More Eero nodes can make your Wi-Fi worse.

You bought a second Eero (or a third) because Wi-Fi was weak in a back room. Coverage improved in that room. But now throughput across the whole house feels slower, devices disconnect and reconnect more often, video calls drop mid-sentence, and the Eero app keeps shuffling devices between satellites unprompted.

This isn't your imagination. It's mesh oversaturation, and it's well-documented in the Eero community though almost never acknowledged in the marketing. The same pattern shows up across all wireless mesh systems — Eero is the most common case only because it's the most common system.

Three mechanisms compound when too many wireless mesh nodes share the same space:

1. Wireless backhaul shares spectrum with client traffic. When your Eeros don't have ethernet between them, every packet a satellite serves to your laptop first travels from the gateway to the satellite over the same radio that's then serving your laptop. The satellite radio handles both — the upstream link to the gateway and the downstream link to clients — and the two compete. In a 3-Eero setup with wireless backhaul, the satellite farthest from the gateway loses roughly half its capacity to backhaul overhead before serving any clients.

2. Overlapping coverage forces unnecessary roaming. If your laptop sits in a room where two Eeros are both visible at decent signal strength, the roaming algorithm constantly re-evaluates which Eero to anchor to. Each switch breaks an active connection for a few hundred milliseconds — long enough to drop a packet in a Zoom call, long enough to interrupt a download. Adding nodes increases the count of locations where this fight happens.

3. Co-channel interference grows with node count. Eeros default to running their satellite radios on the same Wi-Fi channels as the gateway. In a three-node mesh, those radios all transmit on the same channels in overlapping coverage zones. They're not jamming each other — Wi-Fi has collision avoidance — but they are sharing airtime, which reduces aggregate throughput. The 2.4 GHz band is hit harder than 5 GHz because there are fewer non-overlapping channels available.

Three signals together make a confident diagnosis:

1. Throughput on devices close to the gateway is also worse than expected — not just on devices near the far satellite. A pure coverage problem only hurts the far devices; an oversaturation problem hurts everyone.
2. The Eero app's network health view shows satellites with "weak" or "fair" backhaul links between them, even when they're physically close to the gateway. Backhaul quality visualizations in the Eero app — under each satellite's status — tell you whether your satellites are seeing each other well, or whether the satellites themselves are contending.
3. Devices spontaneously switch satellites under load. Open the Eero app, find a device that's been connected for a while, look at "Connected to" — then run a bandwidth test on that device and check again. If the device migrated during the test, the roaming algorithm is fighting your use case.

The fix order, from least disruptive to most:

1. Remove the satellite you added most recently. Counterintuitive but often correct. If you have three nodes and the third was added for coverage in a room that's still adequately served by node 2, take it out. Re-test the whole house. If throughput improves on near-gateway devices without breaking the far room, you've found the oversaturation point.
2. Move the gateway Eero closer to where activity actually happens. Many homes put the gateway next to the cable modem in a basement or closet. That optimizes for the modem; it pessimizes for everyone upstairs. Moving the gateway upstairs and using the basement Eero as a satellite often fixes coverage without adding nodes.
3. Ethernet-backhaul any satellite you can reach with a cable. Running a single Cat6 cable (affiliate) from the gateway to one satellite eliminates the wireless- backhaul problem for that satellite entirely. The difference is dramatic — the wired satellite goes from contending with the gateway for spectrum to acting like a second independent access point.
4. If you can't run ethernet but you do need multiple satellites, use a wired-backhaul mesh instead. The ASUS ZenWiFi XT12 (affiliate) and TP-Link Deco BE65 (affiliate) both ship a dedicated 5 GHz or 6 GHz backhaul radio that doesn't share spectrum with client traffic. They're more expensive than basic Eeros but solve the core oversaturation problem at the hardware level.

Mesh oversaturation is real, but it's not universal. Adding more nodes is the right answer when:

• The house is large enough that there genuinely isn't overlapping coverage between nodes — typically 3,000+ square feet, especially with floors involved.
• You can ethernet-backhaul the new nodes. Wired backhaul defeats most of the oversaturation problems and turns the satellites into independent access points.
• The Eero Pro 7 or Eero Max 7 — the higher-end models have a tri-band radio architecture with a dedicated backhaul channel, which makes wireless backhaul much more viable than on the base Eero or Eero Pro 6.

If none of those describe your setup, removing a node will often outperform adding one.

Eero's hardware is good. The mesh algorithm is conservative — designed to favor stability over peak performance — and that's usually a defensible choice. But the marketing positions every Wi-Fi problem as a "buy another node" problem, and that's not always true. Sometimes the right answer is one fewer node, or a wired backhaul, or a router placement change.

The Eero app makes it easy to add satellites and surprisingly hard to see the cost of doing so. The "network health" view shows a green checkmark when individual links are strong, not when aggregate throughput is healthy. Diagnose the symptom — slow throughput, dropped calls, device ping-pong — not the marketing-friendly metric.

Once you've right-sized your mesh, run the StabilityPulse stability test from a device in the room that was previously slow. Throughput recovery is one signal; the more interesting one is whether loaded latency drops back to two-digit range. Mesh oversaturation often manifests as latency spikes under load rather than as throughput cap, and the stability test will catch what a speed test won't.

For sibling reading, the per-device Wi-Fi diagnostic covers the case where the slow device is the variable rather than the network. Different fingerprint, different fix.