Security shouldn’t feel like a chore. This is a basic checklist that any SaaS CTO (and anyone else) can use to harden their security. Select your startup stage and use these rules to improve your security.
Accustom everyone to security practicesSeed
Humans are often the weakest link in the security of a startup. By explaining how an attacker could infiltrate your company, you will increase your employees’ awareness and thus minimize the chance of them falling for basic traps.
Accustom your team to locking their computersSeed
Your office may be secured, but you will eventually have to receive external people for a party or a meeting. Locking all the machines is a great habit have. If you get into the habit of locking your machine at the office, you’ll be unlikely to forget to also do it in a Starbucks or at a meetup.
Do not share accountsSeries A
Sharing accounts makes it hard to understand who is using the service. You should avoid it as much as possible. Keep a log of shared accounts and make sure you change passwords if an employee leaves.
Encrypt all employee laptops & phonesSeed
By encrypting all laptops, you protect both your company’s assets and your employees’ private files. Encryption of computers or phones used for work should be encrypted during the onboarding process. It will protect against both malicious activities and accidents (e.g. an employee’s child accidentally wiping a mailbox).
Follow an onboarding / offboarding checklistSeed
This checklist should contain a list of all the steps you need to enforce when an employee, contractor, intern, etc., joins your company. A similar list should also be used when someone is leaving your team to ensure that they no longer have access to any of your company’s resources.
Require 2FA in your servicesSeed
Use a password manager to ensure you only use strong passwordsSeed
Use centralized account managementSeries A
A centralized place with all user authorizations is the best way not to forget anything once you need to update a user profile (e.g., if an internship came to its end).
Set up your own custom SAML application on Google apps
Backup, then backup againSeed
Backup all your critical assets. Ensure that you attempt to restore your backups frequently so you can guarantee that they’re working as intended. S3 is a very cheap and effective way to backup your assets:
Centralize and archive your logs and make them meaningfulSeries A
Logs are necessary to trace what happened after an incident. Understanding how an attack was performed is key to make sure it never happens again. Many solutions exist to gather your logs. The more advanced solutions will help you cut through the noise and find the important data. One advice is to always make sure the system time on your machines is in sync. This allows you to cross-correlate logs.
Check your website’s basic securitySeed
Isolate assets at the network levelSeed
Only your public APIs should be exposed to the Internet. You should isolate your networks to prevent any unauthorized accesses to your database. This will prevent attackers from connecting to it and attempting to crack the password, or exploit vulnerabilities.
Keep a list of your serversSeed
This is built-in if you are using a cloud service and all your machines are registered or spawned through it. Otherwise, you will need to create and maintain a list of your assets (servers, network devices, etc.). Review it on a regular basis to determine if you still need them, keep them up to date, and ensure that they benefit from your latest deployments.
Keep your OS up to dateSeed
You should download all of your OS’s security updates and regularly update your machines. As a young startup, you should probably use a PaaS that will handle these updates for you (Heroku, AWS Beanstalk, etc.).
Know how to redeploy infrastructure from scratchPost-Series A
Protect your application from DDoS attacksSeries A
Restrict internal services by IP addressesSeed
Connections to your infra and non-public properties (hosted CIs, Admin interfaces, databases etc.) should only be accessible through a bounce host (VPC, VPN etc.)
Use SSL certificates to secure people using your websiteSeed
Encrypting communications is not only about privacy, but also about your users’ safety, since it will prevent most attempts at tempering with the data they receive.
Use Let’s Encrypt to get a free SSl certificate.
You can also choose your own custom certificate (which may allow you to get a beautiful green bar if you pay for the extra “Extended Validation”):
Watch for unusual patterns in your metricsSeries A
Be honest and transparent about any data you collectSeed
In the case of a breach, the attackers may disclose the data they gathered. Your customers need to be aware of what data you’re storing.
Do not share WiFiSeed
Sharing WiFi networks with guests or neighbors may give them the opportunity to gather information on your network, and allow them to access resources protected by source IP. Use an isolated and dedicated guest WiFi network. Set up a calendar reminder to change the password every two months, since this password is shared.
Ensure your domain names are securedSeed
Domain names should be renewed regularly. If you have bought one from a third party, you should also make sure that the authoritative name server is your own.
Have a public security policySeries A
This is a page on your corporate website describing how you plan to respond to external bug reports. You should advertise that you support responsible disclosure. Keep in mind that most of the reports that you receive aren’t relevant. Don’t freak out if you receive so called “critical disclosures”
Have a security incident response planPost-Series A
Having an incident response plan will allow you to communicate in the best and fastest manner about an incident.
Have an internal security policyPost-Series A
The internal security policy is a short document stating the security requirements in your company. It also defines the different point of contacts for security.
Make an inventory of your company’s assetsPost-Series A
Being aware of your company’s assets is an essential first step when starting to scale your security organization.
Make sure all your critical services are securedSeed
As a company you rely on a multitude of services like Google Apps, Slack, Wordpress etc. Don’t settle for the security defaults of these tools. They can be an easy attack vector if not configured properly. Also make sure you update them regularly.
Make sure your email is securedSeed
Email is an important attack vector for cyber attacks. Be aware of phishing attacks! At a later stage, you can also implement the DMARC protocol to secure your emails and domains.
Set up a bug bounty programPost-Series A
A bug bounty program will allow external hackers to report vulnerabilities. Most of the bug bounties program allow you to offer rewards for bugs found. A lot of the reports won’t be valuable and you need security aware people inside your development teams to evaluate the bugs you receive. These programs are good additions to other security initiatives and can’t by no means be considered as enough.
Places to start:
Take special care of your non-tech employeesSeries A
Non-tech employees are less used to technical trickery and can be deceived more easily than others. This can open the door to ransomware or confidentiality issues. They should be trained and empowered to be distrustful and to preserve the company’s assets.
Don’t store credit card informationSeed
Use third-party services to store credit card information to avoid having to manage and protect them.
Perform security auditsPost-Series A
Security audits take an external point of view on your infrastructure and products. They are often required when selling to enterprise or for compliance reasons. But if you can afford it, perform penetration tests regularly. Don’t freak out when receiving your first report, it’s often difficult to make them actionable. Keep your company stage in mind. Not every vulnerability identified should be fixed right away.
Monitor your dependenciesSeed
Run it unprivilegedSeed
If an attacker successfully attacks your application, having it running as a restricted-user will make it harder for the attacker to take over the host and/or to bounce to other services. Privileged users are root on Unix systems, and Administrator or System on Windows systems.
Use a real-time monitoring & protection toolSeries A
A real-time monitoring and protection tool like Sqreen allows you to integrate security into your app quickly. Sqreen will enable you to get full visibility on your security, prevent data breaches, protect your customers, and stop business logic attacks. Customize your application’s response to attacks (block attack, log stack trace etc. ) and get notified when something important happens.
Encourage your users to use 2FASeries A
As you get higher profile customers, you will be required to implement stronger security practices. This includes offering them 2-factor authentication (2FA), role-based account management, etc. It is recommended to use a third-party solution to manage authentications on your app.
Enforce a password policySeed
Your customer data will be much harder to steal if you require complex passwords: mixed case, special characters, minimum length, etc.
Monitor your users’ suspicious activitiesSeries A
For SaaS apps, user’s are often the most significant source of critical cyber attacks. This is simply because the attack surface for logged-in users is much broader. You need to make sure you monitor your users for irregular activities like: suspicious routes access, spikes of HTTP errors, suspicious geolocations, TOR connections, etc. Being able to identify and block an attacker early often avoids a lot of issues down the road.
Enforce a secure code review checklistSeed
Security should always be kept in mind when developing software. But enforcing an ever increasing list of security checks in code reviews is key to avoid a lot of vulnerabilities. The checks should be different depending on where the code is. Dealing with user entry is one thing, dealing with business structures is another: the concerns are related to the context. In addition to common sense, keep in mind the typical security flaws, like the OWASP Top 10.
Keep secrets away from codeSeed
Never commit secrets in your code. They should be handled separately to prevent them accidentally being shared or exposed. This allows a clear separation between your environments (typically development, staging, and production).
Maintain a backlog of security concerns in your issue tracking toolSeed
Every developer should contribute to maintaining a list of security issues for the backlog. Making vulnerabilities available to the rest of the team will increase the security awareness in the company.
Never do cryptography yourselfSeed
Always rely on existing mechanisms, libraries and tools. Cryptography is an expertise. Building your implementations, or using flags and options you don’t fully understand will expose you to major risks.
Perform security oriented test sessionsSeries A
Once in a while, the entire technical team should sit together and attack the whole organization. This is a great time to test for account isolation, token unicity, unauthenticated paths, etc. You will massively rely on your browser’s web console, curl, and 3rd party tools such as OWASP ZAP, Burp, Nmap, or OpenVAS. Keep in mind that these tools generate a lot of false positives. You can cut through false positives and ease the remediation process by installing Sqreen in monitoring mode on your app.
Use a secure development life cyclePost-Series A
The secure development lifecycle is a process that helps tackle security issues at the beginning of a project. While rarely used as is, it provides useful insights at all stages of the project, from the specification to the release. It will allow you to enforce good practices at every step of the project life.
Use a static security code analysis toolSeed
Static code analysis tools can quickly overwhelm you with a lot of meaningless false-positives. But switching-on security-focused tools can help you discover vulnerabilities inside your code and most importantly increase the security awareness inside your team. Integrate these tools into your workflow to reduce friction. Post-commit checks that automatically comment where code reviews are performed are ideal.