How to recognise a fake email in a company in Algeria (2027): an accountant nearly gave her credentials to a fake bank, another nearly paid a fake manager, three questions in ten seconds and what protects you when someone clicks

“An accountant received an email ‘from the bank’ asking her to confirm her credentials; another, an email ‘from the manager’ demanding an urgent transfer. They nearly did it. How do we teach the team to recognise a fake email, and what do we put in place so that a mistake costs nothing?” Three questions before any click: who really sends it, where the link leads, what am I being rushed to do. And, because someone will click one day, accounts that a stolen password does not open.

Symloop13 min read
Recognising a fake email before clickingThree questions in ten seconds, the address read up to the first slash, five nets if someone clicksFAKE EMAILS · ALGERIA · 2027Recognising a fake emailbefore clickingThree questions in ten seconds, the address read up to thefirst slash, five nets if someone clicksAN “URGENT” EMAILcredentials, transfer, attachmentTHREE QUESTIONSwho, where, what, in ten secondssymloop.com

Short answer. Three questions, ten seconds, before any click. Who really sends it: the address after the @, not the displayed name. Where the link leads: the name just before the first slash, not the text of the button. What am I being rushed to do: give a credential, pay, open a file. Any request for money or credentials is confirmed by another channel. And because someone will click one day: two-factor sign-in on the mailbox and the bank, a password manager, your domain protected against spoofing, an address for reporting without being judged, and a payment rule that depends on no email. Four weeks in the example.

This guide is for the manager whose team receives these messages every week, and for the founder based in France whose accounting is in Algeria. The definitions and facts quoted come from public sources: phishing, social engineering, email spoofing, typosquatting, multi-factor authentication, DMARC, WebAuthn, .dz domain names; the figure on two-factor sign-in is the one Microsoft published in 2019. The bank, the addresses and the messages are a constructed example, with fictitious domain names. This guide does not cover legal rules or remedies after a fraud: see your legal adviser and your bank.

Six fake emails, what they ask for

The figure gives, for six common fake emails, what they ask for and the defence.

Figure 1 · Six disguises, only three requests: your credentials, your money, or the opening of a file. The defence is the same idea each time: do not follow the path the message offers, and check by another, the one you already know.

What it is. Phishing is a form of social engineering and a scam in which attackers deceive people into revealing sensitive information or installing malware. Social engineering is the use of psychological pressure to influence people to perform actions or divulge confidential information: urgency, authority, the fear of doing wrong. Only three aims. Behind the six disguises there are only three requests: your credentials, your money, or the opening of a file. Recognising the request is safer than recognising the disguise.

The fake manager. It contains neither link nor attachment: a short, hurried message asking for a transfer or a change of bank details. Email spoofing is the creation of email messages with a forged sender address; the displayed name, for its part, can be chosen freely. The only defence is outside the mailbox: call the person back on a number you already know. If it has already happened, the steps for the first hour are in the guide to the hacked mailbox and the fake bank details.

01

If someone clicks anyway: five nets

The accented layer is the person; the other four serve on the day they get it wrong.

Figure 2 · The accented layer is the person: they are the one reading the message. The other four serve on the day they get it wrong: a domain that cannot be spoofed, a mailbox that labels, an account a stolen password does not open, and a payment rule that depends on no email.

Someone will click. No training brings the risk down to zero: a Friday evening, a well-made message, a tired person. So the set-up is judged by what happens after the click. The account. Multi-factor authentication, also known as two-factor authentication, is an electronic authentication method that requires a proof in addition to the password: a code, an app, a key. According to Microsoft, in a post from August 2019, it can block over 99.9 percent of account compromise attacks. Turn it on first on the mailbox and on the bank.

The domain. DMARC is an email authentication protocol, designed to give email domain owners the ability to protect their domain from unauthorized use: with SPF and DKIM, it prevents a stranger from writing to your customers in your name. Setting them up is described in the guide to professional email. The money. A written rule, which applies to the manager too: no transfer or change of bank details without confirmation by another channel. The rest of the network. A compromised computer must not reach everything: see the guide to securing a company network.

«The set-up is judged by what happens after the click.»
02

A fake email, read in ten seconds

The figure places six markers on a message imitating a bank.

Figure 3 · A message imitating a bank, and six markers: the sending address, the urgency, the general greeting, the request, the link whose destination differs from its text, the attachment. None is proof on its own; two together are enough not to click. Constructed example.

Three questions. Who really sends it: the address, after the @, not the displayed name. Where the link leads: the real address, not the text of the button. What am I being rushed to do: give a credential, pay, open a file. Ten seconds are enough, provided you take them before the click. Spelling proves nothing any more. A careful fake email has no mistakes, and an AI assistant writes correctly in any language. What cannot be imitated is the sending address and the destination of the link.

The attachment. A lot of malware is distributed via email attachments. An unexpected attachment is not opened, even from a known contact, whose mailbox may be compromised: ask them to confirm by another channel. What follows a file opened too quickly is described in the guide to ransomware. On the phone. The screen shows the name, rarely the address: a tap on the sender's name displays it. Make it a habit: that is where messages are read between two tasks.

03

Where a link really leads

The figure takes one address apart, then compares five.

Figure 4 · Where a link really leads. The real site is the name just before the first slash, with its extension; the start of the address can imitate anyone. Five addresses compared: two lead to the bank in the example, three elsewhere. Fictitious domain names.

The rule. The site is the name just before the first slash, with its extension; everything before it can imitate anyone. In the example, the address begins with the bank's name, but the site is compte-verif.net. In Algeria. Names in .dz are registered directly, like exemple.dz, or beneath an intermediate name, like exemple.com.dz: in that second case, the real name has three words and not two. Write down once the exact address of your bank and of your suppliers.

One letter. Typosquatting is a form of cybersquatting that relies on typing mistakes and look-alikes: one letter changed, a hyphen in place of a dot, another extension. What does not get it wrong. A password manager offers the credential only on the address where it was saved: on an imitated site it stays silent, and that silence is a warning. Passkeys go further: the WebAuthn standard is resistant to some phishing attacks, because the authenticator only offers credentials that were registered on the same website. They are described in the guide to sign-in by code.

«Everything before the real name can imitate anyone.»
04

Four weeks, from taking stock to the first exercise, in 2027

Four weeks in the example, the first one to look at what already arrives.

Figure 5 · Four weeks in the example. The accented period is the first: looking at what already arrives. Accounts come before training, because they protect even those who have not been trained yet. Indicative durations from a constructed example.

Week 1, taking stock: the fake emails received this month, those someone clicked on, and what already protects the accounts. Weeks 1 and 2, accounts: two-factor sign-in on the mailbox and the bank, password manager. Weeks 2 and 3, the domain and the mailbox: SPF, DKIM and DMARC checked, a banner on external messages, a reporting address.

Week 3, training, one hour per team. Weeks 3 and 4, the payment rule, written and signed by management. Week 4, the first exercise: an announced practice email, and two figures recorded, the share of those who clicked and the share of those who reported. The second matters more: a company where people report quickly limits the damage of a click. An outside look at the whole is the role of the security audit.

05

Training without humiliating, in six steps

The accented step is the third: give three questions.

  • What the manager receives: the stock-take, two-factor sign-in turned on for the mailbox and the bank, a password manager, the domain protected, a reporting address, a written payment rule, one hour of training per team, and two figures per quarter.
  • What the company no longer does: pay on the strength of an email, type a password from a link, open an unexpected attachment, find out about a click three days later.
  • What is measured: accounts protected by two-factor sign-in, share of clicks and share of reports at each exercise, time between a report and the answer.
  • What Symloop does: the stock-take, two-factor sign-in and the password manager, the configuration of the domain and the mail service, the reporting address, training on your own examples, then the quarterly exercise.
Figure 6 · The accented step is the third: give three questions, which fit in ten seconds. The others build trust: an announced exercise, examples really received in the company, an address for reporting, and thanks for whoever reports.

Why without humiliating. Someone who clicked and is ashamed stays silent, and it is the silence that costs: a password stolen at nine o'clock and reported at five past nine is changed before it has been used. So the rule is stated in advance: nobody is punished for clicking, everybody is thanked for reporting. Examples from here. Ten fake emails really received in the company are worth more than a slide deck: the fake bank, the fake carrier, the fake supplier that everyone recognises.

Reporting. An internal address, or a button in the mail client, and an answer within the hour: “it is a fake, thank you”, or “it is genuine”. Without an answer, people stop reporting. Every quarter. An announced exercise, never a trap, and the same two figures. Newcomers. An employee's first week is when they know the least about how the company works: the three questions are part of their induction.

How do you recognise a fake email?
With three questions before any click: who really sends it, by reading the address after the @ and not the displayed name; where the link leads, by reading the name just before the first slash; what am I being rushed to do, a credential, a payment or a file to open. Two suspicious signs are enough not to click and to report.
How do you recognise a fraudulent email from your bank?
A bank does not ask for your credentials by email. Never use the link in the message: open the site yourself, from your bookmarks or the app, or call your branch on a number you already know. The displayed name proves nothing; only the sending address counts.
How can you tell where a link leads without clicking?
On a computer, hover over it: the address appears at the bottom of the window. On a phone, a long press shows it without opening it. Then read the name just before the first slash: that is the site.
What should I do if I clicked on a fake email?
Report it straight away, without waiting to be sure. Change the password concerned from another device, close open sessions, check that no forwarding rule has been added to the mailbox, and warn the bank if payment details were entered.
How does two-factor sign-in work?
After the password, the service asks for a second proof that only you hold: a code shown by an app, a notification to approve, a physical key. A password stolen by a fake email is then no longer enough to get in.
Is an email with no spelling mistakes safe?
No. Careful fake emails have no mistakes, and AI assistants write correctly in every language. Spelling is no longer a criterion: the sending address, the destination of the link and the nature of the request are.
The manager asks for an urgent transfer by email: what should we do?
Confirm by another channel before any payment: a call to their usual number, never the one in the message. The rule must be written and apply to everyone, the manager included: nobody can demand that it be bypassed “because it is urgent”.
Should we run fake-email exercises?
Yes, provided they are announced and carry no punishment. Record two figures, the share of clicks and the share of reports, and thank those who report. An exercise felt as a trap mostly teaches people to stay silent.
Is a spam filter or an antivirus enough?
No. The filter stops part of the messages, but the fake manager has neither link nor attachment to analyse. Several nets are needed: the protected domain, the mailbox that labels, the trained person, the account with two-factor sign-in, and a payment rule.
How long does it take, and where do we start?
Four weeks in this guide's example. Start today with two-factor sign-in on the mailbox and on the bank: it is the net that protects even those who have not been trained yet.
Tell us how many mailboxes, which mail service, and what you have already received: we take stock, then we protect the accounts before training

A fake email from the bank, a fake manager asking for a transfer, an attachment opened too quickly, or a founder in France whose team in Algeria receives these messages every week: we set the five nets, then we train on your own examples

Stock-take, two-factor sign-in, password manager, SPF, DKIM and DMARC, reporting address, payment rule, training, quarterly exercise. We work from Algiers and Sétif, in French, Arabic and English.