Managed security: what you are buying between EDR, XDR, MDR and SOC

Operatore di un SOC monitora gli eventi su piu schermi: sicurezza gestita per le PMI

In a professional firm of about thirty people the server sits in a storeroom and the backup is copied to the cloud. Support comes from an external technician, competent and helpful, who comes when called. The Microsoft licences expired last month without anyone noticing, nobody has ever written down the map of permissions on network folders, and the server’s users still include people who left in 2012. Nobody in that firm would say they have a managed security problem: they would say they have a time problem.

It is the most common situation in Italian companies of that size, and it explains why the managed security market mostly sells acronyms. An offer with four acronyms looks more solid than one that describes activities and responsibilities.

Managed security acronyms, explained for decision makers

An antivirus, or endpoint protection, blocks what it recognises: known malware, behaviour already catalogued. It is needed, and it is not enough.

EDR, Endpoint Detection and Response, watches how devices behave and flags anomalies even when no malware signature exists yet. It produces alerts and response tools.

XDR, Extended Detection and Response, extends the field of view beyond devices and correlates what happens on endpoints, network, email, identities and cloud environments. An attempt that starts from a compromised mailbox and continues on a server thus becomes readable as a single episode instead of two unrelated events.

MDR, Managed Detection and Response, is a service rather than a technology: it puts dedicated people behind those alerts, an analyst who reads them, tells a false positive from an intrusion and acts. The SOC, Security Operations Center, is the structure that delivers that monitoring, often relying on a SIEM to collect and correlate logs.

Side by side, the four acronyms differ as follows:

Acronym What it covers What it asks of the company
EDR Anomalous behaviour on individual devices, with response tools Someone to read the alerts and act
XDR Correlation of endpoints, network, email, identity and cloud Connected log sources and a mapped perimeter
MDR The service: people who read alerts, filter them and respond A written mandate on who authorises interventions
SOC The structure that delivers the monitoring, often with a SIEM An internal contact who receives reports and decisions

The difference that drives the cost of managed security

EDR and XDR are licences. MDR and SOC are hours of skilled people, available at night too. Comparing two offers by price per workstation puts items of a different nature on the same scale, and the difference emerges at the first real incident.

Thirty people, an on-call technician and users active since 2012

What happened next in the opening case is instructive. The initial request concerned a structured support contract, not security. The review, however, brought several things to light: nobody knew which licences were active, folder permissions had built up over years of one-off requests, and several accounts of people who had left the organisation were still in the system. The VPN, enabled for almost everyone, opened up the internal network.

The concrete risk was not a sophisticated attack. It was a forgotten credential that was still active. No EDR would have flagged a legitimate login by a user who still exists in the system.

The managed security service started from exactly that point. The platform mapped users, licences and permissions across the whole perimeter and in a few hours surfaced what no manual check caught any more: the accounts of people who had left, the layers of years of one-off requests, the VPN open to almost everyone. The account clean-up and the written procedure for joiners and leavers came out of that inventory, and on the cleaned-up perimeter detection and monitoring began working on real signals instead of noise.

In what order managed security is built

In the projects we run, the effective sequence puts basic hygiene before protection, protection before detection and detection before monitoring.

  1. Backups with documented restore tests and at least one copy unreachable from the network it protects.
  2. Identity management: named accounts, multi-factor authentication, removal of unnecessary privileges, a leaver procedure applied without exceptions.
  3. Updates applied regularly and an inventory of what exists on the network, including what nobody remembered.
  4. Protection of endpoints and email, which remains the most used entry channel.
  5. Detection with EDR and, where the perimeter justifies it, with XDR.
  6. Monitoring with MDR or a SOC, when there is someone who acts on the alerts.

Buying the sixth item without having the first produces a well-designed dashboard on top of an exposed infrastructure. It happens more often than you might think, because the sixth item sells better than the first.

Why NIS2 also concerns those who are not subject to it

NIS2 also affects companies outside its scope: obliged entities are accountable for their supply chain and pass requests for evidence on to their suppliers.

The NIS2 directive identifies specific sectors and sizes, and many small and mid-sized companies fall outside it. A company that supplies components, services or software to a NIS2 entity, however, receives questionnaires on the managed security of its own systems, requests for evidence and new contract clauses.

Those who arrive prepared treat that request as a routine task. Those who are not find out within a few weeks that they must document processes they have never formalised, with a deadline set by a customer.

Questions to ask a managed security provider

They apply to any provider, us included, and the answers should be attached to the contract.

  • Who reviews the alerts, during which hours and from where.
  • What happens at three in the morning: who receives the alert and what actions they can take without waiting for authorisation.
  • Whether the service includes isolating a compromised machine or only reports the problem.
  • Which log sources feed the correlation: endpoints only, or firewall, email and identities as well.
  • How often a report arrives that management can read, not just the technical team.
  • What remains outside the contract and is the company’s responsibility.
  • Who carries out the analysis after an incident and whether that activity is already included in the fee.
  • Within what time the provider commits to act between detection and first intervention.

A commitment made verbally in a meeting has no effect at the moment of the incident.

One point deserves attention at contract stage: monitoring reaches full effectiveness when it comes with a mandate to intervene. It is worth setting down in writing who authorises an urgent action outside business hours and which interventions the provider can carry out on its own, so that the response to a night-time incident starts without waiting for a phone call.

What managed security can promise

Managed security promises two measurable results: fewer attempts that succeed and less time between an attacker’s entry and the reaction. Anyone promising invulnerability deserves a cautious hearing, because no combination of technologies and services offers it. That time decides whether the episode stays contained or becomes a three-day production stoppage.

What to measure after managed security goes live

A security contract is assessed on a few indicators, and none of them is the number of threats blocked, which describes background noise more than protection. What counts is the average time between detection and response, the share of devices actually covered by the agent compared with the inventory, the documented outcome of the latest restore test, and the number of critical vulnerabilities still open past the agreed deadline.

At Aesir Srl we keep fixed-fee IT management and security within the same contract, as MSP and MSSP: when an incident requires shutting down a service, the decision does not go through two companies passing responsibility back and forth. The same organisation is certified ISO 9001, 27001, 27017 and 27018 and runs our cloud platform in our own Tier IV data centres in the European Union, with replication to a second site.

Frequently asked questions

Does the existing firewall need replacing?

Not necessarily. A device still supported by the manufacturer and kept up to date can stay, as long as its logs feed the correlation. The case to assess is another: an out-of-support firewall no longer receives fixes and becomes the weak point of the perimeter.

Are EDR and antivirus the same thing?

No. Antivirus recognises what is already catalogued; EDR watches behaviour and leaves a trace of what happened on a device. In post-incident analysis that trace makes the difference between reconstructing the facts and guessing them.

What does a security assessment include?

An inventory of exposed devices and services, a check of known vulnerabilities with a priority scale, a review of identities and privileges, the state of backups and updates, and an examination of configurations. What remains at the end is an ordered list of actions and an estimate of the effort required, not a score to show off.

Does a small or mid-sized company need a SOC?

The answer lies in two numbers: how much a day of downtime costs and who would handle an alert during unstaffed hours. A company producing on three shifts and a firm working Monday to Friday in office hours have incomparable monitoring needs.

Let’s talk

One exercise can be done straight away, without involving any provider: pull the list of people who left the company in the last two years and check accounts, mailboxes, VPN access and ERP logins. The outcome of that check says more than any sales presentation and shows where to start a managed security project.

If you would like to explore the topic or assess your company’s situation, fill in the form at the bottom of the page or write to support@aesir-tech.it: we will book a free consultation and start from your numbers.

Vuoi parlarne applicato alla tua azienda?

Ti rispondiamo con l’esperienza di progetto, non con una brochure.

Request a call