It's kind of difficult to quantify areas for improvement. In the larger picture, one challenge is that the NDR space is very crowded today. I can mention half a dozen names just off the top of my head. There are at least 12 to 20 different players. All of them are well-known brand names, and it's difficult to compare them. They all claim to be giving you the same network difference capability: catching malware, dealing with all the minor taxonomy of attack, all that. Still, it's very difficult to compare them side by side because they all do things a little differently, and they all have different presentations and output. We haven't deployed it, so I can't give you what we felt about it exactly. But in the larger perspective, the critical feature is really giving a clear separation between a low, high, and medium criticality. You need a rating that is really true to the actual attack. There's one other capability we are evaluating them for, and it's for custom alerts detection. A lot of these products are trying to profile the threats that are already out there in the industry. They're very well known and published. Today, there are targeted acts being played against organizations, so you have to be sensitive to how your firewalls, protocols, and your HTTP are all operating. You might have some fine-tuned threats that are targeting you, and you should be able to build custom defenses. They should have some openness in terms of how you specify your threats. You get a standard library of threats. On top of it, every organization builds its own.
