Cyber Security Advice for Small Business

Cyber Security Advice for Small Business

These cyber security for small business recommendations focus on the conventional Small and Medium Enterprise organizations. This text does not cover startup specifics or the application security needs of software development companies. This is just a checklist of the most crucial cyber security measures every small business owner can and must implement.

This is how you can learn if these cybersecurity tips can be applied to your organization. If you have (1) a person responsible for cyber security full time and (2) a separate cyber security budget, most probably you have to show this to your CISO and check if anything from the list is missing. If you lack any of the two – this guide is just for you.

Use password managers

As long as we use passwords, hackers will attempt to crack them to gain access to valid user accounts. According to Verizon’s 2021 Data Breach Investigation Report, 61% of all data breaches involved user credentials.

There are two ways to lower this risk: using unguessable passwords and adding a second authentication factor. Both are simple and ea; however, they require some time to get used to. Good news: after the transition is over, users spend less time to authenticate, as they must not type their passwords anymore: they will fill them in via the autotype function or accompanying browser extensions.

A password manager allows your employees to remember only two passwords – forever. One to enter their workplaces, such as a workstation or a laptop, and another to unlock the password safe in the password manager software, where all other passwords are securely stored.

Simplicity and high usability of password managers, together with the fact that the users won’t have to memorize passwords from now on, allows you to demand creating long, random passwords from your employees. And as a positive side effect, cyber security becomes an enabler of a more productive workplace.

Use two-factor authentication

Two-factor authentication is as easy to implement as password managers, but it has a powerful advantage: it can be technically enforced. Password managers allow users to extend and randomize their passwords, but they could still ignore the opportunity. Two-factor authentication, instead, can be configured to require users to present a temporary code or a physical token.

Virtually all modern applications accept some form of the second factor. The most primitive one is a temporary password sent in an SMS to a user-owned mobile phone. This one is the least secure and must be avoided. We recommend using mobile app-generated one-time passwords (OTP) for all user accounts in the organization and hardware tokens for users with high privileges or accounts in critical systems.

Install security updates automatically

Software with known vulnerabilities remains one of the most looming cybersecurity threats. Hackers rarely use yet unknown security bugs, also known as zero-days, to attack regular targets. Getting rid of known security weaknesses is easy as most modern applications have built-in automatic software updates.

It is true, once in a while, even the most prominent software vendors screw up their update cycles and release poorly tested patches that brick the systems or send them into infinite reboot cycles. But it is not an excuse to turn off automatic updates; it is the reason to test updates before deploying them across the organization, of course, if you have the time. And most importantly, it is the reason always to have backups; more about it next.

Do backups and backups of backups

Backups do not come first on our list, but they are the most important safeguard for your sensitive data. A full-stop cyber security event, such as a ransomware attack, can be devastating if you do not regularly backup your systems and data. In the case of compromise, your business mission will directly depend on the frequency and quality of your backup copies.

It is crucial to make backup copies at an independent facility, and the cloud could be the right choice for an SME. This approach is never free and is charged per storage used, number of users, number of machines, or a combination of the three.

If you can automate backups on your own – good for you, go for it. But no matter if you do it yourself or use a service provider, make sure you have multiple locations for backup copies, check and restore them regularly, and encrypt them before moving data to the cloud.

The most important thing about the backups that many companies get wrong is recovery tests. It is true: modern backup solutions rarely fail, but when they do, you are basically in a situation when your backups are screwed, and so is your business. Our advice is to restore regularly to verify the integrity of restored data and the functionality of restored systems and applications to ensure that you can trust your backup solution.

Scan yourself regularly and reduce the attack surface

To protect your business, you must know how hackers can attack it. Hacking means exploiting systems and applications in a way that their creators and owners do not intend. Malicious hackers aim at harming you and your business. So, first of all, you always need to know what systems you have, where they are, and what services they run. 

Many small companies get compromised via a very dumb attack vector: they access their systems remotely, so they open the related services, such as RDP, VNC, or SSH, to the internet. Hackers often guess passwords to these services by trying the top 100 popular ones found in prior password leaks. Then, they just use the compromised user’s account to get in, encrypt all the data, and demand ransom.

Enabling strong authentication, enforcing random passwords, and using 2FA prevent that, but if the service has known security bugs, it doesn’t matter. Hiding behind a firewall and automatically installing security patches reduces the risk. However, there is still a possibility you could get hacked, and you can eliminate it by removing potentially vulnerable unused services. Yes, you may not even know they exist, and removing them would not affect you at all.

The question is, how can you find them? This one is easy: regularly run discovery scans to maintain an inventory of your “assessment scope,” as we call it. Then run network scans to learn what services are enabled on your hosts and search for known security flaws in them. Not manually, of course; there are tools for that.

When you find a vulnerable service and think you need it, update it to the latest secure version. But even if the service is “clean,” ask yourself: do you need it exposed on the internet? If not, shut it down and disable it, or close network access to it on the firewall.

  • Use SpiderFoot to learn about your company’s internet presence
  • Run NMap (command-line) or (web-based) to scan your network hosts for open ports and known vulnerabilities
  • Use Nuclei or Nessus to do regular vulnerability scans

Hide your websites behind a cloud firewall

A firewall is a well-known concept: put a device at the border of your network to segregate between the trusted internal private network and untrusted external public internet. A firewall is an excellent place to apply your network security policy, e.g., who can access what. It could also be intelligent enough to defend your organization from common attacks at the network perimeter.

In the modern world, though, the concept of the perimeter is out of date. More and more organizations eliminate the “internal vs. external network” mentality and apply the protection “at the edges” regardless of where the connection is coming from.

The easiest way to protect all your web servers from common attacks is to use Cloudflare edge protection. Cloudflare is a service that puts a combination of a Web Application Firewall and an anti-DDoS solution in front of your web service and takes the first hit whenever someone tries to attack you. 

To use Cloudflare, you first have to migrate your DNS record to its facilities, which is as easy as copying and pasting. If you still feel not confident enough, there are guides on doing it safely and without service interruptions.

Next, you will need to configure the Cloudflare parameters to the level of security you want to get. Good news: all crucial functions, such as DDoS protection, basic WAF, and SSL enforcement, are available for free.

Finally, you will need to configure the hosts you have put behind Cloudflare to allow network connections only from Cloudflare itself. This way, no one will be able to attack you directly by circumventing the protection. It is simple as configuring a cron job on the Linux server; see below the link to a tool that does just that.

Warning! Make sure that you do not lose remote access, or else you will need to waste time restoring it. Good news: since recently, Cloudflare has allowed configuring SSH access to your hosts under its protection without a need to expose SSH ports publicly.

Check your configurations in the cloud and on the ground

Security configuration errors are another reason many companies get hacked: people make mistakes, and system administrators are people. According to OWASP Top 10 project, security misconfiguration is the fifth most significant application risk. Good news: software vendors and cloud service providers share recommendations on how you can harden your systems and apps. Even better news: audits of these security settings can be easily automated.

If you have a Microsoft Active Directory managed on-the-ground IT infrastructure, the best security health check would be running Ping Castle and fixing its findings. You don’t even need high domain privileges; however, running it with elevated permissions would produce more results. Read its recommendations carefully, apply them with caution, and in a few days, your domain will become closed for ordinary attacks.

Suppose you have a lot or all of your systems in the cloud, Scout Suite for the rescue. You can view it as Ping Castle for cloud environments, but it is strongly biased towards AWS. It still produces valuable results for GCP and Azure, but AWS is where it shines. Run it regularly, eliminate its findings, and your cloud infrastructure will become more secure after each iteration.

Centralize logging

When a security incident requires investigation, logs are critical. A security incident may mean different things: from unauthorized individuals attempting to connect to the Wi-Fi access point to a full-scale organization compromise, data breach, or ransomware attack. In any case, without logs, you are blind to what has exactly happened and cannot learn how to prevent it in the future.

One critical thing about the logs: same as backups, they must be stored somewhere far from the regular systems and applications. A catastrophic event must not render them unusable, and malicious hackers must not be able to remove the traces of their presence. Another important thing: logs must be stored centrally to correlate the events in different systems to make sense of what is going on in the organization. And, of course, logs must be reliably backed up.

When your logs are collected and stored, you can start analyzing them regularly or even in real-time. Configure alerts to be notified about a possible security incident on time. However, proceed with caution as it might be overkill for a small organization. A viable alternative would be configuring canary tokens as described in the next section.

Use canary tokens

The concept behind canaries is simple: hackers do certain things during cyberattacks, and you can get alerts if these things are done to your systems. Canary tokens stem from the ancient art of network “honey pots” and are an example of what is now called cyber deception technologies. SMEs can use some canary tokens for free.

Canaries are pieces of data, such as folders, documents, or API tokens, that attract hackers’ attention. Say, a passwords.xlsx spreadsheet on a file server or a false password hardcoded into your application source code. Creating canaries and spreading them throughout your infrastructure is an artistic work, but do not get carried away: too many canaries will generate a lot of noise in the company’s normal operations.

Use basic endpoint security software

Let us not forget about the essential protection of the employees’ workstations and laptops, where the actual work is happening. A built-in firewall must be enabled on all computers, and basic antivirus software should be set to real-time protection and daily full scans.

Besides that, security configuration settings must be regularly checked, and OS Query is an excellent open-source project that allows you to collect this data from remote computers using SQL-style queries. If you don’t have time to figure it out yourself, give Kolide a chance: it is a successful attempt to provide OS Query as SaaS.

Hack yourself first

Our recommendations would be incomplete without how BSG as a company could protect you from cyber threats. The two affordable ways to benefit from good relations with hackers available to virtually every company are penetration testing and bug bounties.

Pentesting is a controlled simulation of a realistic cyber attack on your organization. It may involve network- and application-level attacks, as well as social engineering probes of your employees’ cyber security awareness. Pentests are conducted by qualified experts from respectful companies in the cyber security industry, and pentesting services are delivered according to B2B contracts.

On the other hand, a bug bounty program is an organization’s public announcement that it welcomes hackers to assess its security, report findings, and get rewarded. Different ways to organize bug bounty programs exist, from as large as an in-house operation in huge software product companies to as small as just a properly formed security.txt file with your security contacts on the main company website.

Pentesting and bug bounties do not replace but rather complement each other. They may look similar, but in reality, the only thing in common is hackers involved.

Conclusion and next steps

I hope this information proves valuable to your business organization. There are other ways to improve your security, but this is the absolute minimum of what is needed in every company. What could be the following steps, though?

Once you have completed the above checklist and want to dedicate time and resources to progress forward, we recommend focusing on these activities.

  1. Establish a privacy-centric culture in the company. Encourage everyone to use a proven VPN service or install a server and invite them to use it. Algo or Outline is a great start. This will improve the privacy of everyone in the company and allow you to apply IP-based network rules more efficiently, as you will now know where your colleagues may connect from.
  2. For macOS users, there are many free or affordable juicy security nuggets out there. Little Snitch and other tools from its family of products is a fantastic way to control where your client-side applications connect to and decide whether they should have this ability. Objective See is another source of macOS security tools, and we use most of them ourselves.
  3. Once ready, try to put everything into a system. We are not talking about a formal Information Security Management System as in ISO/IEC 27001 standard, although you may get there one day too. We are thinking more about some basic methodology or framework that maps to a well-established set of standards. Once these questions start popping up in your head, get in touch, and we will figure out together what next.

Leave a Comment

%d bloggers like this: