Published:

Most small businesses treat phishing awareness as a one-off training session or an annual simulation exercise — and then wonder why they still get compromised. The problem is not that training has no value. The problem is treating it as the only layer of defence.
For broader context, see Cyber Security hub.
The NCSC's own guidance says phishing simulations and awareness training are the layer that is most often over-emphasised in phishing defences. That is a striking admission from the UK's national cyber security authority, and it carries weight: no training package can teach users to spot every phishing attempt, and asking staff to scrutinise every email in depth leaves them with no time for their actual work.
This article walks through what actually works — the NCSC's four-layer model, how Microsoft 365's built-in controls fit in, and why a report-button culture matters more than a click-rate metric.
The NCSC's Four Layers of Phishing Defence
The NCSC frames phishing defence as four distinct layers. Each layer addresses a different failure point, and the guidance is explicit: you need at least some mitigation from each layer, not a deep investment in just one.
Layer 1 — Make it difficult for attackers to reach your users
This layer is about stopping phishing emails before they reach the inbox. It covers three practical controls:
Email authentication. SPF, DKIM and DMARC records tell receiving mail servers how to treat emails that claim to come from your domain. When properly configured, they prevent attackers from spoofing your brand — which protects your customers and suppliers as much as it protects your own staff. Setting up DMARC is the single highest-impact action in this layer, and it is also the one most SMEs have not done.
Email filtering. Anti-spam and anti-phishing filtering should be enabled on your email platform by default for all users. If you use Microsoft 365, the anti-phishing and anti-malware protections are included in most business licences — they just need to be configured. Many businesses never touch those settings, leaving them at Microsoft's conservative defaults.
Reducing your digital footprint. Information on your website and social media gives attackers the material to craft convincing spear-phishing messages. A simple digital-footprint policy — what your site needs to show versus what it does not — is a low-effort control that makes targeted attacks harder to write.
Layer 2 — Help users identify and report suspected phishing messages
This is the layer where training sits. The NCSC does not say training is useless — it says training is one part of a wider system, and that simulations are the over-emphasised component.
What works here is simpler than most people think:
Ensure staff understand the nature of phishing threats, particularly in departments that handle finances or sensitive data.
Teach people to spot common features of phishing messages — mismatched sender addresses, urgency that does not match normal processes, links to unexpected domains.
Do not reprimand users who struggle. Staff who fear reprisals will not report mistakes promptly.
Create a positive reporting culture. A one-click "Report Phishing" button in the email client is more valuable than any training module. When someone clicks it, the message goes to a central mailbox, it is removed from the user's inbox, and the reporter gets an acknowledgement. That is the behaviour you are building: when in doubt, report, do not click.
The NCSC notes that metrics express an organisation's values. If you only measure how many people click on a phishing email, you incentivise silence. Measure how many people report it instead.
Layer 3 — Protect your organisation from the effects of undetected phishing emails
This layer accepts a simple fact: some phishing emails will always get through. The question is what stands between a click and a breach.
Multi-factor authentication. If a phishing email steals a password, MFA is what stands between that stolen credential and someone reading your mailbox. It is not a phishing-specific control, but it is the single most effective mitigation against phishing-led account compromise.
Endpoint protection. Modern endpoint detection and response (EDR) tools notice unusual process behaviour — a document opening a PowerShell session, for example. This is the layer that catches malware launching after a malicious attachment or link has been clicked.
Email filtering as a secondary layer. Even when SPF/DKIM/DMARC is not set up on the sender's domain, your own filtering catches a large percentage of threats. The NCSC case study it cites shows 1,800 phishing emails sent to an organisation; 1,750 were stopped by filtering.
Layer 4 — Respond quickly to incidents
The final layer is about speed and process when something gets through. This is where many SMEs have no plan at all.
A basic incident-response process for a small business needs:
Two known contact methods for the person who makes the call to isolate affected systems.
A simple flowchart: what to do in the first hour, who to call, what not to do (do not wipe drives).
A conference number or group chat where everyone involved can coordinate.
Notes on legal and regulatory obligations — when to notify the ICO, when to contact Action Fraud, what evidence to preserve.
The NCSC recommends practising this process before an incident occurs. Exercise in a Box provides free, SME-relevant tabletop exercises that take about two hours and cost nothing.
Why "Spot Every Email" Is the Wrong Goal
The goal of phishing defence is not to make every employee a security expert. The goal is to build a system where a phishing email that gets through causes no damage.
That system looks like this:
SPF, DKIM and DMARC stop your domain being spoofed → plain-English guide to SPF, DKIM and DMARC for SMEs.
Email filtering catches the bulk of malicious messages before they reach anyone.
MFA stops credential theft from becoming account takeover.
EDR catches malware that launches after a click.
A fast response process limits damage when something slips through all the above.
Training sits inside that system — it helps people report what they see, and it makes the reporting culture work. But training alone, without the technical controls above, leaves every click as a potential breach.
What Microsoft 365 Already Covers (That You Might Not Be Using)
If your business uses Microsoft 365 Business Standard, Premium, or E3/E5, you already have anti-phishing protections included. These are not add-ons — they come with the licence:
Microsoft Defender for Office 365 (included in Business Premium and E3/E5). Safe Links scans URLs in real time. Safe Attachments inspects attachments in a sandboxed environment.
Anti-phishing policies that detect mailbox-level spoofing and impersonation risks.
Safe Attachments for SharePoint, OneDrive and Microsoft Teams files shared via email.
The gap is almost never licensing. It is configuration. Many businesses leave these at Microsoft's default settings, which are deliberately conservative to avoid false positives. A proper review — the kind covered in our common Microsoft 365 security weaknesses in SMEs article — typically finds that only a fraction of available protections are active.
MFA on all accounts is the baseline. Conditional Access policies are where it gets powerful — restricting access from unmanaged devices, blocking legacy authentication, requiring approval for admin roles. These are all available in M365; they just need someone to configure them.
Phishing Simulations: Honest View
The NCSC's position on phishing simulations is clear and worth stating directly: no training package, including simulations, can teach users to spot every phishing attempt. Simulations carry legal and HR risk — punishing people for clicking on emails you sent yourself starts to resemble entrapment. The NPSA recommends checking with your HR department before running any simulations.
The reason simulations are so popular is that they produce clear, quantitative metrics — click rates, submission rates, reporting rates. But metrics express your organisation's values. If you value the absence of clicks, you incentivise people to keep problems quiet.
A better approach: measure how many people report a simulated phishing email, not how many click it. That metric actually tells you whether your reporting culture is working.
If you do run simulations, use them as a teaching tool — not a test. People who click should land on a brief training page that shows them exactly what red flags to look for. And they should be communicated to the team in advance: "we run simulations regularly, there is no blame attached."
What Should Your Staff Do With a Suspect Email?
The simplest instruction most businesses can give staff is:
Pause. Do not click, reply, approve a payment, or change payroll details.
Inspect. Check the sender address (not just the display name), look at the link destination, note any urgency or unusual requests.
Verify. Contact the person or organisation through a known phone number or established channel — never through any contact detail in the suspicious message.
Report. Use the "Report Phishing" button in your email client, or forward it to your IT provider.
Respond. If credentials were entered, change the password immediately. If an attachment was opened, disconnect the device and notify your IT provider.
This is not a complex process. It works across email, text messages, and even voice calls — which are becoming an increasingly common phishing vector.
Connecting the Dots: How This Fits with Your Overall Security Posture
Phishing defence does not exist in isolation. It sits alongside the controls that protect your payment processes, your backup strategy, and your identity management. The same layered thinking applies across all of them — see our defensible IT security baseline for SMEs for the broader framework. The Modern Workplace hub covers how these layers integrate into a coherent governance approach for your Microsoft 365 estate.
If you are worried about payment redirection, our article on invoice fraud and payment redirection checks covers the verification process that stops fraudulent bank-detail changes. If you want to understand whether MFA is properly configured across your Microsoft 365 accounts, the MFA for Microsoft 365 in UK SMEs article walks through the April 2026 landscape and what it means in practice.
Book a Security Triage Call
If you would like a structured review of your current phishing defences — what is configured, what is missing, and what the priority actions are — Book a Security Triage Call. We walk through your email security posture, MFA coverage, and the four NCSC layers to give you a clear picture of where your gaps are.
FAQ
Are phishing simulations worth running?
The NCSC says no training package, including simulations, can teach users to spot every phishing attempt. Simulations also carry legal and HR risk — punishing people for clicking on emails you sent yourself can resemble entrapment, so the NPSA recommends checking with HR before running any.
Simulations can be useful as a teaching tool if communicated transparently and followed by immediate, constructive feedback. But as a measure of how safe your organisation is, they are close to worthless. The NCSC's own position is that spotting every phishing email is impossible, and that is the central point.
A better metric: measure how many people report a simulated phishing email, not how many click it. That tells you whether your reporting culture is actually working.
What should staff do with a suspect email?
The simplest instruction most businesses can give staff is: pause, inspect, verify, report, respond. Do not click or reply. Check the sender address and link destination. Contact the person through a known phone number. Use the "Report Phishing" button in your email client. If credentials were entered, change the password immediately and notify your IT provider.
Does MFA stop phishing?
MFA does not stop a phishing email from arriving or being clicked. What it does is stop the stolen password from becoming account takeover. If a phishing email tricks someone into entering their credentials, MFA is what stands between that stolen password and someone reading your mailbox. It is not a phishing-specific control, but it is the single most effective mitigation against phishing-led account compromise.
Mid Year It Security Review June Checklist
Cyber Essentials Renewal Timing Smes August September
Cyber Essentials Q4 Renewal Certification Timeline 2026
Q4 It Security Priorities Uk Smes
How fast should we respond to a click?
You should have a plan before a click happens. A basic incident-response process for a 10–25 person business needs two known contact methods for the decision-maker, a simple flowchart of first-hour actions, a conference number for coordination, and notes on legal obligations (ICO notification, Action Fraud reporting, evidence preservation). The NCSC recommends practising this process before an incident occurs — Exercise in a Box provides free, SME-relevant tabletop exercises that take about two hours.

Cyber Security
UK SME cyber security briefing: Q4 2026 — three threats you need to watch

Backup & Disaster Recovery
Business continuity and backup for UK SMEs: a practical checklist
Cyber Security
Three Automatic-Fail Areas in Cyber Essentials

Modern Workplace
Microsoft 365 Support for Sussex SMEs: Who Owns Your Tenant

Cyber Security