What Is Zero Trust Security? A Plain-English Guide for Businesses

Zero trust is a security approach that trusts no user, device or network by default. Every request to reach an app or data is checked against who is asking, the health of their device and the context, and access is limited to what that person needs. It replaces the older model, where anything inside the company network was trusted.

The short answer

Zero trust is a way of running security that starts from one assumption: no user, device or network is trusted automatically. Every time someone tries to open an app, a file or a system, the request is checked. The check looks at who they are, whether their device is healthy and whether the request makes sense. Then they get only the access they need for that task.

The phrase people use is "never trust, always verify." NIST, which publishes the main U.S. reference on the topic, describes zero trust as moving defenses away from the network perimeter and toward the users, devices and resources themselves.

Why the old model stopped working

For decades, company security worked like a castle with a moat. A firewall guarded the edge of the office network. Anyone inside, whether at a desk or connected by VPN, was largely trusted.

That model breaks down for three reasons:

  • There is no clear "inside" anymore. Staff work from home and on the road, and company data lives in cloud apps like Microsoft 365, Google Workspace and Salesforce.
  • Attackers get in through people. A single phished password or infected laptop puts an attacker inside the moat, where they can move around freely.
  • Breaches spread. Once inside a flat, trusted network, ransomware can reach every server it can see.

Google was one of the first large companies to drop the idea of a trusted corporate network. Its BeyondCorp project, published in 2014, moved internal apps onto the internet behind strong identity and device checks. Today the same approach is standard guidance from NIST and from CISA, the U.S. Cybersecurity and Infrastructure Security Agency.

The core principles

Most definitions boil down to these ideas:

  • Verify every request. Access is decided per request, using identity, device health, location and behavior, not just a one-time login.
  • Least privilege. People and systems get the minimum access they need, for as long as they need it.
  • Assume breach. Design as if an attacker is already inside. Limit how far they can move and watch for unusual activity.
  • Protect resources, not network segments. Each application and data set gets its own protection. Being "on the network" doesn't count for much.
  • Keep checking. Trust isn't granted once at login. It is re-evaluated as conditions change.

The five pillars

CISA's Zero Trust Maturity Model groups the work into five areas. It is a useful checklist for seeing where you stand:

  • Identity: strong authentication for every user, ideally phishing-resistant multifactor authentication (MFA), plus single sign-on and regular access reviews.
  • Devices: knowing which laptops and phones connect, and checking that they are patched, encrypted and running endpoint protection before they get access.
  • Networks: splitting the network into smaller segments so a compromised device can't reach everything, and encrypting traffic.
  • Applications and workloads: putting apps behind identity-aware access instead of a VPN, and securing cloud workloads.
  • Data: classifying sensitive data, encrypting it, and controlling who can view, share or download it.

CISA also describes stages of maturity, from traditional to optimal. Zero trust is a direction you move in over time, not a box you tick.

What zero trust is not

  • Not a single product. Vendors sell "zero trust" tools, but no one product makes you zero trust. It is an approach built from identity, device, network and data controls working together.
  • Not "trust no one" for employees. It is about removing automatic trust from systems, not about suspecting your staff. Done well, it often makes logging in easier through single sign-on.
  • Not only for large enterprises. Small businesses can apply the same ideas with tools they may already pay for.

How to start: a practical path for small and mid-sized businesses

You don't need a multi-year program to get most of the benefit. A realistic order of work:

  • 1. Turn on MFA everywhere. Start with email, remote access and admin accounts. Prefer app-based or hardware-key MFA over text messages.
  • 2. Get an inventory. List your users, devices, cloud apps and where your sensitive data lives. You can't protect what you don't know about.
  • 3. Clean up access. Remove old accounts, take away admin rights people don't need, and review who can reach sensitive data.
  • 4. Check devices before granting access. Use your identity provider or device management tool to block devices that are unpatched or unmanaged.
  • 5. Replace broad VPN access. Move to app-by-app access, often called zero trust network access (ZTNA), so a remote user reaches only the apps they need.
  • 6. Segment the network. Separate servers, guest Wi-Fi and devices like printers and cameras from each other.
  • 7. Monitor and respond. Collect sign-in and endpoint alerts, and make sure someone reviews them, in-house or through a managed security provider.

NIST's National Cybersecurity Center of Excellence has published example builds using commercial products (SP 1800-35). They are a good reference when talking to vendors.

What zero trust costs

Costs vary widely, so this guide doesn't quote a single figure. The main factors are:

  • What you already own. Many Microsoft 365 and Google Workspace plans include MFA, conditional access and device management. The first steps may cost staff time more than new licenses.
  • Number of users and devices. Most identity, endpoint and ZTNA tools are priced per user or per device, per month.
  • Legacy systems. Older apps that can't use modern sign-in take extra work to put behind zero trust access.
  • Who does the work. An internal IT team, a managed service provider or a consultant running a project all price differently.

Our cybersecurity services cost guide explains how consultants and managed security providers usually charge.

Questions to ask a provider

  • Which of the five pillars will this project improve, and how will we measure progress?
  • What can we do with the licenses we already have before buying anything new?
  • How will this change the login experience for staff?
  • Who monitors the alerts once it's in place?

If you want help planning or rolling out zero trust, you can get free quotes from vetted security providers. For short definitions of the terms above, see our zero trust glossary entry.

Sources

  1. NIST SP 800-207: Zero Trust Architecture (NIST)
  2. Zero Trust Maturity Model (CISA)
  3. NIST SP 1800-35: Implementing a Zero Trust Architecture (NIST National Cybersecurity Center of Excellence)
  4. BeyondCorp: A New Approach to Enterprise Security (Google Research)

Drafted with AI assistance, then edited and fact-checked by a person. How we use AI.