Millimeter wave 5G hasn't had an easy time of it, as a technology. It's been panned for fickle connectivity, extremely limited coverage, and being too thinly-deployed to even meaningfully use (not to mention the absolute nonsense conspiracies some people are spreading about it). It doesn't help that mmWave only has one truly active booster in the United States—Verizon. The company says its existing plans for 5G mmWave deployments around the country haven't changed in light of the COVID-19 pandemic, but the current crisis raises a valid question: if we won't be gathering in large groups for the foreseeable future—let alone in stadiums or concert halls—does mobile mmWave even make sense anymore, and did it ever?

In answering that question, we have to look forward. We will one day return to the sporting and music venues that we packed in The Before Times, also known as 2019. The NFL will host matches with tens of thousands of fans, all of them using their phones to share photos and video of the big game, utterly crushing our feeble LTE networks in the process. To be clear, there are use cases for mobile mmWave, and situations where its uniquely massive capacity is able to carry a much greater data burden than traditional 4G or even sub-6GHz 5G networks. Even massive Wi-Fi installations frequently struggle at such immense scale, making mmWave 5G a uniquely suited technology. But much as we could solve the problem of battery life by fitting every phone with a 7000mAh cell, or the issues of reception by slapping on big external antennas on them, there's a cost:benefit analysis we have to acknowledge both consumers and businesses alike perform when it comes to new tech. And right now, the costs of mmWave remain very high, while the benefits seem lower than ever.

Little waves, big dollar signs

On Verizon, mmWave's expense is easily quantified. The Galaxy Note10+, launched last year, was priced $200 more for a mmWave 5G variant ($1300) than the 4G unlocked version ($1100) sold by Samsung. Verizon's LG V60 ThinQ 5G ($950) carried a $150 premium versus T-Mobile's mmWave-less V60 ($800) at launch. The mmWave version of the OnePlus 8 ($800) Verizon exclusively stocks comes with a $100 markup over the standard OnePlus 8 5G on T-Mobile ($700). Verizon still doesn't even sell a version of the base model Galaxy S20 because it's waiting for Samsung to ship a special—and once again exclusive—mmWave-equipped S20 so that the carrier can advertise it as 5G-ready. At this time, we don't know if it will be more expensive than the standard and already rather expensive S20 5G. No doubt, part of Verizon's problems here are economies of scale: mmWave smartphones are being produced in comparatively tiny quantities, making absorbing the additional costs of the hardware, development, and certification more difficult for OEMs. As Techsponential handset analyst Avi Greengart puts it, "Even though Qualcomm is including mmWave 'for free' with its top two Snapdragon chipsets [the Snapdragon 765 and 865], vendors still need to incorporate additional antennas and circuitry. They also need to alter the designs in some cases... [And] getting 5G working on wildly different networks requires expensive testing and network certification."

We've seen relatively few mmWave smartphones to date, and none have cost less than $800 at launch.

These high costs have meant we've seen relatively few mmWave smartphones to date, and none have cost less than $800 at launch. So, where are the down-market mmWave phones? Qualcomm implies they're coming, as it currently offers mmWave in the cheaper Snapdragon 765 platform, and says it plans to bring mmWave to the Snapdragon 600 family later this year. But when I asked if Qualcomm had secured any smartphone design wins utilizing mmWave on its Snapdragon 600 or 700 5G platforms, the company declined to comment. (For what it's worth, I think they may have a handful at this point—I doubt the number is zero.)

The Galaxy S20 Ultra features mmWave 5G. It'll just cost you $1400.

Expense aside, the biggest stumbling block to mobile mmWave adoption in smartphones seems to be that nobody else is meaningfully deploying it in the way Verizon has. AT&T and T-Mobile have hedged their bets on 5G by moving more quickly on repurposing existing low-band spectrum for broader coverage across the US. While both do have mmWave 5G deployments in various places around the country, their commitment to mmWave as a smartphone technology remains limited. To date, AT&T and T-Mobile have launched just three 5G mmWave smartphones on their networks, all of them built by Samsung: the Galaxy S10 5G, the Galaxy S20+, and the Galaxy S20 Ultra. Other 5G handsets like the OnePlus 7T Pro McLaren 5G, LG V60, OnePlus 8, Galaxy S20, and Galaxy Note10+ 5G have only supported these carriers' sub-6GHz 5G networks, leaving "full" 5G support to the most expensive parts of the portfolio. Neither AT&T nor T-Mobile have a mmWave-capable handset that costs less than $1200—think about that. PCMag's lead mobile analyst (and, if you ask me, the best 5G reporter around) Sascha Segan thinks that's telling. When I asked Segan if he saw AT&T and T-Mobile guiding their handset partners away from mmWave to drive down phone prices, his sentiment was clear, "Yes, I think that especially T-Mobile will focus on sub-6-only handsets this year so as to keep phone prices down. We've seen that with OnePlus and with the smaller Samsung S20, and we'll see that going forward."

This makes the economics of Verizon's pricey mmWave handsets seem potentially untenable, now more than ever as they're thrown into stark relief against a badly damaged American economy. Greengart tells me that even absent COVID-19, Verizon had a tough sell on mmWave as a feature: "There just aren’t enough places to use mmWave to justify a premium. With the economic impact of COVID-19 pushing super-premium smartphones out of reach for many, Verizon is almost certain to have too much inventory to sell at today’s prices." And that's exactly what's happening: Verizon has already slashed $400 off the MSRP of its LG V60, a brand-new mmWave phone in the super premium segment. Greengart believes that, overall, US consumers are just going to start spending less money on smartphones as the economy slows down—and he has a great analysis detailing just why that is, and how consumer preferences will change.

Covid-19 is decimating large segments of the economy. While there are undoubtedly some people who got stimulus checks that supplemented secure incomes, tens of millions of Americans are already unemployed or underemployed. Even more are likely to be unemployed in the coming months as travel, events, and entertainment will recover slowly until a vaccine or cure is fully deployed...

Hopefully, this disruption will not last that long, but for the duration of the recovery period, consumers who have lost income will be forced to keep their current devices indefinitely. When those phones do inevitably fail, many will be replaced by sub-$250 phones because their owners have no choice.

Those who are still employed but are experiencing economic insecurity will be looking for better values. This is where the mid-tier market opens up: saving “just a few dollars a month” will now be necessary. In the past these buyers would have stretched their budget without hesitation, but with appealing options now coming to carriers from Apple, Samsung, and others, they will be much more willing to compromise - because it won’t feel like as much of a compromise.

Relief for high 5G prices could be coming later this year and in 2021, though. Everyone I spoke to for this article basically agreed: the cost of 5G has to come down. When I asked if Qualcomm was likely to walk back price hikes with the Snapdragon 865's successor (expected to be announced at the end of 2020), Segan said "Certainly. The 865, with the non-integrated modem, is almost certainly more expensive to produce than an integrated chipset would be, and Qualcomm's lock on millimeter-wave antenna modules has helped them drive up prices. Next year we should see more integrated chipsets." Greengart said that Qualcomm had the clear power to lower prices, "Sacrificing margins for volume is a strategy that can be implemented without any design, engineering, and manufacturing lead times."

In an interview with PCMag, OnePlus CEO Pete Lau also says he's expecting the cost of implementing 5G, including mmWave, to come down over time. When I spoke to Qualcomm for this article, the company strongly implied that we could expect the broader forces of scale and technological advancement to make 5G more economical, but declined to say whether the 865's successor would be cheaper. But cost isn't millimeter wave's only challenge.

The capacity conundrum (or: mid-band madness)

Verizon alone has licenses for over 1000MHz of mmWave bandwidth in key markets.

mmWave's big promise lies in reducing congestion on the crowded, comparatively narrow networks we're stuck with right now. And carriers have paid good money to make real that promise, bidding over $10 billion so far on the FCC's three mmWave spectrum auctions. They're not paying out the nose for no good reason. That reason is bandwidth, and mmWave presents a lot of it: Verizon alone has licenses for over 1000MHz of mmWave bandwidth in key markets (for comparison, most LTE bands in the US are 10-40MHz wide). But the reason they're paying quite so much is debatable, and I believe it has far less to do with mmWave being the most desirable electromagnetic real estate than it does with our federal government mucking up our 5G strategy from day one.

The United States faces an especially crowded market for wireless spectrum, featuring dozens of incumbent and legacy bandwidth users like the military, broadcast and satellite TV, GPS, radio, plus various marine and aeronautical systems. The chart below is one of my favorites, because it illustrates just how ridiculously, crazily crowded wireless is in the US. It's absolutely nuts.

Note: This is an old chart (2016), and doesn't reflect things like recent consolidation of broadcast TV in the 600MHz band.

You may notice a big hunk of pink and purple in the second row of that image, and carriers have noticed it, too: it's called the C Band, and it sits from around 3700 to 4200MHz here in America. This spectrum was leased to satellite companies ages ago, and much of it is being used in an incredibly inefficient manner today. The FCC has taken its sweet time, but it's finally prying 280MHz of C Band away from the satellite cartel in an auction at the end of the year, where AT&T and Verizon are expected to be exceptionally active bidders.

This "mid-band" spectrum is extremely valuable stuff, because radio waves in that relatively low part of the EM spectrum propagate (travel through the air and around obstacles) reasonably well. In aggregate, a 5G network operating in the 3700-4000MHz range could easily achieve many square miles of outdoor range using only existing cell tower sites, even if that coverage may not prove especially effective at reaching indoors, and might need to be augmented with additional sites for proper "blanket" coverage. But its particular value is not only in that range, but the capacity relative to that range: a 5G network operating with 100MHz of bandwidth could easily achieve 200-300 megabits per second of end user connectivity, and over time as the equipment becomes more capable and technologies like 5G carrier aggregation come into play, it could get a lot faster. Millimeter wave promises much quicker speeds than mid-band 5G—well in excess of a gigabit—but its range is far, far more limited, especially relative to that impressive peak performance. As a result, mmWave has proven vastly more difficult to scale to the kind of saturation coverage our big cities and suburbs demand. And yet, mmWave spectrum is the only new spectrum the FCC's made available during 5G's critical first stages.

This is precisely why Verizon has acted so aggressively on mmWave, because it literally has no other choice.

The way we've handled this in America—prioritizing mmWave over freeing up more mid-band spectrum—was a royal screwup, according to Segan, and I fully agree: "Mid-band frequencies are absolutely the best way to install first-generation 5G. The US's original sin was that our regulators took the coward's way out and couldn't face down C-Band incumbents, so we didn't do a mid-band 5G auction and instead went for mmWave, which was much less used." And he says the fallout of that decision is going to continue to haunt us, "We still don't have enough dedicated mid-band for a good nationwide multi-carrier 5G strategy, and we won't until the C-Band auction." With the C Band auction scheduled for December 2020, that means any spectral spoils carriers acquire won't yield new network deployments until Q1 2021—at the earliest. And this is precisely why Verizon has acted so aggressively on mmWave, because it literally has no other choice: It's a bit like saying your dog's favorite food is dog food, because it's what he always eats. If you gave him his choice of a raw prime ribeye or his standard kibble, you might find that assumption challenged.

Verizon has a mmWave OnePlus 8—but this case won't fit on it. They had to move the buttons for the antennas.

Once these auctions close, industry consensus is that the big winners will start deployments quickly, be it Verizon or AT&T (or both). Segan says he sees Verizon "going all in with the C-Band auction at the end of this year, to get that spectrum and then roll out a sub-6 5G network on C-Band." At that point, my suspicion is that Verizon's hard-charging mmWave "UWB" 5G marketing largely goes out the window and gives way to some other exciting acronym they'll unveil shortly thereafter. AT&T is also in rather acute need of mid-band spectrum, though its vast and heavily-aggregated LTE network still tops many national speed tests, making that need a bit—if only slightly—less urgent than Verizon's.

There is one mid-band player in the US market, though: the new T-Mobile. The freshly merged supercarrier has embarked on a vast expansion of its newly acquired Sprint 2.5GHz mid-band spectrum, and anticipates it will deploy at least a thousand mid-band sites a month throughout 2020. The combined mid-band holdings of Sprint and T-Mobile are vast: in many major metros, the new T-Mobile controls up to 160MHz of such spectrum, enough for a seriously speedy network. And that coverage is proving effective—in Philadelphia alone, T-Mobile says its mid-band coverage exceeds that of Verizon's entire national deployment of mmWave by a factor of 2.5. The power of propagation is hard to understate, and it's something Verizon and AT&T are going to struggle to market against until they have shiny new mid-band holdings of their own.

Verizon is telling consumers 5G is here, but COVID-19 has ensured that even for those people in covered cities, it may as well not be.

As network testers and smartphone reviewers start to smell what the new T-Mobile is cooking (i.e, as mid-band coverage spreads), I think you may see the narrative tide truly start to turn against mmWave in a big way: T-Mobile did 5G right, Verizon did 5G very wrong, and mid-band is the way forward. Of course, this isn't to say Verizon is at fault for the choice it made; it really had very few other viable options. If the FCC could have freed up more mid-band spectrum a year ago, Verizon's 5G rollout would almost certainly look very, very different today. But I think the damage being done to mmWave 5G as a result of Verizon's heavy marketing of its UWB 5G network is real. Verizon is telling consumers 5G is here, but COVID-19 has ensured that even for those people in covered cities, it may as well not be (let alone for the vast, vast majority for which it isn't here period). The pain could get worse for Verizon, too: top Apple analysts now believe the company's mmWave iPhones are at risk of delay, possibly by months. That means Verizon could be on an uneven playing field this fall, forced to wait for its mmWave models to ship while other carriers go ahead sub-6GHz only. It could also choose to offer iPhones without the company's signature 5G marketing feature. The latter would put tremendous pressure on Verizon to own up about how realistic its mmWave plans are, and just how committed to mobile mmWave it is.

Fixing wireless

When I talk to people a lot smarter than I am about mmWave, one of the things I always hear is "fixed wireless." For the uninitiated, the consumer-facing implementation of this concept is simple: mmWave 5G home internet. And to me, it represents the easiest way for carriers to convince investors that their costly mmWave licenses weren't just panic purchases driven by a weak FCC unwilling to act decisively on 5G. Those billions of dollars sunk on mmWave for mobile are unlikely to be written down just yet, but all of America's carriers have expressed interest in fixed wireless applications for the technology. From Segan: "In a country with poor home Internet competition and relatively lazy incumbents, there's room for the wireless carriers to do multiple-play deals. Verizon, T-Mobile and US Cellular have all said they have home Internet strategies coming up, and millimeter-wave may be the way that gets carried in urban and suburban areas."

mmWave actually does make a good amount of sense as a broadband solution, if not in every single city or situation. The technology's limited range and fickleness with obstacles can more easily be mitigated in a fixed application for two big reasons. First, none of the connected devices (antennas attached to homes or other buildings) are actually moving. This makes it much easier to "point" signal where it will be reliably received. Second, fixed home wireless can operate at significantly higher power outputs, and with much larger antennas, than your smartphone. Together, these factors substantially reduce the challenges for which mobile mmWave 5G has been so maligned. In a typical scenario, a mmWave base station at some distance would transmit to a mmWave antenna on your roof or in a window, which would connect back down to a Wi-Fi router inside your house (as in, your phones and other electronics still connect to the internet via Wi-Fi—not mmWave).

Verizon's 5G home internet: a mmWave antenna in your window connects to a Wi-Fi router to act as your ISP.

As a business opportunity, fixed wireless is highly attractive to carriers. Verizon is the first to offer mmWave home internet to any significant group of customers, but that business hit a snag due to modem delays on Qualcomm's end, which Verizon says won't resolve until close to the end of the year. Still, the opportunity is an easy one to see: building residential fixed wireline fiber internet is exorbitantly expensive, so much that Verizon ended new deployments of its national FiOS broadband a decade ago. Challenges for mmWave at home remain, however, as even things like rain, fog, or strong wind can significantly attenuate signals, potentially requiring more dense deployment of transmitters (not to mention topography, which tends to be a far greater concern)). That means costs could vary considerably depending on region before even considering the question of minimum viable subscriber density.

There's also the rather unintuitive fact that mmWave fixed wireless deployments have no synergy with mobile mmWave: the two technologies, while very similar, are being deployed on a totally independent basis. Right now, no mmWave base station exists that can both provide fixed wireless broadband access and high-speed mobile mmWave access, because there just doesn't seem to demand for one. That may sound baffling, but it's down to the differing economics and deployment considerations of fixed wireless mmWave versus mobile mmWave. When you're deploying base stations for a network that provides saturation coverage for moving smartphones—on the ground—that's going to look very different from one designed to reach the roofs of a suburban neighborhood equipped with loads of fixed, high-powered antennas.

It's entirely possible mmWave could find a place as the backbone of digital infrastructures in various relatively novel deployments, even if it really doesn't end up mattering very much as a data connectivity feature in our smartphones.

Fixed wireless mmWave's applications also extend far beyond the home, and that's where the big use case may arise, according to Segan, "I'm still a big believer in the value of 5G quality-of-service in wireless factory and medical applications, for instance, as well as in smart city and public safety applications. There are a lot of applications where it's annoying to use WiFi because of interference and WiFi's poor quality-of-service guarantees. (Think about how much more unreliable WiFi calling is than LTE phone calls are.) mmWave's true use may be as a private, secure super-WiFi for corporations and governments, where bandwidth can be reserved for specific applications in a way you can't do with WiFi." And I agree with him here: there's still just so much we don't know about the advantages mmWave will provide in these contexts, and more importantly, how compelling they'll be to the parties concerned. It's entirely possible mmWave could find a place as the backbone of digital infrastructures in various relatively novel deployments, even if it really doesn't end up mattering very much as a data connectivity feature in our smartphones.

mmWave as the Wi-Fi killer

The questions of Wi-Fi's continued scalability and capacity aren't without merit, but didn't it just get an unprecedented, massive expansion of capacity? The extra 1200MHz of channels now being opened up to Wi-Fi and other unlicensed uses like 5GNR-U (stay with me here) are going to unleash a huge swath of bandwidth, and Wi-Fi remains the dominant standard among most commercial businesses and venues, small and large. That seems unlikely to change. Wi-Fi is an "any device" technology, in that it's deployed in basically anything with a screen. That makes the barrier to entry far lower than a private mobile network, both on cost and complexity. I asked Anshel Sag, a 5G and chipset analyst at Moore Insights, if Wi-Fi's new grazing area would potentially depress the demand for mmWave. He thinks even if it could initially, it won't indefinitely: "I don't think that Wi-Fi 6e will be the golden bullet to all of the congestion problems of Wi-Fi. It will alleviate them, but many will just move onto 6GHz as soon as they can."

Wi-Fi works with everything, and its much lower frequencies mean it is exceptionally easy to achieve near-100% coverage.

Still, the more I dug into 6e, I simply couldn't find evidence of common consumer use cases mmWave would solve that greatly expanded Wi-Fi bandwidth would not. The single biggest advantage mmWave would have over a massive Wi-Fi network is, ironically, one its greatest weaknesses: range. In ultra-dense venues like sporting arenas, mmWave's short range, coupled with its ability to "point" signal using beamforming, would result in reduced interference with other mmWave base stations inside that venue (since it's rare the signals from two nodes would ever meet). This would allow significantly improved scalability when compared to Wi-Fi, which does a lot of talking over itself by design in order to provide continuity of coverage. But the sense I got as I researched was that, at best, even in venues where mmWave would make sense from a scale perspective, it makes less sense from a coverage and cost one. For one, if you want to potentially cover every user (and realistically, only a subset of them), every carrier has to come into your venue and install mmWave base stations, which is already extremely difficult from an RF design perspective, let alone the challenge this presents for the business owner economically. For another, Wi-Fi works with everything, and its much lower frequencies mean it is exceptionally easy to achieve near-100% coverage. At best, it appears that mobile mmWave 5G could augment existing Wi-Fi coverage inside some stadiums and other large venues, but there is simply no compelling evidence to suggest that it will replace massive commercial Wi-Fi deployments.

At best, it appears that mobile mmWave 5G could augment existing Wi-Fi coverage inside some stadiums and other large venues, but there is simply no compelling evidence to suggest that it will replace massive commercial Wi-Fi deployments.

The list of problems with mmWave as a Wi-Fi replacement doesn't stop there. Unlike Wi-Fi, which can be deployed easily and relatively cheaply by businesses of any size, mmWave comes with all the strings and costs of working with a wireless service provider. There are much more complex FCC compliance issues to consider, and a relative paucity of vendors—Qualcomm remains the world's only mmWave modem supplier. When will will that change? We don't know. It also seems that situations where mmWave's capacity provides a meaningful quality of service benefit to consumers over Wi-Fi are rare, and when a technology's use case is defined so narrowly, that's going to bode poorly for adoption. And at that point, if you're a business owner, why would you support something 90% of your customers don't even have? These are questions that nobody seems to have a good answer to.

Dreams of distributed access

Given what I've told you about mmWave so far, it should not surprise you to learn that I haven't even covered all of the challenges associated with it. One of the biggest among them is backhaul (basically, your cell tower's hard line out to the internet). Right now, every single mmWave base station, of which dozens are needed to cover even a very small neighborhood, must provide its own backhaul connection to your carrier's infrastructure. That means every single one of those little transmitters must be hardwired not just for power, but for internet. Doing this exponentially increases the challenge of building out an already ludicrously challenging technology, and it's a serious stumbling block for mmWave.

Right now, every single mmWave base station, of which dozens are needed to cover even a very small neighborhood, must provide its own backhaul connection to your carrier's infrastructure.

According to Sean Kinney, Editor-in-Chief of the wireless trade publication RCR Wireless, there are two technologies that offer promising solutions to this issue: integrated access backhaul (IAB) and mmWave distributed relays (mDR), "Integrated access backhaul [IAB] can help quickly densify--that's coming in Rel. 16, expected in RAN products 2021. The thing with IAB is it would solve your backhaul problem but you still need a site and power."

Both IAB and mDR were new concepts to me when I started working on this article, but I believe I can give you a layman's interpretation that's relatively simple to understand. IAB would allow multiple 5G mmWave base stations to talk to one another in much the way a mesh Wi-Fi network does: a 5G node could provide internet access without dedicated backhaul by transmitting wirelessly to a second 5G node that does have backhaul. Together, groups of these nodes would be able to form a larger "mesh," providing connectivity in a far more easily scaled manner without the need for site-by-site backhaul. The second solution, mDR, would seek to solve this problem on a much more localized scale, by helping those nodes see around corners and get into buildings. Through a system of relays (it's just like it sounds), a mmWave node would be able to get signal around corners, into alleys, or otherwise out of line-of-sight areas. These relays would be simpler and smaller than a full-on mmWave node, acting as transmitters and receivers—more like a home Wi-Fi range extender than a mesh network.

Image courtesy Intel via RCR Wireless

The real-world viability of these solutions is still up in the air, though as Kinney says, IAB is at least in the next upcoming 3GPP spec, expected to be finalized early this summer. Both could make the economics of mmWave more attractive, but just how much more is a bit of an unknown. Until carriers commit to concrete deployment plans and get this equipment on the ground (or a utility pole, or a street sign, or a window), it all still exists in the realm of seemingly-pretty-good ideas. mDR in particular is a pretty new thing, and one which is still in the very early stages.

The immediate problem I see with a solution like mDR is cost: the number of repeaters and access relays necessary, as well as the know-how to position them in the most efficient way, sounds like a troubleshooting and equipment nightmare. I'm not saying it's impossible, but it sure does make you wonder whether it's worth doing—a mmWave distributed relay system needs two relays just to get around a single 90-degree corner. Multiply that problem by ten thousand alleyways, and the deployment costs have got to start really adding up. We're talking about a system that would need far denser distribution than even blanket coverage public Wi-Fi.

An economy lacking scale

If you're still with me, I promise we're closing in on the end here. Carriers' investments in these mmWave networks, Verizon in particular, are in part premised upon the idea that the demand for handsets supporting this technology are going to drive up their average revenue per user, and also that they will increase the competitiveness of the business in the larger carrier landscape. While these investments are based on multi-year plans and the deployment of technology that may not even meaningfully exist at the time spectrum licenses are issued, there's still an expectation of ROI: carriers didn't pick up this spectrum for fun, they did it to make money. Verizon's pitch is much more heavily leveraged on the ultra-high-speed mobility angle for mmWave, and it has adopted a pretty aggressive posture in proclaiming 5G supremacy in the US. But posturing only gets you so far, and investors are really only concerned with what you can put on a balance sheet when all's said and done.

I'd argue that while technological barriers like scaling and distributed access do pose a challenge for mmWave, the current economic situation could pose an even greater one. The COVID-19 pandemic has led unemployment in the US to approach Great Depression-era levels, and while we have good reason to believe many people out of work today will return in the coming months and year, it's clear the damage to the economy is growing. In a highly advanced economy heavily driven by consumer spending, as opposed to industrial output or agriculture, widespread unemployment is bound to cause consumers to cut back. Expensive wireless bills are a place they may start: Verizon reported losing over 500,000 net postpaid subscribers during quarterly results last month, while AT&T squeaked out a little over 150,000 net adds. T-Mobile was the apparent beneficiary, netting almost 500,000 postpaid phone subscribers, perhaps in significant part because of its highly competitive pricing on multiline family plans and frequent handset promos (subscriber bleed from Sprint probably helped a fair bit, too).

If carriers like Verizon are unable to convince consumers that investing in a pricey new mmWave-ready handset is a good idea, these networks start to lose some of their reason for existing.

If carriers like Verizon are unable to convince consumers that investing in a pricey new mmWave-ready handset is a good idea, these networks start to lose some of their reason for existing. The Galaxy S20+ and S20 Ultra, the highest-profile mmWave phones yet, have experienced legitimately terrible sales in the US since launching. A huge part of that is lockdown closing retail stores, where most Americans buy smartphones, but many in the industry agree that sky-high prices are acting as a major turnoff for suddenly cost-conscious consumers. In retrospect, there truly could not have been a worse time to launch a $1400 smartphone than late March, especially with such insane premiums that, in part, depended on the promise of next-generation mmWave 5G technology. Verizon was so bold as to assume it'd be able to charge a premium access fee for its 5G UWB network, something that now sounds hilarious bordering on actually insane. Avi Greengart doesn't think it's a sustainable position, "Long term, I don’t think that Verizon will be able to charge a premium for 5G." When I asked Verizon if it had plans to drop the access fee or start introducing 5G phones without mmWave connectivity, the company declined to comment, citing a standard policy regarding future products or decisions.

In the end, if the cost:benefit analysis on mmWave 5G dips too deep into the red, I think there's a real chance we see carriers pull back deployments. While no one I spoke to for this article sees any of the big three carriers abandoning millimeter wave wholesale, most agreed that the promise of mobile mmWave was simply not as attractive as a technology in the current environment. Greengart said "I expect that mmWave deployments from AT&T and T-Mobile will also be scaled back or delayed until people can congregate tightly together again," and Kinney's sentiment largely mapped on, saying "I think the hesitancy around aggressive mmWave deployment becomes more pointed with shifts toward [working from home] and a precipitous drop in use of large, public venues." Segan thinks that on AT&T and T-Mobile in particular, the mmWave smartphone use case will get played down, not up, "I don't see AT&T and T-Mobile making a big deal about mmWave on phones. I see them using mmWave as a way to create new businesses with new revenue streams."

While we can't know for certain what's next for 5G and mmWave, there can be little doubt that the uncertainty just became a much bigger factor.