: Govern, Identify, Protect, Detect, Respond and Recover | A simple structure for your own SaaS security program and policy |
| CSA Cloud Controls Matrix v4.1 | 197 cloud control objectives in 17 domains, with the CAIQ questionnaire built in | A ready-made question set for assessing cloud and SaaS providers |
| HIPAA | US rules for protecting health information | Applies when SaaS handles health data; see HIPAA-compliant software development |
How to Write a SaaS Security Policy
A SaaS security policy sets the rules for how your organisation buys, configures and uses cloud applications. Keep it short enough that people read it, and cover these points:
- Approval: who can approve a new SaaS tool, and the assessment it must pass.
- Identity: SSO and MFA required, admin roles limited and reviewed.
- Data: which data classes may go into which tools.
- Integrations: who can connect apps and grant OAuth access.
- Monitoring and response: which logs are kept and who acts on alerts.
- Offboarding: how access is removed from people and from retired tools.
Mapping the policy to the six NIST CSF 2.0 functions, with Govern first, makes gaps easy to spot (NIST CSF 2.0).
SaaS Security Best Practices If You Build SaaS
If you build a SaaS product, your customers rely on you for the provider side of the shared responsibility model, so security has to be designed in rather than added later. CISA and its US and international partners frame this as three principles: take ownership of customer security outcomes, embrace radical transparency and accountability, and lead from the top (CISA). This checklist turns that into practice:
| Control | What to do | Source |
|---|---|---|
| Secure by default | Ship MFA switched on and remove default passwords | CISA pledge goals 1–2 |
| Eliminate vulnerability classes | Use frameworks and patterns that prevent injection, cross-site scripting and similar flaws at scale | CISA pledge goal 3; OWASP Top 10:2025 |
| Secure development lifecycle | Adopt the four SSDF practice groups: prepare the organisation, protect the software, produce well-secured software, respond to vulnerabilities | NIST SP 800-218 (SSDF) |
| Open-source dependencies | Keep an inventory of components and patch known vulnerabilities; 98% of audited codebases contain open source and 87% have at least one vulnerability | Black Duck OSSRA 2026; OWASP A03 Software Supply Chain Failures |
| Fast patching | Prioritise vulnerability management, now the top breach entry point at 31% | Verizon DBIR 2026 |
| Tenant isolation | Derive tenant context from the verified identity and scope data, caches, files and logs to each tenant (see multi-tenant architecture) | OWASP Multi-Tenant Cheat Sheet |
| API security | Authenticate and authorise every API call and test APIs regularly (see API security best practices and API security testing) | OWASP Top 10:2025 (A01, A07) |
| Disclosure and transparency | Publish a vulnerability disclosure policy and issue CVEs for serious flaws | CISA pledge goals 5–6 |
| Customer-facing logs | Give customers audit logs that show evidence of intrusions | CISA pledge goal 7; OWASP A09 |
For the architecture behind these controls, see SaaS architecture best practices, and for where security fits in the delivery process, B2B SaaS product development.
SSPM vs CASB vs SaaS Discovery Tools
These tools solve different problems, and many teams need only one or two of them. Start with the gap you have, not the category.
| Tool | What it does | Use it when |
|---|---|---|
| SaaS security posture management (SSPM) | Checks the security settings of your sanctioned SaaS apps and flags risky configurations | You run several critical SaaS apps and can't review settings by hand |
| Cloud access security broker (CASB) | Sits between users and cloud services to apply policy on access and data movement | You need to control how data flows into and out of SaaS |
| SaaS discovery and management | Finds the SaaS apps and integrations people actually use | You don't know your full SaaS inventory |
How GainHQ Can Help
GainHQ is a custom software development company with more than 15 years of experience, over 13 products built and a team of more than 30 people. We apply the practices above when we design and build SaaS products, from secure defaults to tenant isolation and API security. See our software development services.
Frequently asked questions
What is SaaS security?
SaaS security is the set of controls that protect cloud applications and the data in them from unauthorised access, misuse and loss. For SaaS you use, it covers identity, access, configuration and vendor risk. For SaaS you build, it also covers secure design and development.
Who is responsible for security in SaaS?
The provider and the customer share it. In Microsoft's model for SaaS, the provider runs the network, operating systems and physical infrastructure, while the customer keeps responsibility for its data, configuration and identities (Microsoft Learn).
What should a SaaS security assessment include?
Check independent evidence (a SOC 2 report, ISO/IEC 27001 certificate scope or CSA STAR entry), authentication and SSO support, patching and vulnerability disclosure, logging, sub-processors and integrations, data location, breach notification and data return when you leave.
What is a SaaS security checklist?
It is a list of controls to confirm for each SaaS app, such as MFA, least-privilege roles, hardened sharing settings, approved integrations, audit logging, backups and offboarding. Use one checklist for new tools and repeat it at renewal.
Which compliance standards apply to SaaS security?
The most common are SOC 2 and ISO/IEC 27001:2022 for vendor assurance, NIST CSF 2.0 for structuring a security program, and the CSA Cloud Controls Matrix for assessing cloud providers. Sector rules such as HIPAA apply when you handle that type of data.
How often should you review SaaS security?
No standard sets one frequency. A practical rhythm is to assess each app before purchase and at renewal, review admin access quarterly, and recheck settings after major changes or a security incident.
What is the difference between SSPM and CASB?
SSPM checks the security settings inside the SaaS apps you have approved. A CASB sits between users and cloud services to enforce policy on access and data movement. SSPM fixes misconfiguration; a CASB controls use.
What security practices matter most if you build a SaaS product?
Ship secure defaults such as MFA and no default passwords, follow a secure development framework like NIST SSDF, prevent common flaws listed in the OWASP Top 10:2025, manage open-source dependencies, isolate tenants and patch quickly. CISA's Secure by Design pledge sets out seven concrete goals (CISA).