What needs improvement?
The biggest issues within the product were that it had become stagnant. For about four or five years, there was very little real innovation going on.
It felt as if they were just sitting back. They were lacking in regards to keeping up with the developments within the cloud.
Overall, I think they had a good, solid product. I think they failed to add features. It was not as feature-rich as other products. I would say the biggest problem was the lack of features, they just hadn't kept up. Under Raytheon, they were starting to correct this, but it was a work in progress. Overall, the biggest problem with the product itself was the lack of features.
I knew that they needed to handle web sockets in some way, all we could do was effectively bypass it. There were too many times when the connections just didn't work right through the proxy. Our customers would have to bypass and basically go around the product. There were various levels to this and it was a real pain for our customers to diagnose those problems. There needed to be an end-product protocol analyzer output (for lack of a better way to put it), that would help administrators understand why the connection wasn't working.
There was so much legwork involved: someone would have to take a laptop and set up in front of the proxy; then they would have to load Wireshark (as I used to call it) and pull their captures; then they would have to give that to Forcepoint, or they would have to try to reason it out themselves. That caused a lot of problems because most administrations weren't confident or competent enough to do it.
They didn't have the skill-set needed to make proper use of those tools in the first place for analysis. There were a lot of customers who could've gotten value from the product but who were put in a position where they had to basically bypass the product because of certain connections. Some form of connection-troubleshooting should be included within the product, more than just looking at a log that nobody knows how to read except for support.
There was no way to troubleshoot connections in an effective manner that didn't require a lot of legwork by the user. Whenever you ask a user to do that, nine out of 10 times, they're not going to do it. They're just going to take the easy way out, bypass it, and then they'll bitch about the product, but they won't actually fix it. They won't want to make the extra effort. The problem just remains unsolved. They needed something like a connection analyzer tool to explain why, or at least give a better indication of why this was failing.
Again, it was the lack of development. The GUI is quite nice. I think it's very natural for people once they get used to it. Ironically, the company I'm working for now is actually POCing the DOV product and one of the things they like is the interface. They had a lot of good synergy with their other products. They failed to capitalize on it, ultimately. They're getting there. They got better, but it might be too little too late. That's the problem.
For how long have I used the solution?
I have been using this solution for nine and a half years.
What do I think about the stability of the solution?
Forcepoint Secure Web Gateway was pretty stable.
What do I think about the scalability of the solution?
Scalability-wise, It was pretty good. We had deployments that were easily over 10,000 — it could handle it if you scaled it properly with the right hardware. There were deployments that were 34,000 at that point. Now, with very large enterprises (with over 60,000 employees), that's where it starts to get a little tricky. They managed to get the product to that point but it was not cheap either, but it was achieved on the scale pretty well.
How are customer service and technical support?
The biggest issue that hampered Forcepoint was the lack of development and good consistent support. Because of this, they went from 50,000 to 60,000 customers, down to 16,000 in the space of about three years.
That's a huge drop. A lot of that came down to Palo Alto and a couple of others jumping in the game with filtering but at the same time, Forcepoint had people who were loyal, but they were giving up because of the support situation.
I constantly dealt with customer support. There wasn't a week that went by where I didn't have to talk to somebody.
They knew me by name and I knew them by name and we were all one big happy family, but when they moved from San Diego to Austin, they lost people.
They gave good, consistent support for about a year, but then they began losing staff because they weren't paying enough. They hired a bunch of people and claimed that they adjusted their wages, but then the same cycle occurred another year later.
That's been the biggest problem. It wasn't the product that caused them to lose people (even though they had some innovation issues), it was the support. At the end of the day, the innovation was causing them to look weak in customers' eyes; when you couple that with the support problems, that's when they started really losing customers.
How was the initial setup?
The initial setup was pretty straightforward.
In fact, you could literally load it, install it, and have it up and working in no time. All you had to do was point a browser to it and you could prove it worked. Now, if we wanted to take it and fully integrate it onto the network, then nine times out of ten, we used one particular method called WCCP.
What other advice do I have?
Overall, on a scale from one to ten, I would give this solution a rating of seven.
The product was sitting at about a seven. Support was dragging it down to a six. So, if I had to be honest, they had some very good analytics — just counting the product itself, I would say it's a seven.
It wasn't perfectly stable, but it was pretty good. I would say it would need to be more feature-rich. It would need to be more helpful for troubleshooting problems with connections and it would need more stability — then I would give it a higher rating. Under all of those things, it's the support that's the problem because that's a different question from the product itself — it's the Achilles heel.
Which deployment model are you using for this solution?