Tuesday, 8:47 a.m., at a 22-person property management firm in Montréal. Everyone receives the same email: "Your parcel is on hold. Confirm your delivery address." Three people click. One types their password into the fake page. Two others forward the message to the office manager asking, "Is this legit?"
This time it was an exercise. The person who entered their password lands on a page that walks them through the red flags in about 90 seconds. The two who reported it get a thank-you. And leadership now knows where to focus.
That is the idea behind a phishing simulation. Done well, it turns your staff into a detection system. Done badly, it breeds resentment and teaches nobody anything. Here is how to get the first outcome.
TL;DR
- A phishing simulation is a training exercise, not a trap or an exam.
- The best scenarios look like your real business email, without cruelly exploiting emotions.
- The most useful metric isn't the click rate; it's the report rate.
- Individual results are personal information, so limit who can see them.
What a phishing simulation is (and isn't)
A simulation is the controlled sending of fake malicious emails to your staff, from a platform that records who opens, clicks, enters credentials or reports the message. Each action triggers an immediate learning moment.
The Canadian Centre for Cyber Security explicitly recommends that organizations run internal phishing simulations to improve employees' understanding of the risks. And the stakes are real: industry data cited by Insurance Business Canada (October 2025) puts the average cost of a phishing incident for Canadian SMEs at $89,000. Meanwhile, a 2025 Coalition study found that 77% of Canadian small businesses think they are too small to be targeted.
What it isn't: a way to catch people out, a performance review, or a substitute for technical controls. Even a well-trained team will eventually click something. That's why simulations go hand in hand with multi-factor authentication, solid email filtering and a clear process for when a real click happens.
Planning a campaign in five steps
- Get leadership on board. If the CEO is exempt from testing, the message to everyone else is obvious. Everyone participates, executives included.
- Announce the program. Explain why you're doing it and what happens after a click (training, not discipline). Don't announce dates.
- Set up the technical side. A dedicated sending domain, a landing page and allowlisting in your email filter. In Microsoft 365, that means the advanced delivery policy for third-party phishing simulations. Do a dry run with two or three people first.
- Send in waves. If all 22 employees get the same email at the same minute, word of mouth skews the results. Spread delivery over a few hours or days.
- Analyze and train. Look at results by scenario and by team, then adjust the training and the next scenario.
Choosing realistic scenarios without being cruel
A good scenario looks like the email your people actually receive. For a manufacturer in London, Ontario, that's a supplier invoice; for a clinic, a document shared by a "colleague"; for a law office, an urgent client request.
| Scenario | Why it works | Difficulty |
|---|---|---|
| Parcel on hold, delivery fee | Everyone orders online | Easy |
| Shared document (Drive, OneDrive) | Mimics a tool used daily | Medium |
| Password or MFA reset | Creates technical urgency | Medium |
| Invoice from a known supplier | Targets accounting and purchasing | Medium |
| QR code to "validate" access | Bypasses link scanning in email | Hard |
| Urgent request from the boss | Prepares staff for business email compromise | Hard |
What to avoid: fake bonuses, fake layoff notices, health scares or family emergencies. These lures work precisely because they play on emotions, and that is the problem. They damage trust in the employer for a long time, and what the employee remembers is being made a fool of.
Don't stop at email, either. The Cyber Centre's guidance also covers phishing by text message (smishing), by phone (vishing) and through QR codes (quishing). Once your email campaigns are running smoothly, consider an occasional phone test aimed at the front desk or the accounting team: "Hi, it's IT, I just need the code you received to finish the update." If someone reads out an MFA code over the phone, you have found a gap no email filter will ever close. The same rules apply: announce the program, keep it respectful, and turn every slip into a coaching moment.
Measuring what matters: reports before clicks
Click rate is the most quoted metric, but it's misleading. It swings wildly with scenario difficulty, and some security filters automatically "click" links to scan them, which inflates the numbers.
Track these instead:
- Report rate: the share of employees who use the "Report phishing" button or alert IT. This is your network of human sensors.
- Time to first report: the time between sending and the first report. In a real attack, that first report lets you pull the message from every inbox.
- Credential submission rate: more serious than a click, because that's what hands over an account.
- Trend over time: a single campaign says little; the trajectory over several months is what counts.
If employees have no easy way to report a suspicious email, fix that before your first campaign.
After the click: train without shaming
The moment right after a click is the most valuable. The employee is paying attention, a little embarrassed, and ready to learn. A short page showing the missed clues (sender address, misleading link, artificial urgency) beats an hour-long e-learning module three weeks later.
A few simple rules:
- No public leaderboard of "clickers," ever.
- For repeat clickers, supportive one-on-one coaching rather than discipline.
- Results reported to leadership in aggregate, by team, not by name.
- Restricted access to named results and a defined retention period, since this is personal information.
Also remind people what happens in real life: if someone types their password into a fake page, multi-factor authentication is often what stops the attacker. And every employee should know what to do in the next few minutes, which is exactly what your incident response plan should spell out.
Checklist before your first campaign
- Leadership has approved the program and takes part in the tests.
- Staff know a program exists and that a click leads to training, not discipline.
- A report button or address is in place and everyone knows it.
- The sending domain and landing page are ready; allowlisting is configured.
- A dry run has been done with a small group.
- Scenarios are realistic and don't exploit health, jobs or family.
- Metrics are defined: report rate, time to report, credential submissions.
- Access to named results and their retention are controlled.
Our team designs and runs simulation campaigns tailored to your organization, as part of our ongoing security partner service or as a one-off engagement.
Frequently asked questions
Is it legal to run phishing simulations without warning employees?+
Generally, an organization may test its own systems and train its staff, provided it acts reasonably and transparently. Common practice is to announce that the program exists, in a policy or a staff memo, without revealing campaign dates. Because results identify individuals, they are personal information under Québec's Law 25 or other applicable privacy laws, and employment rules may also apply, so validate your approach with your legal advisor.
How often should we send phishing simulations?+
Small, regular campaigns, for example monthly or quarterly, work better than one big annual exercise. The goal is to build a reflex, not to run an exam. Vary scenarios and send times so employees can't spot the test by its format.
Can we run simulations ourselves with Microsoft 365?+
Yes, if you have the right licence: Microsoft's Attack simulation training requires Microsoft 365 E5 or Microsoft Defender for Office 365 Plan 2. Open-source tools also exist, but they need a dedicated sending domain, allowlisting and time to design scenarios. An outside partner mainly brings neutrality: it isn't your IT colleague who is trying to trick everyone.
Sources
- 1.Canadian Centre for Cyber Security: Don't take the bait (ITSAP.00.101)
- 2.Insurance Business Canada: AI-powered attacks push cyber risks higher for Canada's SMEs (October 2025)
- 3.Canadian Underwriter: Canadian small businesses underestimate cyber risk (Coalition, 2025)
- 4.Microsoft: Configure the advanced delivery policy for third-party phishing simulations
- 5.Microsoft: Get started using Attack simulation training
- 6.Act respecting the protection of personal information in the private sector (CQLR c. P-39.1)