Planning a Mesh System Before You Spend a Cent

The 3-pack of a mesh system costs substantially more than the 2-pack, and nothing on the box tells you which one your home needs. The square-footage rating printed on the side assumes a friendly building: few dense walls, one floor, nothing metal in the path. Your home is probably not that building.

You can settle the question on a floor plan in about ten minutes, before you order any hardware.

Pack size is a geometry problem, not a square-footage one#

Mesh kits are sold by coverage area because area is the only thing a manufacturer knows about your home. It is the wrong unit. Two homes of 2,000 sq ft can need one node and three nodes respectively, depending on whether the interior walls are stud-and-drywall or solid masonry, and whether the router sits near the middle or in a utility closet by the front door.

Node count comes down to one question: how many places in your home have both a strong link back to the router and a clear path into the room that failed?

A satellite needs to hear its parent well — as a rule of thumb, somewhere in the -55 to -65 dBm range for a backhaul that is not itself the bottleneck. Past about -70 dBm the satellite still lights up green and still publishes full bars to your phone, but it is relaying a degraded link, and every device behind it inherits that degradation.

So a third node is either genuine extra coverage or an expensive paperweight, and which one it turns out to be depends entirely on whether such a spot exists in your home. That is a question about layout, and a floor plan is the only place you can settle it before the box arrives.

Know what is actually in the box#

In most systems, one unit becomes the router and the rest become satellites. So a 2-pack is one router unit plus one satellite, and a 3-pack is one router unit plus two satellites.

That matters when your ISP equipment has to stay where it is — a fiber ONT, a cable modem, a VDSL modem. You either put that box into bridge or modem-only mode and let the mesh unit route, or you keep the ISP router in charge and spend one mesh unit doing little more than being a client of it.

The second case is common, and it quietly changes the math: a 2-pack sitting behind an ISP router you cannot bridge buys you one satellite's worth of new coverage, not two. Work out which situation you are in before you compare prices, because the real comparison is price per useful node.

Build the plan that answers the question#

The whole exercise runs on your phone. Drawing the plan, standing the router in it and reading the baseline heatmap are free; revealing the recommended spot and the before/after result are the paid part.

1. Sketch the floor plan. Pace the rooms out rather than guessing; proportions matter far more than exact dimensions. Drawing a plan the simulation can use walks through the whole step.
2. Set the material preset to the way the house is built, then nudge the wall and floor loss values if it is heavier than that preset assumes. This is the step people skip and the one that moves the answer most. A brick chimney running up the middle of the house can add a node all by itself.
3. Place the router where it really is — inside the closet, on the floor, behind the TV, whatever the truth is. Planning from where you wish it were produces a plan for a different house.
4. Read the baseline heatmap and write down which rooms fail before you add anything. Some homes discover here that they need no nodes at all, just a better router position.
5. Nominate the room that matters most and run the placement search. It weighs the spots a node could occupy and ranks them by how much each one lifts that specific room.
6. Open the before/after comparison on the winning position and note how many dB the priority room gained.
7. Change the priority room to your second problem room and run the search again. This is the step that answers the pack question.
8. Leave the plan on the phone. It keeps one saved draft, so Continue Plan reopens this one when you are standing in a store, or when a sale tempts you into a bigger box than you need.

The two-search test that settles 2 versus 3#

Run the search once for each room you actually care about, then compare where the two winners landed.

The size of the lift matters as much as the position. A gain of 3 dB is inside the noise of any prediction. A gain of 12 to 15 dB is the difference between a room that buffers and a room that does not. Reading a before/after comparison honestly covers what counts as a meaningful change and what is wishful thinking.

When a winning position is inconvenient#

The search will point at positions, and some of them will be awkward. Two constraints decide whether an awkward position is still the right one.

Each satellite must hear a parent, not just exist. A node placed past the useful range of the one before it does not extend the network; it republishes a slow link with a confident-looking status light. If the only candidate spot for a third node sits deep in the shadow of the second, the third node is not buying coverage.

Nodes in a long home should form a chain, not a ring around the router. In a long or L-shaped layout the second satellite usually belongs partway between the router and the first satellite's territory, not on the far wall. The node placement rules of thumb go into spacing, hop limits and node-to-node targets in detail.

When the honest answer is not mesh#

Four cases where the coverage map should send you away from the checkout.

• One bad room, everything else fine. A whole mesh system is an expensive way to fix a single bedroom. One well-placed extender or one cable does the same job for a fraction of the price — you can plan the extender position the same way.
• Full bars, slow speeds. If the failing room already reads a healthy signal, more radios will not help. Confirm you have a coverage problem before you buy coverage; what a dead zone actually is covers how to tell the two apart.
• The router is in the worst possible spot. Moving it is free and often worth 5 to 10 dB across the far half of the home. Do that first, re-run the plan, and see whether you still need anything.
• A concrete slab or masonry core that nothing gets through. When no candidate position lifts the room by more than a few dB, the answer is a wire. Coax, MoCA and a single Ethernet run usually beat three wireless nodes, and one wired access point beats all of them.

There is no shame in the plan telling you to spend less. That is the plan doing its job.

From coverage map to shopping list#

Once the map has given you a node count and positions, the rest of the buying decision is mechanical.

• Node count comes straight off the two-search test.
• Power. Is there an outlet within a few feet of each winning position? A node that belongs in the middle of a hallway wall with no outlet is a plan you cannot execute.
• Backhaul band. If a satellite lands two or more rooms from the router, a tri-band system with a dedicated backhaul radio earns its extra cost. If every satellite lands one wall away, a dual-band kit is fine.
• Ethernet ports on the satellites. If a winning position is next to the TV or the desk, that node doubles as a wired switch for the device that needs it most.
• Wired option. If any winning position happens to sit near a coax outlet or a route a cable could follow, wiring that node lifts the ceiling for everything behind it.

Decide on the map, then order#

Everything above happens inside one app. Range Up takes the layout you draw, the material preset you set for the build and the router's true position, and returns the coverage your home has right now. Name a failing room and it will weigh the spots a satellite could occupy and set its favorite against that baseline. Name the second failing room and repeat. Two winners in the same corner mean one satellite; two winners at opposite ends mean two; and that is your pack size decided before a courier is involved. It predicts from the building rather than reading your network, and the plan stays on the phone — no account, no cloud upload, no scanning. Try it before you order the box.