Wi-Fi 7 Speed Tests & Placement Guide

Quick start checklist (what you need + what good looks like)

To see real Wi-Fi 7 gains, you need a Wi-Fi 7 client on 6 GHz. With 320 MHz (or Auto) and a multi-gig WAN or LAN test, gigabit-plus throughput is normal. If you are seeing only ~300 Mb/s everywhere, treat it as a stability floor, not a success.

If your ISP tier is sub-gig, run a local LAN test (iPerf or WiFiman) to measure the router separately from the internet.

• Best Wi-Fi 7 mesh systems (2026)
• Wi-Fi 7 upgrade guide

Case study: one small move, big upload gains

In a plaster-and-lath colonial, moving the living-room node from a TV cabinet to an open bookshelf raised uploads from about 40–60 Mb/s to 120–180 Mb/s at the far couch. Calls stopped stuttering, and streaming DVR uploads finished reliably.

We started at 320 MHz on 6 GHz, then stepped down to 160 MHz for steadier mid-room results. We reduced 5 GHz to 40 MHz to cut overlap with a nearby office AP and labeled final node locations for future service.

Placement rules of thumb

• Mount or place high on open shelves; avoid behind TVs
• Center of home where possible; avoid corners and closets
• Keep roughly 30–50 feet between nodes in typical wood-frame homes; closer in plaster, brick, or metal-heavy layouts
• Use Ethernet for backhaul where feasible (2.5GbE/10GbE if available)

How to test room-by-room

Use a Wi-Fi 7 client and the same test server (or a local LAN test), then record near, mid, and far results in each room. Compare before-and-after small placement changes. Focus on consistency and upload stability for calls.

Run each test twice and average. Flag rooms where upload or jitter dips; that is where placement or backhaul changes pay off most.

Multi-floor testing (stairwells matter)

Stairwells act like vertical hallways, so test near the stairs and one room above or below. If speeds collapse between floors, your node placement is likely too far from the vertical path.

A small shift toward the stairs or a wired node at the landing usually stabilizes performance without adding more hardware.

Measurement protocol (repeatable results)

Verify the test device is Wi-Fi 7 and actually connected to 6 GHz (not 5 GHz). Test 6 GHz and 5 GHz separately when supported. Disable VPNs and background sync.

Use the same server (or a wired LAN test), run two passes per spot, and average. Record download, upload, and jitter. Favor stable upload and low jitter for video calls. Note wall materials between node and test spot.

Channel width guidance (performance first)

Start with 320 MHz on 6 GHz (or Auto) to unlock Wi-Fi 7 speeds. Step down only if range or stability suffer in your space.

• 6 GHz: 320 MHz first; drop to 160 MHz if results are unstable or the link is too short-range
• 5 GHz: 80 MHz if clean; 40 MHz in dense areas or for range
• 2.4 GHz: keep at 20 MHz for legacy devices

Evaluation thresholds (what good looks like)

Move a node or add one?

If one room is weak but the rest of the floor is strong, move a node a few feet before adding hardware. If multiple rooms fail the same threshold, you likely need another node or wired backhaul.

Adding nodes without wiring can make things worse, so confirm each new node improves results before committing to mounting.

Apps and tools we use

• Speed tests: Ookla, Fast.com, or local iPerf/WiFiman LAN tests
• Wi-Fi scan: Airport Utility (iOS), WiFiman/NetSpot (Android/desktop)
• Logging: spreadsheet with room, near/mid/far, notes, and after-move retests

iPerf local testing (optional but precise)

For LAN throughput independent of ISP, run an iPerf3 server on a wired computer (preferably 2.5GbE/10GbE) and point your test device at it over Wi-Fi. WiFiman’s local speed test can do the same with less setup. This isolates placement effects from internet variability and makes before-and-after comparisons clearer.

ISP tests vs LAN tests

Internet speed tests are useful, but they can hide local Wi-Fi problems if the ISP is the bottleneck (a 1 Gb/s plan will cap around 900 Mb/s). A quick LAN test with iPerf shows whether placement or backhaul is the true issue.

Use both and compare. If LAN is strong but ISP is slow, focus on modem, plan tier, or provider issues instead of moving nodes. If LAN is weak, fix placement and backhaul before calling the ISP.

Backhaul decision tree

If a room fails the upload target after placement tweaks, backhaul is the next lever. Ethernet is best; MoCA is a strong second choice when coax already exists.

For Wi-Fi 7, prioritize 2.5GbE or 10GbE links (Cat6/6a). A 1 GbE link can cap wireless performance even when the Wi-Fi is faster.

• Have Ethernet? Use 2.5GbE/10GbE for the primary and busiest secondary nodes
• Have coax? Use MoCA 2.5 with 2.5GbE adapters to stabilize the longest hop
• No wiring? Shorten the wireless hop and reduce channel width

Troubleshooting patterns

• Near-room strong but mid/far collapse: backhaul hop is wireless; wire it or reduce channel width
• Spiky results across the home: too-wide channels or overlapping nodes; tune power and placement
• Good download but poor upload: interference or contention; move node and retest

Common placement pitfalls and fixes

• Behind TV cabinets: move up and forward; avoid enclosures
• Too close together: add separation or reduce power to cut overlap
• Forcing 320 MHz on long hops: drop to 160 MHz (80 only if needed)
• Testing with a Wi-Fi 6 client: speeds will look capped

Checklist

• Pick test rooms and mark near/mid/far spots
• Confirm Wi-Fi 7 client and 6 GHz link; start at 320 MHz
• Run two passes per spot on the same server; average
• Move node a few feet and retest if mid/far dips
• Wire backhaul or use MoCA where a hop causes dips
• Document final node locations and settings

Retest cadence

Re-run the same test once per season, after major furniture changes, or anytime you change ISP speed tiers. Consistent testing prevents gradual drift in performance.

If you rely on Wi-Fi for work calls, keep a simple log so you can spot trends before problems show up in meetings. A short 10-minute quarterly check is usually enough for most homes. Include a quick upload test in your busiest room.

Document the final layout

Save a simple map with node locations, backhaul type, and channel widths. This makes future troubleshooting faster and keeps changes from eroding performance over time.

If you work with an installer, ask for a one-page handoff that lists SSID, admin access, and where cabling terminates. Include the test device used so future comparisons are apples-to-apples.

Recommended tools

FAQs