The third challenge is more telling, and gets us closer to understanding the "fix." Assuming you hurdle the initial cost and complex deployment, there's a good chance you still end up unhappy with your new DLP system. Why? Two reasons: 1) your success relies on the imperfect knowledge of your programmers; and 2) today's DLP systems are incomplete. A DLP solution is simple in practice: you set up a rule with a trigger, and when the trigger is set off the DLP system denies the action. However, most of us are not omniscient and cannot anticipate all the rules that are needed to prevent data leakage. We don't have to look far for a good analogy. A bank wants to avoid money leakage, and thus installs a lock and an alarm on its door, and then cameras that record. Why have the camera if you have locks and an alarm? Because you don't know what you don't know, so if you're system is foiled you want to be able to replay the video tape to see who did it and most important, how they did it. Current DLP systems do not include this "record" functionality and thus provide clients no feedback loop for adjusting rules and, most importantly, no forensics when a breach does occur.
So, in effect, most DLP buyers soon discover that they're taking a shot in the dark and hoping for the best.
The answer to these problems lies in the architecture and the packaging. Let's start with architecture. Current DLP solutions have been designed as network-based systems, which make them expensive to buy, complex to deploy, and blind to anything that occurs outside of the rules set by their human administrators. Further, they're protecting data near its source, not where it's being used, which is where the greatest risk of breach lies. An endpoint DLP solution gives businesses complete visibility and control of the users' behavior since the endpoint is where all the action is. An endpoint agent can allow you to record all computer activity (whether or not a rule is triggered), creating a feedback loop administrators can use to perfect policies over time, as well as providing a play-back feature if and when there is a breach.
Now, the packaging. Delivering the endpoint solution in a Software-as-a-Service (SaaS) model solves the more obvious and easier problems of cost and deployment complexity. With a SaaS solution, businesses effectively eliminate the upfront purchase and deployment costs.
To be sure, protecting data wherever it resides or is in use -- whether in storage, on the network or at the endpoint -- should be a critical objective for any business. But if DLP providers want those 20%-plus growth rates to go on forever, they not only need happy enterprise clients, they ultimately need SMBs to buy the vision. And to make that happen, the endpoint is the place to be.
Brad Miller is CEO of Awareness Technologies.