Your Router Is Not That Fast: How Wi-Fi Works vs. False Marketing

You must have experienced it. You got a new (possibly expensive) router, and it ended up failing your expectations in performance or coverage. Sometimes, it's a defective product. Often, though, it's your unreasonable expectations. To be fair, it's not entirely your fault.

Over the years, I've received plenty of emails and messages about how my Wi-Fi recommendations turned out to be duds simply because they could not match the false expectations. So much so that, since mid-2021, I have stopped offering recommendations on specific situations—no product can deliver the marketing nonsense.

This post aims to explain in simple terms how Wi-Fi works and the marketing language used by the hardware vendors, which is often repeated by online clickbait content disguised as "reviews". When through, chances are you'll know how to set your expectations straight, which is the first step in picking the proper hardware for your needs.

Let's start with how the Wi-Fi signal works.

Dong's note: I first published this post on February 4, 2024, and updated it on May 23 to add up-to-date relevant information.

Home router and your expectations: How Wi-Fi works vs. misleading marketing numbers

Wi-Fi uses radio frequencies, measured in Hertz, to transmit data from one party to another. It shares the same principle as any other technologies that use radio waves, including the radio itself.

One Hertz vs. lots of Hertz

To understand Wi-Fi radio frequencies, we need to know what constitutes one Hertz.

Heinrich Hertz is a German physicist who conclusively proved the existence of electromagnetic waves in the late 19th century.

As shown in the GIF below, in the simplest terms, Hertz is the number of radio wave crests—or wave cycles—in 1 second. How frequently a wave crests per second is its frequency. It's simple enough.

The higher the frequency, the closer the distance between two consecutive wave crests, which translates into a shorter length the wave itself can travel. However, in radio transmission, that also means the more information you can put on it.

FM and AM radio broadcasting stations use frequencies measured in megahertz (MHz), kilohertz (kHz), or even lower frequencies. At these frequencies, a broadcasting station can cover a large area, like a big city.

Signal coverage also depends on the station's broadcasting power. At the same broadcasting power level, signals travel further using lower frequencies than higher ones.

Traditional Wi-Fi broadcasters (routers or access points) use much higher frequencies measured in Gigahertz (GHz), including 2.4 GHz, 5 GHz, and 6 GHz frequency bands. Additionally, per regulation, they use no more than 1 watt (or 30 dBM) of broadcasting power. As a result, generally, a single Wi-Fi broadcaster can only blanket a modest home in physical size.

Wi-Fi is also available on two unique frequencies: the short-lived 60 GHz band (802.11ad), which has an extremely high bandwidth at an extremely short range, and the upcoming 900MHz band (Wi-Fi HaLow), which has mile-long ranges but extremely low bandwidth.

Real-life Wi-Fi visualization and interference

You can visualize how Wi-Fi, or any wireless radio transmission, occurs by dropping a rock in a still pond and watching the ripples move outwards on the water's surface.

The size of the rock and how hard you throw it equal the "broadcasting power".

Pick a particular ripple and count the number of times it reaches its highest point in one second. If it crests only once per second, you get one Hertz, twice equals two Hertz, and so on. That's the idea.

Radio wave crests can't be counted with the naked eye—we can't see them to begin with—and, as mentioned, Wi-Fi uses frequencies in GHz. For example, 5GHz means there are 5,000,000,000 wave crests in a second. So, I'd leave the counting to the science!

Now, if you drop another pebble at a different spot, that'd be your neighbor's Wi-Fi signal. Toss a rubber duck in the water! That's a microwave. See what happens when the ripples collide? Those are signal distortions—it's when your Wi-Fi signals drop, disconnect, or degrade.

Here's the thing: the pond was never entirely serene. Wind, insects, fish, debris, the liquid's viscosity, etc., are always there to affect the ripples. Similarly, visible and invisible stuff around us can adversely affect Wi-Fi signals. The point is that at any given time, there are more things in the air that hinder a router's Wi-Fi signals than those that don't. And there's always something in the air.

Signal distortions and degradations are part of radio transmission. As radio waves travel through the air, their integrity is degraded by distance and other factors—the fact that Wi-Fi works at all is remarkable.

The nature of Wi-Fi bands and their actual speeds

As mentioned, Wi-Fi generally uses three popular frequencies, called "bands", including 2.4GHz, 5GHz, and 6GHz. They have significant Hertz gaps and, therefore, are different in range and bandwidth. The links in the previous sentence explain things in great detail, but briefly, how Wi-Fi works depends on the band, specifically:

• 2.4GHz: This is the original Wi-Fi band with the most extended range but the lowest bandwidth. Since Wi-Fi 6, this band has never delivered real-world speeds of over 300Mbps, no matter what its specs indicate.
• 5GHz: This band has medium range and good bandwidth—over 10 times that of 2.4GHz. It became prominent with Wi-Fi 5 and has been the mainstay since Wi-Fi 6.
• 6GHz: This is the band with the shortest range and the highest bandwidth—about double that of 5GHz, depending on the Wi-Fi standards. It was first introduced with Wi-Fi 6E and much improved with Wi-Fi 7, which is the first Wi-Fi standard that incorporates all three bands.

Generally, a Wi-Fi broadcaster includes two first bands (Dual-band) and all three bands (Tri-band). You can also find tri-band or Quad-band, where the 5GHz or 6GHz is split into two to increase bandwidth.

However, from the client's point of view, a connection generally happens on a single band at a time—even with Wi-Fi 7's MLO feature. Most importantly, in a Wi-Fi connection, the lowest denominator, often the client, decides the connection speed, as shown in the max theoretical speed—the middle column—in the table below.

Standard (name)	Debut Year	Channel Width (in MHz) and Theoretical Speed (in Mbps) per Stream (rounded numbers)	Max Number Streams Used in Clients (Max Speed Theoretical(•) /Real-word)	Security	Bands	Status (in 2024)
802.11b	1999	20MHz/11Mbps	Single-stream or 1x1 (11Mbps/≈6Mbps)	Open WEP	2.4GHz	Obsolete
802.11a	2000	20MHz/54Mbps	1x1 (54Mbps/≈30Mbps)	Open WEP	5GHz	Obsolete
802.11g	2003	20 MHz/54Mbps	1x1 (54Mbps/≈35Mbps)	Open WEP	2.4GHz	Obsolete
802.11n (Wi-Fi 4)	2009	20MHz/75Mbps 40MHz/150MBps	Quad-stream or 4x4 (600Mbps/≈400Mbps)	Open WEP WPA	2.4GHz,  5GHz, Dual-band	Legacy
802.11ac  (Wi-Fi 5)	2012	20MHz/108Mbps 40MHz/217Mbps 80MHz/433Mbps	4x4 (1732Mbps/≈1000Mbps)	Open WPA WPA2	5GHz, Dual-band, Tri-band(••)	Common (Phasing out)
802.11ad (WiGig)	2015	2.16GHz/multi-Gigabit	n/a	Open WPA WPA2	60 GHz	Obsolete
802.11ax (Wi-Fi 6)	2019	20MHz/150Mbps 40MHz/300Mbps 80MHz/600Mbps 160MHz/1200Mbps	Dual-stream or 2x2 (2402Mbps/≈1500Mbps)	Open WPA WPA2 WPA3	2.4GHz 5GHz Dual-band, Tri-band(••),	Common
802.11axe (Wi-Fi 6E)	2021	20MHz/150Mbps 40MHz/300Mbps 80MHz/600Mbps 160MHz/1200Mbps	2x2 (2402Mbps/≈1500Mbps)	OWE WPA3	6GHz, Dual-band, Tri-band, Quad-band(••)	Common
802.11be (Wi-Fi 7)	2023	20MHz/225Mbps 40MHz/450Mbps 80MHz/730Mbps 160MHz/1.45Gbps 320MHz/2.9Gbps	2x2 (5800Mbps/≈3000Gbps)	OWE WPA3	6GHz, 5GHz, 2.4GHz, Dual-band, Tri-band, Quad-band(•••)	Common (Latest)
802.11ah (Wi-Fi HaLow)	2024	1MHz 2MHz 4MHz 8MHz 16MHz	(85Mbps to 150Mbps)	OWE WPA3	900MHz	Emerging

Again, often, the client limits the real-world performance of a Wi-Fi connection. No matter how powerful a Wi-Fi router is, the connection speed between it and a device is always that of the latter.

For example, if you use a 4x4 Wi-Fi 7 router, such as the Asus RT-BE96U, with a 2x2 Wi-Fi 6 client, such as the Intel AX200, the best theoretical connection speed you will get is 2402Mbps, with the real-world speed capping around 1500Mbps (or Gig+) at most, after overhead.

Additionally, in terms of range, a consumer-grade broadcaster of the three above bands can't have Wi-Fi coverage of more than 3000 ft² (279 m²) indoors, often much less.

The A-B-C of clever marketing language

Still, hardware vendors always use the broadcaster's rosy numbers to form marketing language that promises the impossible. They count on the fact that consumers would take things at face value and make assumptions about the rest.

All hardware vendors do this to some extent, but some are worse than others. Let's take Netgear, with its flagship Wi-Fi 7 Orbi 970 series, as a specific example. The screenshot below shows part of the product's webpage (which is subject to change.)

Pay attention, and you'll note a couple of tricks the vendor used to create false assumptions in consumers to lure them into purchasing the product. By dissecting this particular example, we have these standard A-B-C marketing tricks:

A. Ambiguity via the use of non-definitive terms

The vendor claims that the hardware can deliver (uncertain) high numbers in speed and coverage. Specifically, this 3-pack Orbi 970 Series can deliver "speeds up to 27Gbps" and "up to 10,000 ft² in Wi-Fi coverage."

Clearly, "up to" is an uncertain term, but users will likely overlook it and focus on the superimposed phony numbers. In case you're confused: If somebody tells you that they'd pay you "up to" $1000, they can give you a single dollar and still call themselves generous. (I'll do better: if you consume content on this website with some attention, you'll be up to 100% smarter!)

Where do these absurd numbers come from? That brings us to the next trick.

B. The impossible best-case scenarios presented as the collective norm

Often, the vendor implies the best possible range of the 2.4GHz band, which is low in bandwidth, as the hardware's overall coverage.

Technically, a 2.4GHz Wi-Fi broadcaster can indeed cover around 3000 ft² when used outdoors or in an empty warehouse. In real-world usage, its range is much shorter. And the ranges of the 5GHz and 6GHz bands are significantly worse.

However, hardware vendors' misleading norm is to suggest that the range of the slow band (2.4GHz) is that of the device as a whole. This practice becomes even more egregious when considering what the vendors do with the device's Wi-Fi bandwidth.

C. The combined the bandwidth of all bands and presented it as “speed”

When a Wi-Fi broadcaster has more than one band—Dual-band, Tri-band, or Quad-band—each of the bands has its own bandwidth and generally works independently from others.

In the case of the Orbi 970, which is a quad-band broadcaster, Netgear combines these bands' theoretical non-existent bandwidths—1147Mbps (2.4GHz) + 8647Mbps (5GHz-1) + 5765Mbps (5GHz-2) + 11530Mbps (6GHz)—to get the total bandwidth of 26909Mbps and rounds it up to form the 27Gbps number to present as the "speed" of the hardware.

It's important to note that bandwidth is quite different from speed, and Netgear's speed calculation above is false at many levels. Specifically:

1. The Orbi family uses one 5GHz band—the 5GHz-1—exclusively for wireless backhauling. As a result, the hardware needs to remove this band's bandwidth (8647Mbps) from the pool.
2. As mentioned, generally, a Wi-Fi connection happens on one band at a time. So, the best speed would be on the 6GHz band (11530Mbps). But even that's not true since:
3. The client limits the speed of any particular connection. Since we only have 2x2 clients, the best speed would be around 5800Mbps. But that is also not true because:
4. A Wi-Fi connection has massive overhead. The best real-world sustained Wi-Fi 7 connection I've experienced out of a 2x2 client topped out at around 3000Gbps, or 3Gbps, or one-ninth of 27Gbps.

So, the gist of these tricks is that the vendor, Netgear in this case, combines the bandwidth of all bands (2.4GHz, 5GHz, and 6GHz) as a massive speed and the range of the slowest band (2.4GHz) as extensive coverage to insinuate the false assumption that the hardware can deliver both of these impossible qualities simultaneously.

In reality, on a good day, you can get either the extended range or the high bandwidth (of a short-range band). To put things in perspective, you can be an impressive marathoner on one day and a fast sprinter on another. But it would be foolish to claim that you can sprint an entire marathon. Most of us can barely train for either.

As noted in the in-depth review, the Orbi 970's real-world performance is indeed a far cry from the numbers above. Scientifically, it had to be so.

It’s a BS Wi-Fi speed game of chicken

It's worth noting that 3Gbps is already extremely fast. It's so fast that most people don't know what to do with it or even have the right equipment to experience it, and Netgear could have used this honest number and remained impressive.

The problem is that the company (and other vendors) have gotten so deep into the BS game of Wi-Fi speeds for so long—since Dual-band became a thing—that they have had to keep finding new ways to make ridiculously large numbers to show progress and competitiveness.

Again, to put things in perspective, it's like once you happened to impress your date by digging a hole on a beach, you then got so consumed with digging deeper and deeper holes that you never realized it wasn't the hole but the fact you were moving around shirtless that was impressive the first time. The whole thing is comical if it's not pathetic.

And there's more. When looking at a Wi-Fi router's product page, you likely note the mention of 8K streaming, video conferencing, gaming, smart home stuff, and so on. They are there to create a relatable false sense of "quality."

Indeed. But the thing is, none of these items require a lot of bandwidth. You only need around 100Mbps for 8K streaming. Video conferencing and gaming depend more on the quality of your broadband, which the Wi-Fi broadcaster has no control over. (By the way, the Orbi hardware is, in fact, quite terrible when it comes to latency unless you use wired backhauling.) Finally, the best way to deal with IoT devices is by segmenting your network, which has little to do with bandwidth.

The point is this: Believe anything mentioned on the hardware's product page, and you'll be hugely disappointed. None was close to reality. Most are outright nonsense backed by the insidious notion of ambiguity.

How Wi-Fi works: The final thoughts

It's human nature to take things for granted and to possess confirmation bias. Online media and hardware vendors take advantage of that to further their interests. As Wi-Fi has improved over the years, we've expected more from it, and sometimes, it's easy to believe in magic. Ultimately, it's up to you to decide what is real.

When shopping for new Wi-Fi hardware, keep the following in mind:

• Almost everything claimed by the hardware vendor is false—always discount the numbers that follow "up to" by at least 60%.
• Your Wi-Fi bandwidth is, at best, as high as the real-world speed of the broadcaster's fastest network port. If it's a 10Gbps port, then you can expect around 7000Mbps after overhead.
• Your Wi-Fi speed is, at best, as fast as your client. If it's a 2x2 Wi-Fi 7 client—currently the fastest on the market—you'll generally get around 3Gbps of real-world speed in best-case scenarios—often much lower. (And 3Gbps is remarkably fast for any application.)
• Due to invisible elements in the air and distances, a Wi-Fi connection constantly fluctuates in real-world rates. It's never as stable or reliable as a wired connection via a network cable.
• No home's environment is ideal for Wi-Fi—vacuum space is best for radio signals, but you won't be able to last long, if at all.
• When mixing hardware of different Wi-Fi standards or performance grades, a connection's performance is always that of the lowest denominator.

Here's the bottom line: Don't expect Wi-Fi to suit your expectations, feelings, or wishful thinking. Instead, use the correct hardware on both ends of the connections as best you can.

Most importantly, don't fall for the fantastical numbers offered by networking hardware vendors via marketing or even the hardware's specs. Those are never true and, at best, are grossly exaggerated.