Real User Monitoring and Visual Monitoring are both user-centric strategies to ensure that quality metrics important to users are maintained. Their underlying methodology, however, is different, leading to a separate set of advantages and disadvantages for each approach. The proactive nature of visual monitoring can help you discover and remedy problems before users even notice them.
Real User Monitoring (RUM) is a type of passive monitoring that tracks application availability and responsiveness by constantly observing a system using background services. A RUM system collects data exclusively from actual user sessions, which requires analyzing incoming internet traffic. This type of monitoring captures and analyzes each transaction individually as users interact with websites and applications.
Unlike RUM, Visual Monitoring doesn’t require any actual website visitors or application users to perform its tests. Instead it simulates human behavior by replaying the interaction paths (series of clicks, drags, etc.) end users take through that website or application. Obviously in order to probe your relevant interaction paths (or test cases), you have to identify those business-critical processes that you want to monitor.
Visual monitoring is a type of active monitoring, and since its interaction stream is artificially generated, the resulting traffic unlike with RUM is just a byproduct of testing. In other words, RUM infers results by searching through the entirety of user traffic and extracting only what it needs, while visual monitoring creates the specific traffic it needs and then measures just that traffic, so it doesn't need to store it long term.
Both methods collect data and measure application performance issues to ensure that potential problems don't reduce employee productivity and increase costs. But since RUM infers problems from inspecting real user traffic, by definition a problem has already occurred and the user has probably already noticed it.
In this case, visual monitoring tools are clearly more useful: because they have actively attempted to discover problems (by repeatedly running their test cases at regular intervals), they will be much more likely to find any problems before users do. It thus keeps many potential user complaints from ever arising because it can alert system administrators early, allowing them to solve issues before users can come across them.
That's not to say RUM is useless – for instance it's very helpful at understanding long-term trends. But if you stay awake at night worried that users will swamp the support department, you should definitely consider adding visual monitoring to your suite of tools. And both RUM and visual monitoring focus on the perspective of the user experience, where other more traditional monitoring techniques look at the performance and characteristics of individual services and hardware, apart from the user.
Visual monitoring, like other proactive monitoring strategies, can help you resolve problems before your users are even aware they are occurring. The overlap between its capabilities and drawbacks and those of RUM is quite complementary, so the best solution may be to combine real user monitoring together with visual monitoring.