A datacenter fire, a ransomware attack, a simple human error… and the whole business stops. The question is no longer “if”, but “when”. A Disaster Recovery Plan (DRP) and a Business Continuity Plan (BCP) turn a potential catastrophe into a controlled interruption. Provided you build them on the right foundations — and test them.

Many organisations think they have one… until the day they need it. Here’s how to build a genuinely reliable plan, starting with the two numbers that structure everything: RTO and RPO.

1. DRP, BCP: what are we talking about?

The BCP (Business Continuity Plan) aims to keep the business running, even in degraded mode, during the incident — the business view. The DRP (Disaster Recovery Plan) aims to bring the IS back into service after a disaster — the technical view.

2. The 2 numbers that structure everything: RTO and RPO

Before any technical solution, two objectives to set per service: how long can you go without it, and how much data can you afford to lose.

RPO (Recovery Point Objective): the amount of data you accept to lose — driven by backup frequency. RTO (Recovery Time Objective): the maximum time to restore the service — driven by the recovery strategy.

3. Where to start: the 5-step method

4. The strategy depends on the target RTO

The shorter the RTO, the more sophisticated — and costly — the solution. So size by criticality: vital services first.

5. Backups: the non-negotiable foundation

• 3-2-1 rule: 3 copies, 2 media, 1 offsite / offline.
• Immutability: copies that neither an attacker nor an error can alter.
• Isolation: backups not reachable with the current IS credentials.
• Restore tests regularly: a backup never restored doesn’t really exist.

6. The misconceptions

7. Your DRP / BCP checklist

8. Frequently asked questions

Case study: Datacenter migration & IS rebuild for a mutual-insurance group →

“We’ll move the mailboxes to the cloud — a weekend’s work.” That’s how most Microsoft 365 migrations that go wrong begin: missing emails on Monday, locked-out users, file shares open to everyone, security left at the door. A successful migration isn’t a copy-paste — it’s a managed project, where you prepare, test, cut over in waves, and secure from day one.

Here’s what you actually migrate, the possible approaches, and the method for a controlled cutover — with no data loss and no open gap.

1. What you really migrate (not just the mailboxes)

Reducing Microsoft 365 to Outlook ignores 80% of the project. A migration touches five interdependent workstreams, and it’s their coordination that makes it a success — or a failure.

2. Three approaches, three risk levels

3. The 5-phase method

Whatever the approach, a controlled migration always follows the same sequence. You never cut over before you’ve piloted.

The pilot wave (phase 3) is decisive: a representative group migrates first, reveals the surprises (mail rules, shares, third-party apps) and validates the runbook before scaling up.

4. Secure from day one

A migration is the best moment to lay a clean foundation — far harder to fix once everyone is settled in. The fundamentals to enable right away:

• MFA for everyone: multi-factor authentication blocks the vast majority of account takeovers.
• Conditional access: restrict sign-ins by device, location or risk.
• Controlled sharing & guests: by default, you don’t share “with the whole internet”; you frame external access.
• Data protection & backup: leak-prevention policies (DLP) and third-party backup — the cloud is not a backup.

5. The misconceptions

6. Your migration checklist

7. Frequently asked questions

Case study: International reinsurer — Full-cloud migration & security →

“We’ll do the audit when we have time.” Then comes a regulator’s inspection, a client demanding certification, or a data breach. An audit is neither a box to tick nor a punishment: it’s the honest snapshot of your security posture — the one that turns a vague hunch (“we’re more or less protected”) into a prioritised action plan.

The trick is knowing which audit you mean — because a “GDPR audit” and an “ISO 27001 audit” don’t target the same thing — and where to start without drowning. Here’s the method.

1. GDPR, ISO 27001: two complementary logics

They’re often confused. GDPR is a legal obligation centred on protecting personal data. ISO 27001 is a voluntary standard that structures the security of all information through a management system (ISMS). One protects people, the other organises the company — and the two reinforce each other.

2. The 3 kinds of audit not to confuse

Behind the word “audit” hide three different exercises. Mixing them up is comparing apples and oranges.

3. The 5-step method

Whatever the framework, a reliable audit always follows the same thread: scope, collect, analyse, prioritise, follow up.

The key deliverable isn’t the findings report, but the risk-prioritised action plan: what to fix first, with what effort and what impact.

4. GDPR side: the 4 foundations

• Records of processing: the map of who processes which data, why and for how long. A non-negotiable starting point.
• Legal basis: every processing activity must rest on a valid ground (consent, contract, legal obligation, legitimate interest…).
• Data subject rights: access, rectification, erasure, portability — procedures that respond within the legal deadlines.
• Security & breaches: appropriate technical measures and a 72-hour notification procedure in case of a breach.

5. ISO 27001 side: the ISMS, not just controls

ISO 27001 doesn’t ask you to blindly apply a list of controls, but to manage security by risk: identify your assets, assess the risks, select the Annex A controls that treat them, then measure and improve continuously (PDCA logic). Certification attests that this system exists and works.

6. The misconceptions

7. Your “before you start” checklist

8. Frequently asked questions

Case study: IS rebuild & compliance for a mutual-insurance group →

Too often, security shows up at the end of a project — when the architecture is frozen, the budget spent and deadlines tight. The result: costly patches, compromises, and flaws that could have been avoided. Security by Design flips the logic: you build security in from the very first line of the specification.

Good news: it’s not about expensive tools, but about habits. Here are the five that change your projects’ trajectory — and why adopting them early costs far less than fixing late.

1. What is Security by Design?

It’s a simple principle: design a system by building security into every stage of its lifecycle, rather than bolting it on afterwards. Security becomes an intrinsic property of the architecture — like performance or availability — not an add-on.

2. Why fixing late costs so much

The later a flaw is found, the more expensive it is to fix: you have to undo, retest, redeploy, sometimes under the pressure of an incident. That’s the whole “shift left” logic: move the security effort upstream, where it pays off most.

3. The 5 habits to adopt

4. The misconceptions

5. Your Security by Design checklist

6. Frequently asked questions

Case study: Cloud environment audit & hardening of a major institution →

One Monday morning, the screens display a message: “Your files have been encrypted.” Email is dead, business apps are down, and a countdown demands a ransom in cryptocurrency. Thousands of organisations live this scene every year. The good news: a ransomware attack is not inevitable. It follows a known — therefore predictable — scenario, and is prepared long before D-day.

In this article, we take apart how ransomware works step by step, list the mistakes that cost dearly, then detail a concrete remediation plan and a checklist to keep at hand. The goal: that you, as an IT director or manager, know exactly what to put in place — and what to do if the incident strikes.

1. What exactly is ransomware?

Ransomware is malicious software that encrypts your data to make it unusable, then demands a ransom in exchange for the decryption key. But in recent years, attackers no longer just block: they first steal the data, then threaten to publish it. This is double extortion — and sometimes triple, adding blackmail of customers or a denial-of-service attack.

2. Anatomy of an attack, step by step

Unlike the image of a “virus” striking at random, a ransomware attack is a driven process, often over several days. Understanding its phases means spotting as many opportunities to detect and stop it.

The most common entry points

• Phishing: a booby-trapped email that steals a password or installs an initial payload.
• Exposed remote access: an RDP or VPN reachable from the Internet, without two-factor authentication.
• Unpatched vulnerabilities: a server or device behind on updates.
• Supply chain: a compromised provider or third-party software.

3. The misconceptions that cost dearly

4. Prepare: defence in depth

No single measure stops ransomware. The right approach stacks complementary layers: if one gives way, the next hold. This is the defence-in-depth principle.

5. In the event of an attack: the remediation plan

If the incident occurs, improvising makes everything worse. A clear, known, rehearsed plan saves hours — sometimes days. Here’s the five-step outline.

Alongside the technical response, two reflexes: trigger the crisis unit (management, IT, communications, legal) and notify the relevant authorities — in France, the CNIL within 72 h for a personal-data breach, and file a complaint. Preserve the evidence (logs, affected systems): it’s essential for analysis and insurance.

6. Your anti-ransomware checklist

7. Frequently asked questions

Case study: Operation & security of a critical multi-site IS →

Smartphones and tablets now access the same data as desktops — email, files, business apps — but often escape the same security discipline. Yet a lost phone, a booby-trapped Wi-Fi or a malicious app is enough to open a breach. Two complementary building blocks let you regain control: MDM and MTD.

Often confused, they don’t play the same role: one manages and frames the devices, the other defends them in real time against threats. Here’s how they fit together, and how to deploy a genuinely controlled mobile fleet.

1. Why mobile is a blind spot

Mobile stacks up risk factors: it carries sensitive data, constantly leaves the corporate network, mixes work and personal use (BYOD), and multiplies entry points — loss or theft, public Wi-Fi, booby-trapped apps, SMS phishing (smishing).

2. MDM and MTD: two complementary roles

MDM (Mobile Device Management) manages the devices; MTD (Mobile Threat Defense) defends them. The first frames and administers, the second detects and blocks threats in real time. One without the other leaves a gap.

3. The 3 vectors MTD covers

A mobile threat doesn’t only come through apps. A good MTD watches all three vectors at once, while respecting the user’s privacy.

4. Deploying a controlled fleet

• Enrolment of devices (corporate and BYOD) and an up-to-date inventory.
• Security policies: lock, encryption, enforced updates.
• Work / personal separation via a container, respecting privacy.
• MTD across the three vectors, with automated remediation.
• Conditional access: a non-compliant device doesn’t reach resources.
• Remote wipe (of the work container) on loss or departure.

5. The misconceptions

6. Your mobile fleet checklist

7. Frequently asked questions

Case study: Mobile fleet protection for a public body →