The licence was fine.
Then you added a switch.
Monitoring is one of the few things you buy where the bill is driven by how much of your own estate you decide to look at. That shapes behaviour, usually badly.
Most monitoring is licensed per sensor, per device, per module or per network size, and the model matters more than the initial price because it decides what happens as you grow. Per-sensor pricing counts every individual metric, so the bill rises with monitoring depth as well as estate size. Per-device is more predictable but still grows linearly. Per-module puts the capabilities you actually need behind separate purchases. The failure mode common to the first three is the same: the cost of looking more closely creates a reason to look less closely.
How each one behaves as you grow.
Per sensor
A sensor is one thing being watched — an interface, a service, a metric. A single switch can easily be twenty. The count grows with depth as well as breadth, so monitoring something properly costs measurably more than monitoring it superficially.
The tell: teams stop instrumenting access switches. The blind spot is created by the pricing model, not by an engineering decision.
Per device or per node
Simpler to predict and easier to explain to finance. Still linear in estate size, and it can penalise you for the many small devices that are cheap to own and countable all the same.
The tell: debates about whether a device is worth licensing, which are really debates about whether to have visibility.
Per module
The platform is a base plus purchasable capabilities — flow analysis here, configuration management there, packet capture as a separate product. Each is reasonable on its own; the total rarely resembles the first quote.
The tell: the capability you needed is in a different SKU than the one you evaluated.
Per network size
Priced on the shape of the network — ports, WAN lines, installations — with everything the platform does included. The bill does not move when you monitor something more closely, which is the point.
This is how Net-Monitor is licensed: three variables, quoted per deployment, no separate modules.
Four things to establish before signing.
- 1
Get the quote at 150% of today
The current price tells you almost nothing. The slope between today and three years from now is the actual decision.
- 2
Count what "one device" costs you in practice
For a sensor-based model, take a real switch and count what fully monitoring it consumes. Multiply by your access layer. That is the number to compare.
- 3
List every capability in a separate line item
Flow, configuration management, capture, reporting, high availability. Ask which are included and which are not, in writing.
- 4
Ask what happens at renewal
Support uplift, version entitlement, and what the price does if you have grown. Renewal is where surprises usually arrive.
How Net-Monitor is licensed
- Three variables: ports, WAN lines, installations. Quoted per deployment, because those three describe the network rather than the monitoring.
- Everything is inside the licence. Discovery, mapping, traffic analysis, configuration management, the sniffer and the AI. No separate modules to buy.
- Monitoring something more closely does not cost more. Which means the depth of your visibility is an engineering decision rather than a budget one.
Short answers.
What is per-sensor licensing?
A sensor is one monitored element — an interface, a service, a specific metric on a device. A single switch commonly needs a dozen or more to be monitored properly. Because the count grows with how thoroughly you monitor as well as with how many devices you have, the model prices depth of visibility rather than estate size alone.
Why does network monitoring cost so much?
Usually not because of the headline licence but because of how it scales. A model that counts elements grows with the estate, capabilities frequently sit in separate modules, and renewal uplift compounds. The quote for today's network is rarely the number that matters; the shape of the curve is.
What are the hidden costs?
The common ones are modules that turn out not to be included, per-element counts that grow faster than the device count, storage for anything retained, high availability licensed separately, and the engineering time spent tuning a system that produces noise. The last is the largest and never appears on a quote.
How is Net-Monitor licensed?
On three variables: how many ports are on the network, how many WAN lines you run, and how many sites or data centres. Everything the platform does is inside that — there are no separate modules and no per-element count that climbs each time you add a device. Pricing is quoted per deployment because those three numbers differ so much between networks.
Related questions.
Want to know what yours would cost?
Three numbers — ports, WAN lines, sites — and we can give you a real figure rather than a range.