What is our primary use case?
Microsoft Defender that you get by default on Windows is an unmanaged solution. It detects, but it is conventional EDR in the sense that it can detect malicious code on the machine, but it is not good from an enterprise point of view because you can't see what is being detected. The difference between Defender and Defender ATP is that you get what's called the execution chain, which is its classic use case.
When I try to open an attachment to an email, Defender tells me that this is malicious, but when you are in an enterprise and you do receive an alert that the file is malicious, the problem usually for the analyst is that they don't know what the person clicked on. They know there was a malicious file but was it an attachment? Was it something on the USB stick? Did they download it from the internet? That's not clear. Defender ATP gives you the execution chain. In this particular example, you can see that it was outlook.exe that launched the suspicious file which then launched or tried to download various components. You can see the whole execution tree because very often, the initial thing you get is a dropper, which then downloads subsequent components, and very often, the subsequent components get missed.
It essentially gives you visibility into the execution chain. So, you are better able to do a risk assessment. For instance, if something came from Outlook, then you know that you need to go and look in exchange or look in the mail system. If the trigger came from winword.exe, then you know that it was a document, and the person had opened a document from the email. You might see Internet Explorer, when it was still there, spawn PowerShell or a command shell, which is unusual, or you might see calc.exe open a command shell. All of this detection is invaluable for identifying whether something is suspicious or not. Your EDR might not detect any of this, but ATP would see this suspicious sequence of opening and flag it. So, essentially it is the visibility and the ability to detect unusual behavior that conventional EDR would not necessarily do for you.
Its version is usually up to date. It is a cloud solution.
How has it helped my organization?
Its visibility is the most useful part of it, and it also increases the effectiveness of your response. You spend less time asking the users the standard question of what did they click on. To which, they usually say that they didn't click on anything. You can go in ATP, and you can see that they opened an email and then clicked on a link, and the link is this. There is no hiding this. Users do lie.
You can detect threats that are not necessarily known because of a behavior. If you have Internet Explorer opening a command shell, that is not normal. That does not happen unless there is some kind of malicious activity. It is also very good for visibility into what PowerShell scripts do. PowerShell is a double-edged sword. It is very powerful, but in a lot of cases, there is no visibility on what it is doing. With ATP, we generally have that ability.
What is most valuable?
I like the process visibility. This ability to visualize how something was executed is valuable, and the fact that Defender ATP is also linked to the threat intelligence that they have is also valuable. So, even if you have something that doesn't have a conventional signature, the fact that you get this strange execution means that you can detect things that are normally not visible.
The other feature that I like in Defender is that because it is up in the cloud, when you're trying to do any kind of managed service, it is fairly easy to set up if you're just within one tenant, but there are a lot of things wrong with the way Microsoft does it as compared to other products like Palo Alto Cortex, SentinelOne, or CrowdStrike.
What needs improvement?
The catch with ATP is you have to have the right Microsoft license. The licensing of ATP is linked to the licensing of Office 365. You have to have an E3 or an E5 license. If you have a small office license, it is not possible for you.
Another challenge is that it is not a multi-tenant solution. Microsoft's tenant is a licensed tenant. I'm an MSSP. So, I have multiple customers. In Microsoft's world, that means that I can't just buy an E5 license and give that out to all my customers. That won't work because all of the customer data resides within a single tenant in Microsoft's world. Other products—such as SentinelOne, Palo Alto Cortex, CrowdStrike, et cetera—are multi-tenant. So, I can have it at the top of the pyramid for my analyst to look into it and see all the customers, but each customer's data is separate. If the customer wants to look at what we see, they would only see their data, whereas in the Microsoft world, if I've got multiple customers connected to the same Microsoft tenant, they would see everybody else's data, which is a privacy problem in Europe. It is not possible to share the data, and it is a breach of privacy. So, the licensing and the privacy aspect makes it problematic in some situations.
It is also very complicated. If you decide to outsource your monitoring through an MSSP, the model for allowing the MSSP to connect to your Defender cloud is very complicated. In Office 365, it is relatively simple, but because of the way it has been done in Defender—because Defender is not part of the same cloud—it is a mess. It is possible, and it is workable, but it is probably one of the most complicated integrations we do.
It is still clunky as compared to products like Cisco AMP, SentinelOne, and CrowdStrike. Microsoft took the Defender product, and they bolted on the extra features, but you can see that there are different development teams working on it. Some features are well integrated, and some features are not. They keep on improving it, and it is better than it was. It is better than an unmanaged solution, but it is far from perfect.
For how long have I used the solution?
I have been using it for about two years. I've got a couple of customers today with it.
What do I think about the stability of the solution?
Its stability is lesser than some of the competition. I've seen machines having a blue screen. I've seen machines block, but it is usually a problem related to the lack of resources. I wouldn't deploy it on a machine with less than 16 gigs of memory. All the issues that we had on the laptops were essentially related to memory because it does all the analysis in memory, and it eats a lot of memory to do that. So, stability is more a function of making sure that your endpoint farm has what's available. If you've got less than 16 gigs, I would not recommend it. You need to either change your endpoints or consider using another solution because although it'll work, it can be very slow.
What do I think about the scalability of the solution?
It is like Microsoft Office. Its scalability is good, but I don't know how manageable it would be on a big scale. The biggest deployment I've worked on was about 5,000 endpoints, and it seemed to be okay.
How are customer service and support?
It is Microsoft support. It can be very good, and it can be very bad. It depends on who you get on the phone. I would rate them a five out of ten.
How would you rate customer service and support?
How was the initial setup?
It is very simple. You can deploy it through the normal tools that you use, such as SCCM. The deployment for it is linked back to your tenant.
We use it as a headless install. It is pushed out onto all the machines. Our normal rollout process rolls out about 50 to 100 machines in no time. They can pull the agents from the internet, or they can pull the agents internally, deploy them, and turn them on. For an antivirus, it is quite quick.
In terms of maintenance, it is pretty much like other Microsoft solutions. If you are able to do the auto-update functions, that's good. The downside to it is that it is fairly heavy on network traffic. On one of the large deployments, we found we had problems with the internet gateway because the console and all the telemetry and everything else is in the cloud. It was problematic.
It runs in the background. It is like any other antivirus solution. Sometimes, it needs tuning. An example would be that we have developers who do a lot of source code compiling. They might have tens of thousands of files that get touched or accessed when they do a compile. We have to make sure that those particular file types and certain directories are not scanned on read when they're opened. Otherwise, what normally might take an hour to compile can take more than 12 hours. That's not a problem specific to Defender. It is a problem in general, but it is fairly easy to create profiles to say that for those particular groups of machines or those particular groups of users, these file directories are exceptions to the scanning.
What's my experience with pricing, setup cost, and licensing?
The licensing fee is a function of your Office 365 license. The feature set you get is a function of the license as well. There is probably an E2 version, an E3 version, and an E5 version. There are several versions, and not all features are the same. So, you might want to check what features you're expecting because you might get shocked. If you only have an E3 license, the capability isn't the same.
You have to look at the total cost of ownership (TCO) because the license component is only one aspect of the block. So, if your internal IT teams know well about IBM cloud solutions, then Defender is very easy because there is nothing new. What hurts the projects is integration. It is a hidden cost because it is beyond licensing. It can be problematic if you don't have some of the other integration tools from Microsoft. So, if you don't have the package deployment platforms and all the cloud equivalents, then there is a lot of manual work involved.
The other aspect that comes into the cost is that there is an option to store. You can make the agents report a lot more information, but if you increase the storage, then you increase your Azure storage costs, which can be painfully expensive. You typically have about 7 to 30 days of basic detection data included, but if you want to keep a more detailed log so that your IT guys can go back and figure out what was going on, it would increase your storage requirements, and that can get expensive. I know customers who turned on some of the features to increase the detection rate, and they got a huge bill from Microsoft.
What other advice do I have?
A weakness, as well as an advantage, of Defender is that it is always on the cloud. There is no on-prem. You deploy additional agents into the customer infrastructure, but the console and the feedback are through the cloud.
Customers often say that Microsoft has included it in their license. So, it is license-cost neutral, but just because it is included in the license and appears to be cheap, it isn't necessarily a good reason for doing it. It isn't equivalent to other EDR or XDR solutions, but to an extent, you get what you pay for. ATP is a work in progress. To me, it is not a complete product.
Customers also go for it because it gives them visibility, and it means it is one less system to manage. They have the license for it, and they just want everything in the same ecosystem. There isn't much that we can do about that. As an MSSP, we're agnostic from a technology point of view. If the customer says, "This is what we want to do," we'll take it over.
I would advise asking yourself:
- What do your endpoints consist of?
- Which operating systems, such as Windows, Linux, iOS, or Android, will you have to support? The functionality that you get depends on your license.
- What is it that you're trying to achieve by taking Defender?
- Are there more capable XDR-type solutions out there?
If I was comparing them, from most effective to least effective or least integrated, I would put SentinelOne, Palo Alto Cortex, Cybereason, Microsoft Defender, and Cisco AMP.
If you want to get into the advantages of XDR solutions, which is about the detection capability coupled with artificial intelligence (AI) and data leaking, then it may not be the solution that you want. If you also want to be able to do threat intelligence, it is not the solution for you. That's because essentially the threat intelligence features are not there. You can get some threat intelligence from Azure, Microsoft Sentinel, etc, but it is not in the product like with Palo Alto Cortex, SentinelOne, or Cybereason.
I'd give it a cautious six out of ten.
Which deployment model are you using for this solution?
If public cloud, private cloud, or hybrid cloud, which cloud provider do you use?
Disclosure: PeerSpot contacted the reviewer to collect the review and to validate authenticity. The reviewer was referred by the vendor, but the review is not subject to editing or approval by the vendor. The reviewer's company has a business relationship with this vendor other than being a customer: MSSP