Security awareness: preparing your team against scams
How phishing, vishing and payment fraud target people, how to build an awareness program that changes behavior, and how to measure the result.

Firewalls, EDR and two-factor authentication cut risk a great deal. Even so, a significant share of incidents begins with a person: someone who opens an attachment, types a password into a fake page, approves a sign-in prompt they did not request, or changes a supplier’s bank details in reply to a convincing email. An attacker does not need to break the technology if they can talk the people who use it into helping.
A security awareness program exists to change that behavior. It is not a yearly talk or a mandatory video. It is ongoing work that teaches the team to recognize real scams, gives them a simple way to report what looks odd, and measures whether that is happening.
How scams against people work
Social engineering exploits predictable levers: urgency, authority, fear of losing something, the wish to help. MITRE ATT&CK lists phishing as an initial access technique (T1566), with variants by attachment, by link and by service, such as messages in apps and social networks. These are the most common forms.
Phishing and spearphishing
Generic phishing goes out in bulk: a delivery notice, a bill, an expired password warning. Spearphishing is crafted for one person or team, with real names, roles and subjects, which makes it much harder to spot. The goal is usually one of two: steal credentials through a page that mimics the company’s sign-in, or get the victim to run a file.
An important point for training: two-factor authentication does not solve everything. Modern phishing kits sit between the victim and the real site and capture the code or the session cookie as well. Another trick is MFA fatigue: an attacker who already has the password fires approval prompts until the person accepts one out of weariness. Teaching the team never to approve a sign-in they did not start themselves is one of the most useful lessons. On the technology side, phishing-resistant authenticators, such as FIDO2/WebAuthn security keys and passkeys, close much of that gap.
Vishing and smishing
Phone calls (vishing), text messages (smishing) and scams in messaging apps take the attack out of email, away from email filters. A frequent pattern is a call to the help desk on behalf of an employee, asking to reset a password or enroll a new MFA device. Another is a call to an employee on behalf of the help desk, asking them to install a remote access tool.
Payment fraud
Here the goal is money, not access. The attacker poses as a supplier and announces that its bank account has changed, or poses as an executive and asks for an urgent, confidential payment. Tampered invoices and instant payment requests outside the normal flow belong here too. The defense is more process than technology: no change to bank details without confirmation through a channel you already know.
What an awareness program needs
NIST publishes a reference for this kind of program, SP 800-50 Rev. 1, Building a Cybersecurity and Privacy Learning Program, which treats it as a cycle: assess needs, plan, deliver, measure and adjust. In practice, these are the elements that make a difference.
- Content drawn from the company’s own day. Finance needs to recognize payment fraud; the help desk, fake reset requests; leadership, spearphishing that uses their own names. One generic course for everyone teaches each of them little.
- Concrete behaviors, not concepts. Check the sender and the link’s real domain, distrust urgency, do not approve sign-in requests you did not start, confirm payment requests through another channel, use a password manager.
- An easy way to report. A report button in the email client, or a single, well-known channel. If reporting is a chore, people do not report.
- A blame-free culture. The person who clicked and told someone straight away is the one who helped most. Punishing mistakes teaches people to hide them, and a hidden mistake gives the attacker hours of head start.
- Handling sensitive data. How to treat confidential information and personal data, in line with data protection law (in Brazil, the LGPD).
- Frequency. Short campaigns through the year work better than one big event.
How to measure whether it works
The most quoted metric, the click rate in phishing simulations, is the least useful on its own. It swings with the difficulty of the scenario, and a simulation that is too easy produces a good-looking result that says nothing.
More useful indicators:
- Report rate: how many people told the security team about the simulation. This number should rise over time.
- Time to first report: the sooner someone raises the alarm, the sooner the SOC can block a real campaign.
- Data entered: how many people, beyond clicking, typed their password. That is the exposed credential an attacker would use, even where other controls would stop the sign-in.
- Results by team and by scenario: where the team holds and where it gives way, to steer the next topic.
Results should be presented by team, not by name. The goal is to improve processes and training.
An illustrative example
The scenario below is a composite of common situations, not a real customer. A retailer runs a first simulation with three scenarios: an expired password notice linking to a fake sign-in, an invoice with an attachment, and a call to the help desk asking to reset a manager’s password.
The diagnosis shows that many people in finance enter their password on the fake page, that almost nobody reports because they do not know whom to tell, and that the help desk resets passwords over the phone with only a name and an employee number, both easy to find.
The program that follows has three fronts. First, a report button in email and a clear message about what to do with anything suspicious. Second, targeted sessions: payment fraud and fake pages for finance, identity verification for the help desk. Third, a process change: password resets only with a call back to the number on file, or with the manager’s approval.
In the next simulation, with scenarios of similar difficulty, more people report, the first report comes sooner, and the fake help desk request is refused. The work does not end there: each round points to the next topic.
What to ask a supplier
- Is the content adapted to my company? Ask how the program treats different teams and whether it uses examples from your routine.
- How are results measured? Be wary of anyone who measures only clicks. Ask for the report rate, data entered and the trend between rounds.
- Do the simulations cover more than email? Vishing and smishing against the help desk and finance often reveal process gaps.
- How are employees treated in the reports? Ideally, results by team, with no individual exposure.
- Does the program include process recommendations? Some gaps are closed not by training but by a verification rule.
- What is the calendar? An ongoing program has topics and campaigns through the year, not a single event.
Next step
Awareness works best when it is measured with realistic simulations and adjusted after each round. If you want to know how your team reacts today, a social engineering test gives you a starting point. To build an ongoing program for the whole team, see Leonnes’ awareness programs.