Every Mesh Node You Add Makes Your Network Worse. The Only Question Is Whether You Notice

Keith Parsons · 2026-08-11

A consumer tech piece is again pitting mesh Wi-Fi kits against traditional routers, and selling the comparison on square feet of coverage. I think the people who get anything out of it are those who are selling the mesh kits, not the people who are buying them.

If you add another mesh point, your throughput will be cut in half. It does give you more coverage, but at a cost. That is the sentence the box should carry, and it is the one nobody prints.

What would an honest number on the box say?

Every mesh node makes things worse. The more mesh nodes you add, the worse it gets, so it is a cumulative effect. Adding one mesh node cuts your max throughput to one half. Adding two mesh nodes cuts that to one quarter. Adding three gets it to one eighth. It is an exponential function.

Adding 1 mesh node cuts your throughput. Your max throughput in 0.5. Adding 2 mesh nodes cuts that to 0.25. Adding 3 gets it to 0.125. It's an exponential function.

So it starts hurting at the very first node, but you might not care. If you have 100 gig backhaul and that is not your bottleneck, then maybe mesh is not a problem. The penalty for buying mesh is based on what you need for capacity and download speeds, and not based on anything else.

Should a 12-person dental office buy a three-node kit instead of pulling Ethernet?

In a dental office you do not have a lot of need for speed, so this might be a cheaper alternative than pulling Ethernet. Just realize that your customers will have less throughput. Again, you might not have a need for more throughput. If you want long term, I would go with Ethernet and then non-mesh APs, because you will have a better solution for the long term. But it does cost more.

What happens in an apartment building where every unit runs three nodes?

It is just totally trashed. Especially in MDUs, when you are in a multiple dwelling unit, you are close together, and the RF goes through walls. Adding mesh units increases the amount of traffic in the air, lowering airtime, and thus lowering everyone's ability to transmit data to the Internet, which is what they really want. You need to pull cable, not mesh units.

How would somebody in that building even tell that is what is wrong?

You could use any of a variety of tools to look at what the airtime is currently being used for. I like Wi-Fi Explorer Pro from Adrian Granados, and it allows me to look and see what the current channel utilization is, just on my phone with a spectrum clip attached to the back. That number, channel utilization, shows how much free time is available. If there is not much free time, there is not much traffic going to be sending.

Isn't 6 GHz backhaul finally good enough?

It might be good enough for mesh if they do not have 6 gig clients. If you have 6 gig clients and they are trying to also use the 6 gig frequency that is being used for mesh, that can be the problem. So the statement is true as long as you are not talking about 6 gig clients. Yes, 6 gig mesh is fine, and it does not mess with your 5 gig. That is a true statement. But it does mess with your 6 gig clients.

What should corporate IT be able to see on a home mesh network?

It does not really matter to corporate IT, as long as the person on the other end is receiving the information at a speed and latency and ping times that they would want. I recommend using a service called Orb. The Orb client can run on the remote worker's laptop or phone or tablet and tell them whether the slowdown is on their end or the other end.

For my customers, I have their remote workers leave Orb on their dashboard on the screen all the time. So if they are in the middle of a Zoom call and something is having trouble, they can just look up and see whether this is my network or someone else's.

What do you do when the mesh gear is already installed?

When I have to deal with mesh gear, I do one of two things. I use MoCA, which is Ethernet over coax, and most houses today still have coax, and that solves the mesh problem. Or I set up a point-to-point wireless bridge inside their house that skirts around the mesh. Many times we do not have to remove the mesh at all. We just put something around it to make sure that one remote user's location, his home office or whatever, has the best possible connection they can.

By making our changes, the channel utilization dropped. So the kid downstairs watching Netflix, or those doing their homework, do not interfere with the worker trying to get his Zoom call to work properly. You just have to move those onto separate networks.

Do this

• Find a tool that shows your channel utilization, and check it. If your channel utilization is low, mesh is not your problem.
• If Zoom calls are dropping or gaming is lagging, get the traffic you care about off the mesh.
• Pull Ethernet where you can, and use non-mesh APs for anything you intend to keep for the long term.
• Use MoCA over the coax that is already in most houses, as an alternative to pulling new cable.
• Put in a point-to-point wireless bridge for the one location that has to work, and leave the mesh in place around it.
• Have remote workers run Orb and keep it visible, so they can tell whose network is at fault.
• Move the heavy household users, Netflix and homework and gaming, onto separate networks from the work device.

Mesh's problem is that it uses up airtime. If you do not have a lack of airtime, then mesh is an inexpensive solution. If you do, all three of those fixes will augment a mesh system and offload the traffic you actually need to work.