Cyber security platforms rarely stand still. Yet perceptions of them can.
For many organisations, their view of Microsoft Sentinel may still be shaped by how the platform operated several years ago, particularly around data ingestion, storage and the potential cost of bringing large volumes of security data into Log Analytics.
Those concerns were fair. But Sentinel has evolved.
For organisations reviewing their security operations or considering the future of an existing SIEM platform, relying on an old understanding of Sentinel can lead to decisions based on a model that no longer reflects how the technology is deployed today.
The short version for busy readersMicrosoft Sentinel is a cloud-native security information and event management (SIEM) platform. It now includes a data lake for longer-term security data, it is available in the Microsoft Defender portal, and after 31 March 2027 it will no longer be supported in the Azure portal. How well it fits your organisation depends on your data strategy, your architecture and your wider Microsoft environment. Table of contents |
One o the persistent concerns around cloud-native SIEM has been cost.
The logic is understandable. Security teams generate vast amounts of data across endpoints, identities, networks and cloud platforms. If every piece of that data is treated in the same way, retained in the same place and analysed at the same frequency, costs quickly become difficult to manage.
But not every security log has the same value.
Some data needs to be immediately available for analytics, detection and investigation. Other data mainly needs to be retained for longer-term investigation, compliance or historical purposes.
That distinction matters.
A more considered approach to data lets you focus higher-value analytics on the information that supports active security operations, while using different storage approaches for data that does not need the same level of immediate analysis. Sentinel's data lake provides another option for retaining large volumes of security data for longer, with Microsoft supporting retention of up to 12 years.
The conversation therefore shifts from:
"How much will Sentinel cost us if we put all our data into it?"
to:
"Which data actually needs to be analysed, how quickly do we need it and where should the rest live?"
That is a much more useful question.
It is tempting to think about Microsoft Sentinel the way you would think about a traditional security product: purchase a licence, switch it on and begin using it.
The reality is different.
Sentinel is a cloud-native security information and event management platform. The commercial and operational model is influenced by the data being collected, how it is used, how long it is retained and the wider Microsoft security environment surrounding it.
That means the quality of the architecture and data strategy can be as important as the technology itself.
Before making decisions about Sentinel, organisations should understand:
which security data is genuinely valuable for detection and investigation;
which sources need real-time or near-real-time analytics;
where longer-term data should be retained;
how existing Microsoft security technologies fit into the wider architecture;
what third-party technologies need to be integrated;
and how usage, retention and ingestion choices could affect ongoing costs.
For organisations already invested in Microsoft technologies, Sentinel should be considered alongside the rest of the estate.
Identity, endpoint, cloud, productivity and security data all contribute to the picture available to a security operations team. Sentinel is also available in the Microsoft Defender portal, helping bring security operations into a more unified environment.
The value lies in bringing the right signals together so analysts have the context they need to understand what is happening across the environment and respond appropriately.
That becomes more important as infrastructure becomes more distributed.
Security teams are no longer monitoring a clearly defined corporate perimeter. They are protecting users working across multiple locations, cloud applications, on-premises infrastructure, endpoints and third-party services.
The SIEM therefore needs to help teams connect those signals without creating an unmanageable volume of data, alerts and operational overhead.
There is a temptation in security to collect everything because it might be useful later.
But collecting more data is not the same as gaining greater visibility.
The more important question is whether the right information is available when analysts need it.
A large volume of poorly prioritised logs creates cost and complexity without materially improving an organisation's ability to detect or investigate threats.
A more mature Sentinel strategy starts with the security outcomes the organisation needs to achieve and works backwards.
What threats are you trying to identify?
Which systems and identities matter most?
What information would an analyst need during an investigation?
What data needs immediate analysis, and what simply needs to remain available?
Those questions help determine the architecture, rather than allowing the technology to determine it.
This is not just a theoretical consideration. When Royal Cornwall Hospitals NHS Trust moved from Splunk to Microsoft Sentinel, the Trust worked with Trustmarque Ultima to assess the value of different log types and how that data should be handled. The result was a more considered approach to ingestion, giving the security team greater flexibility over the data available for detection, investigation and threat hunting.
For organisations reviewing a legacy SIEM platform such as Splunk, this is particularly important.
A comparison based on Sentinel's capabilities or commercial model from several years ago may no longer provide an accurate basis for deciding what comes next.
Sentinel will not automatically be the right answer. It does deserve to be assessed against the organisation you have today, the security outcomes you need and the way the platform operates now.
A meaningful assessment looks beyond licence comparisons and feature lists. It considers data volumes, retention requirements, existing Microsoft investments, third-party integrations, operational processes and the capabilities available within the security team.
Only then is it possible to understand the potential technical and commercial implications of a move.
There is also a date to plan around. After 31 March 2027, Microsoft Sentinel will no longer be supported in the Azure portal. Customers using Sentinel there will move to the Microsoft Defender portal, making this a sensible time to review how your Sentinel environment is set up, what you ingest and where your data sits.
Microsoft Sentinel continues to evolve, but changing technology is only half of the equation.
Organisations also need to revisit the assumptions they have built around it.
If your perception of Sentinel is still dominated by concerns about putting huge volumes of logs into Log Analytics, unpredictable ingestion costs or simply replacing one SIEM licence with another, it may be time to look again.
The better starting point is the security outcomes you need, the data required to achieve them and how your existing technology environment can support that approach.
In January, Trustmarque Ultima will be hosting a webinar exploring Microsoft Sentinel and the considerations organisations should be thinking about as they review their approach to SIEM.
Pre-register for our Microsoft Sentinel webinar to be among the first to receive the session details and registration information.