HomeNews / ArticlesDigital TransformationEDR vs MDR: Which Security Model Fits Your Business

EDR vs MDR: Which Security Model Fits Your Business

43% of UK businesses and 30% of charities experienced a cyber breach or attack in the 12 months before publication of the Cyber Security Breaches Survey 2025, roughly 612,000 businesses and 61,000 charities. Only 30% of businesses had formal incident-management capability, against 61% of medium-sized and 85% of large businesses. That second number is the one that decides the edr vs mdr argument: almost every organisation can now generate an alert, and far fewer can act on one at 11pm.

An EDR licence is a small line in the budget. The staffing, authority and governance that make it work is the expensive part, and it rarely appears in a quote. Treat this as a total operating model question, where the internal hours you commit sit on one side of the ledger and the MDR fee sits on the other. Licence prices tell you almost nothing on their own.

EDR vs MDR at a Glance

EDR is software installed on laptops, servers and other devices that collects telemetry, flags suspicious behaviour and lets an authorised person isolate a device. MDR is a managed service in which security analysts watch those alerts, investigate them and coordinate response, often outside office hours.

The benefits are real on both sides, and they overlap. EDR gives your team continuous telemetry from every protected endpoint, behavioural detection that catches fileless attacks a signature-based antivirus misses, and the ability to isolate a compromised machine in seconds rather than hours. MDR gives you continuous monitoring by analysts, triage and escalation against an agreed playbook, and a defined response process where internal security capability is thin. Both depend on the same endpoint telemetry, so the table below shows who delivers each benefit rather than which model wins.

EDR and MDR compared across the key criteria

CriteriaEDRMDR
What it isEndpoint software and telemetry on laptops, servers and other devicesManaged service run by security analysts, usually built on EDR telemetry
Who watches the alertsYour own IT or security staffThe provider's analysts, usually around the clock
Investigation and triageCarried out by your team in the consoleCarried out by the provider and escalated under an agreed playbook
ContainmentIsolation and quarantine available, normally actioned by your teamActioned by the provider within agreed authority
Staffing neededTrained internal staff available to respond to alertsMinimal internal time, with one accountable contact
Pricing modelPer-device licence plus internal staff timeRecurring monthly or annual fee
Best fitOrganisations with tested in-house security operationsOrganisations without out-of-hours cover or specialist security staff

MDR usually runs on top of EDR telemetry, so the two are layers rather than rivals. A provider without endpoint sensors is guessing; sensors without anyone watching them are a log file.

First check: could your team act on a high-severity alert within the hour on a Friday evening, or on the second day of a bank holiday weekend? If the honest answer is no, that gap is the thing you are actually buying.

How Each Model Handles a Live Attack

It's 22:40 on a Friday. A finance manager's laptop spawns a PowerShell process that pulls an encoded script from a newly registered domain, then starts enumerating local administrator accounts. The EDR agent on the device sees the behaviour, scores it as high severity and raises an alert. The detection has already happened, and it took milliseconds.

Everything that decides the outcome happens afterwards, and that is where the two models part company.

Triage is where the paths split

Under EDR, that alert lands in a console queue and waits. Nobody is signed in on a Friday night, so the first person to see it is whoever opens the dashboard on Monday morning, by which time the account has been used from a second country. This isn't a failure of the tool. It's the predictable result of buying a sensor without buying attention.

Under MDR, a monitoring analyst picks it up within the agreed window, pulls the process tree and host context, and confirms it is hands-on-keyboard activity rather than a patch script a contractor left running. That decision requires someone who has seen both before.

A comparison infographic showing how EDR and MDR services respond differently to a live cyber attack scenario.

Containment needs authority, not just a button

Isolating a laptop is a click. Deciding to isolate the finance director's machine twenty minutes before a payroll run is a judgement call, and somebody has to own it. In an EDR-only model, the person on the console needs written authority to cut a senior manager off the network. In an MDR model, that authority is agreed in advance, so the provider isolates the host and disables the compromised account without waiting for a call back.

The NCSC's Windows platform security guidance recommends installing and configuring antivirus or an endpoint detection and response solution as part of device security. That framing is telling. The NCSC treats EDR as a control you configure on a device, not as a complete incident-management capability.

Investigation and recovery come next. Someone has to establish how the attacker got in, check whether other devices carry the same payload, and restore clean systems from backup. Then comes the reporting tail: the business owner, your insurer, and where personal data is involved, the regulator and affected individuals. Correlation across endpoint, identity and email logs is what turns five separate alerts into one story, which is why a managed SIEM service usually sits alongside whichever model you pick.

Write down who owns each of those five steps in your business today. Detection, triage, containment, recovery, reporting. Blank lines are what you're really buying.

Staffing and Skills Each Model Demands

EDR has an ongoing staffing cost, even when the licence price looks modest. Someone must tune detections to the organisation's devices, review alerts outside office hours, investigate suspicious activity, preserve evidence and coordinate communications with the board, insurer and regulator. Without people assigned to that work, alerts can wait until the next available staff member checks the console.

A cybersecurity team works together in an office, monitoring digital threats on computer screens during late night hours.

The Cyber Security Breaches Survey 2025 reports that 30% of businesses had formal incident-management capability, including 61% of medium-sized businesses. Having that capability is not the same as having a written incident response plan. For an EDR-only setup, the practical question is whether staff have enough time and authority to monitor, investigate and act, as well as maintain the procedures behind those actions.

What an EDR-only model quietly asks of your IT manager

In a mid-sized organisation, an IT manager may already be responsible for infrastructure, Microsoft 365, line-of-business applications and security. If there is no separate cover for nights, weekends or holidays, the organisation needs to decide who reviews alerts when that manager is unavailable. Otherwise, the working arrangement amounts to checking the queue whenever time allows.

Incident response also requires clear responsibilities, tested restoration procedures and an agreed approach to reporting. A procedure that has never been tested may not work under pressure.

Practical rule: if the person authorising isolation of a director's laptop is also responsible for explaining the resulting downtime to the board, agree who has authority before an incident occurs.

An outsourced monitoring service can provide out-of-hours cover and agreed escalation paths. Managing security operations describes the operational work that sits around security tools. The trade-off is that outsourcing monitoring does not remove the need to define decision rights, internal contacts and responsibilities for recovery.

Check three things in writing: who reviews a high-severity alert outside office hours, who can authorise isolating a director's device, and whether the incident plan has been tested. If any answer is unclear, the gap is in the operating model, not the endpoint software.

What Each Model Costs in Practice

Comparing an EDR licence against an MDR subscription is the wrong sum. The licence is the visible part of one model, and the service fee is the visible part of the other. What differs is where the labour sits.

The UK Government's Cyber Security Sectoral Analysis describes how managed security service providers typically deliver security protocols, monitoring, management and threat detection for a monthly or annual fee. That recurring delivery model is what separates MDR from a boxed product. The same analysis records service providers, including managed service and managed security service providers, generating approximately £8.4 billion in UK cyber-security-related revenue, up from £7.4 billion the previous year, with threat intelligence, monitoring, detection and analysis accounting for 45% of sector activity and incident response and management for 38%.

Where the money actually goes

Cost elements that differ between EDR and MDR

Cost elementEDR run in-houseMDR service
Platform or licencePaid to the EDR vendor per deviceUsually included in the service fee
Out-of-hours alert coverStaff time or on-call allowance, priced at your own loaded hourly rateIncluded in the service, subject to contract coverage hours
Alert triage and investigationInternal staff hours, priced at your own loaded hourly rateProvider analyst time under the agreed playbook
Tuning and maintenanceInternal staff hours to reduce false positivesProvider-managed, with agreed review cycles
Incident response supportInternal staff or an external retainer, quoted separately in GBPProvider-led response, with any surge work defined in the contract
Training and evidence handlingInternal training and record keepingEvidence retention and reporting defined in the service terms

Two rows do the damage. Out-of-hours cover and triage are continuous obligations, so their cost scales with nights, weekends and bank holidays rather than with device count. A forty-device estate still needs someone awake on Boxing Day.

A method you can run this month

Estimate the internal hours your team spends each month on alert triage, detection tuning and out-of-hours response. Multiply by a fully loaded hourly rate in pounds, including employer costs, not the salary figure. Add any on-call allowance and any retainer you already pay for incident support. Then put that total next to an MDR quote for the same device count and coverage hours.

Two rules make the comparison fair. Ask for every quote in GBP, and state your device count and required hours of cover in the request so providers price the same scope. If a figure can't be verified, such as an assumed number of alerts per month, mark it as unconfirmed rather than assuming it.

Watch the break-even: it moves with alert volume and with how many staff hours you have already committed elsewhere. A team with spare capacity gets more from EDR; a team already at capacity gets more from MDR.

Choosing EDR or MDR by Situation

The right answer depends on four things: the team you already have, whether anyone covers out-of-hours, your regulatory exposure, and the budget you can sustain rather than the one you can stretch to once. It isn't a religious question, and a provider who sells only one of the two is answering a different question from yours.

External help is already the norm in the mid-market. In 2025, 54% of medium-sized UK businesses used an external IT provider for cybersecurity, according to the UK Government's Cyber Security Breaches Survey 2025 as reported by IT Europa. If you already have that relationship, adding monitoring around your existing endpoint tooling is an extension rather than a new function.

A comparison chart showing when to choose EDR or MDR based on staffing, availability, and regulatory requirements.

Which model fits which situation

SituationBetter fitReason
Tested in-house security team with out-of-hours coverEDRYour team can act on alerts directly and already owns the response plan
Small business with no security specialistMDRNobody is available to watch or act on alerts overnight
Medium-sized business with an existing IT providerMDR or a hybrid modelThe provider can add monitoring around EDR without a new internal function
Large organisation with its own security operations centreEDR as the sensor, MDR for surge and out-of-hours coverInternal analysts keep control while the provider covers gaps
Charity holding sensitive records with a small IT teamMDREvidence handling, reporting and recovery need defined ownership

The mid-market hybrid problem

Many mid-sized businesses run a hybrid estate, with on-site servers integrated into cloud services. Coverage has to span both, because an attacker who lands on a server and pivots into Microsoft 365 using stolen credentials won't appear in an endpoint console alone. Any contract needs to name the endpoint estate, the identity logs and the cloud workloads explicitly.

We recommend solutions that suit the customer's needs, meet the customer's risk acceptance and fall within the customer's budget, which is why the conversation usually starts with your coverage hours rather than a product sheet. Where an organisation wants endpoint protection without a full managed service, Microsoft Defender for Business is available as an optional enhancement, adding endpoint protection, threat detection and data loss monitoring at desktop level. For medium-sized businesses our IT support is generally based on a hybrid cloud system that integrates existing on-site servers with the cloud, so endpoint and identity coverage are designed together.

A verdict is only worth the terms behind it, so check the contract items before you sign anything.

What to Check in an MDR Contract

Most disappointment with managed detection comes from the paperwork, not the technology. An alert that is acknowledged but not investigated has cost you money and bought you nothing.

A professional analyzing a legal contract with a magnifying glass while considering key business agreement clauses.

  1. Coverage hours. Confirm nights, weekends and bank holidays in writing. "24/7" in a sales deck and "24/7" in a schedule of service are different documents.
  2. Analyst-to-alert ownership. Name who picks up an alert and who signs off containment. Shared ownership means no ownership.
  3. Response authority. List the actions the provider can take without asking first, such as isolating a device or disabling a compromised account. If isolation needs your approval, you are the out-of-hours service.
  4. Escalation times and metrics. Ask for time to contain, not only time to acknowledge. An acknowledgement email proves someone read a screen.
  5. Evidence retention. Confirm how long logs and investigation records are kept, and whether they are usable for insurer and regulator reporting.
  6. Integration. Confirm coverage of Microsoft 365, Entra ID sign-in logs and your cloud workloads, not just endpoints.
  7. Reporting support. Confirm who helps with reporting to the ICO, to customers and to your cyber insurer, and in what timeframe.
  8. Recovery support. Confirm what restoration help is included and what is charged as surge work after the first 24 hours.

Items one to three belong in the contract as named obligations with times attached, not in an appendix. Ask for a sample incident report before signing, so you can see the level of detail you'd receive at 7am after a bad night.

We don't use "Response Time" service level agreements, because a response can be an acknowledgement email typed by a receptionist. The measures that matter are containment, evidence and recovery, which is why the terms above are worth arguing over before you sign a master service agreement rather than after the first incident.

Confirm Your Next Step Before You Buy

Four actions over the next month will settle this better than any vendor comparison.

  1. Log two weeks of alerts. Record every alert your current tools produce and note how many arrived outside office hours. That count is your exposure.
  2. Cost the internal hours. Tally triage, tuning and out-of-hours response per month, then price them at a fully loaded rate in pounds sterling.
  3. Map ownership and testing. Write down who owns each step of the incident process and whether the plan has been tested or only written.
  4. Shortlist two or three providers. Request quotes in GBP against the contract checklist above, with your device count and coverage hours stated.

If the out-of-hours gap is real and nobody is accountable, MDR is usually the safer option, but only once the contract confirms coverage hours and response authority. If you have a tested team with genuine night cover, EDR under your own control is defensible and often cheaper.

F1Group is an East Midlands IT partner providing managed IT services and cyber security. We can help you compare EDR and MDR options properly, and recommend solutions that suit your needs, risk acceptance and budget rather than the most expensive line on the sheet. Our staff hold enhanced DBS and CRB checks, so the people advising you are the people who would be accountable.


If the out-of-hours gap is real in your organisation, F1Group can help you price both paths on the same basis: we scope EDR, MDR or a mix of the two against your device count, coverage hours and reporting obligations, then recommend what fits your risk acceptance and budget. Call 0845 855 0000 today to talk it through, or send us a message and we'll put the options in writing. You can see the full range of our managed IT and cyber security work at F1Group.