Here’s the question I get more than any other, usually from a friend standing in the router aisle with two boxes in their hands: “Do I need one of these mesh things, or is a single router fine?”

It’s a fair question, because the boxes are designed to make you feel like the answer is always “buy more.” I spent eleven years shipping customer-premises equipment — the routers and gateways that go inside millions of US homes — to Tier-1 internet providers. I’ve sat in the meetings where we argued about whether a home needed one access point or three, and I’ve watched what actually happens when those devices land in real houses with real walls. So let me answer it the way I’d answer it for a friend, not the way a spec sheet wants me to.

The short version: for most normal-sized homes, mesh is not the first thing I’d buy. I’d start with one well-placed, quality router — and only add mesh if coverage proves to be the real problem. Mesh solves a specific problem, and if you don’t have that problem, you’re paying double for a downgrade. But there’s a real set of homes where mesh isn’t optional — it’s the only thing that works. The trick is knowing which house you live in. (If you’re still deciding whether Wi-Fi 7 is even worth it before you get to this question, start with my honest Wi-Fi 6 vs Wi-Fi 7 breakdown — this guide assumes you’ve already landed on Wi-Fi 7.)

The one-line answer (and the exceptions that matter)

If your home is roughly 2,000 square feet or less, on one or two floors, with normal interior walls, a single quality router placed centrally will cover it. You do not need mesh. Full stop.

You start needing mesh when one or more of these is true:

• Your home is larger than about 2,500–3,000 sq ft, or spread across three levels (basement, main, upstairs).
• You have dead zones a single router physically can’t reach — a detached garage office, a finished basement under concrete, a bedroom on the far side of the house.
• Your walls fight you: brick, plaster-and-lath, stone, or a foil-backed insulation layer that eats signal.
• The router has to live in a bad spot — jammed in a closet or at one end of the house because that’s where the ISP’s line comes in.

If none of those describe your house, keep reading just long enough to talk yourself out of mesh, then go buy one good router.

What a single router actually does — and the myth that sells mesh

A router broadcasts in roughly a sphere around itself. Range falls off with distance and with whatever the signal has to pass through. This is where the marketing trick lives: eight antennas look like more coverage. They aren’t, really. Extra antennas mostly help a router talk to more devices at once and aim signal more cleverly (beamforming) — they don’t meaningfully push the signal through three more walls. Going from a mid-range to a flagship single router buys you capacity, processing headroom, and faster ports far more than it buys you reach. You’re paying for what the router can do in the rooms it already covers, not for a dramatically bigger coverage bubble.

This is the single most expensive misunderstanding I see. People buy a $500 flagship router expecting it to blanket a 3,500 sq ft house, the far bedroom is still bad, and they conclude Wi-Fi 7 is a scam. A high-end router may hold a cleaner connection at the edge — better front-end design, antennas, thermals, and beamforming do buy you something — but it does not magically turn a far room through three dense walls into a strong-signal room. You’re mostly paying for capacity, processing headroom, radios, and ports, not a dramatic expansion of physics. The router was rarely the problem. Distance and dense walls were. That’s exactly the gap mesh exists to fill.

What mesh actually does (and the part the box doesn’t mention)

A mesh system is two or three (or more) units that blanket your home in overlapping bubbles and hand your phone off between them as you walk around, all under one network name. Done right, you get strong signal everywhere and no manual switching. That’s the upside, and it’s a real one.

Here’s the part the box is quiet about: the mesh units have to talk to each other, and that conversation costs you bandwidth. This is called backhaul, and it’s the whole ballgame.

If the units talk to each other over Wi-Fi (the default for most people), they’re spending part of the same airtime your devices need. In many real-world wireless-backhaul setups, especially on basic dual-band mesh, a satellite can deliver dramatically less throughput than the main unit — sometimes roughly half — because one radio is doing double duty. This is why a cheap two-pack mesh sometimes feels slower in the far room than a single good router did: you traded a weak-but-direct signal for a strong-but-taxed one.

Good mesh systems fix this by adding a dedicated band reserved just for backhaul (tri-band and quad-band systems), so your devices get the full pipe. That’s most of what you’re paying extra for when you go from a $200 mesh to a $600 one. It’s not magic — it’s a third radio.

The gateway engineer’s footnote: why ISPs don’t just ship everyone mesh

People assume that if mesh were strictly better, the carriers would put three pucks in every install. They don’t, and the reason is instructive. When you’re shipping equipment into millions of homes, every extra node is another thing that can fail, another support call, another unit drawing power and generating heat 24/7. ISPs ship a single capable gateway by default and only add extenders when the home actually needs it — because the operational math (this is the same cost-per-subscriber thinking that drives every CPE decision) says simpler wins unless coverage forces your hand. That’s a good rule for your own house too: don’t add nodes you don’t need.

The decision, the way I’d actually walk through it

Forget speed ratings for a second. The mesh-or-not decision is a coverage decision, and coverage comes down to four honest questions:

1. How big, and how many floors? Square footage matters, but vertical matters more — signal through a floor is worse than across a room. Three levels almost always wants mesh.
2. What are the walls made of? Drywall and wood: forgiving. Brick, concrete, plaster, stone, metal: signal killers. A 1,800 sq ft brick house can need mesh when a 2,800 sq ft drywall house doesn’t.
3. Where does the internet line actually come in? If your only option is to put the router in a corner or a closet, you’ve already lost half your coverage to placement. Mesh lets you put a node where the people are.
4. Where do you actually use Wi-Fi? If the dead zone is a closet you never sit in, ignore it. If it’s your home office or the kid’s bedroom, it’s a real problem worth solving.

Run your house through those four. If you come out the other side with one or two real, daily-use dead zones that a central router can’t reach, you want mesh. If you’re chasing a weak bar in a room nobody uses, save your money and place a single router better.

The thing that makes mesh genuinely good: wired backhaul

If you take one piece of insider advice from this whole article, take this: if you can run an Ethernet cable to where your second node will go, do it. A mesh node connected to the main unit by a wire (instead of Wi-Fi) gives you the satellite’s full speed with none of the backhaul tax, and the whole system gets dramatically more stable. Coax-to-Ethernet adapters (MoCA) can do this over the cable-TV wiring you may already have in the walls.

This is the difference between mesh that feels like real wired internet in every room and mesh that feels like a compromise. In the homes where I’ve seen mesh truly shine, it was wired. Where it disappointed, it was wireless backhaul fighting thick walls. Plan for wire if you possibly can.

If you decided you want a single router

Then you’re in the easier, cheaper world, and you don’t need this page — you need my picks. I’ve already laid out exactly which single Wi-Fi 7 routers are worth it (and which to skip) in Best Wi-Fi 7 Routers in 2026. Buy one good box, place it centrally and high, and pocket the difference. Come back here only if a far room stays stubbornly bad after good placement.

If you decided you want mesh: what I’d actually buy

These are the systems I’d put in a friend’s home, sorted by the situation they fit. I’ve deliberately matched the system to the house, not to the biggest number on the box. Prices move constantly, so I’m not quoting them — check current pricing at the link.

Read this before you buy a TP-Link Deco

If you’ve shopped mesh at all, you’ve seen TP-Link’s Deco line — the BE63 and BE85 in particular — and you’ve seen them top a lot of “best value” lists. On raw hardware-per-dollar, that reputation is earned; the Deco systems are capable and often cheaper than the equivalents above.

I’m not putting them in my recommendation boxes, and I want to be straight about why, because it’s not a performance complaint. TP-Link has been under sustained US regulatory and national-security scrutiny, including reported DOJ antitrust attention and state-level legal action. TP-Link has denied allegations that its products or user data are controlled by the Chinese government, and the company says its US operations and user data are managed in the United States. My concern here isn’t that a Deco system will fail tomorrow — the hardware is real. It’s that a home router is a long-life, firmware-dependent device that sits at the center of your network for years, and long-term vendor trust matters as much to me as the spec sheet. That’s the same lens carriers use when they vet a vendor for a multi-year program: the cheapest bill of materials doesn’t win if there’s an open question over the supply chain.

So this is a judgment call, not a verdict. If you buy a Deco knowing all of this, that’s a legitimate informed decision — I just won’t make it for you, which is why my picks above are eero, NETGEAR, and ASUS. (The regulatory side of this is a bigger story I’ll dig into separately.)

---

FAQ

---

The Bottom Line

Standing in that aisle, here’s what I’d tell you in thirty seconds. If your home is normal-sized, normal walls, and you can put the router somewhere central — buy a single good Wi-Fi 7 router, place it well, and don’t think about mesh again. If you’ve got a big or multi-level house, signal-killing walls, or a forced bad router location, mesh is the right tool, and the eero Pro 7 is where most people should start. Go up to the Orbi 970 only if you’ve got the square footage and the multi-gig plan to justify it, and reach for the ASUS ZenWiFi BT10 if you’re wiring your backhaul or want real control.

And whatever you buy: if you can run one Ethernet cable to your second node, do it. That single wire does more for real-world mesh performance than any spec on the box. The biggest number on the box is almost never the thing that matters in year three.

Next in this series: best routers for 1Gbps and 2Gbps fiber plans — matching the box to the speed you’re actually paying for, not the speed on the sticker.