Blog

More Starlink Speed Tests

Christian Leave a comment

Last month, speed tests from beta testers of SpaceX’s Starlink leaked. This month, a few tests conducted by StarLink itself were shared in a public FCC filing. The results are outstanding:

Test 1

  • Ping: 19ms
  • Download speed: 103Mbps
  • Upload speed: 42Mbps

Test 2

  • Ping: 18ms
  • Download speed: 103Mbps
  • Upload speed: 41Mbps

Methodology

SpaceX almost certainly cherry-picked the tests to put Starlink in a good light, but the results are still impressive. I wasn’t convinced Starlink would ever deliver on Elon Musk’s claims about sub-20ms latency. Hell, Ajit Pai, the FCC’s chairman, was skeptical Starlink would deliver sub-120ms latency.

The test results come from screenshots in a presentation slide. I don’t know much about the methodology behind the tests, but it looks like they came from Ookla’s speedtest.net.

Along with the test screenshots, the slide includes some text. Here are a few of the bullets:

  • High-speed, low latency broadband to any location on earth
    • Tested at over 100 Mbps using standard user equipment
    • Latency <40-50ms round trip to the internet

I’m not sure if SpaceX was meaning to imply that the screenshots shared in the slide came from tests using standard equipment. While the screenshots do show speeds over 100Mbps, the latency results are lower than 40Mbps (possibly a lot lower if the tests are measuring round-trip as I expect).1

Here’s the graphical portion of the slide:

Graphical portion of slide showing speed test results

At first glance, I thought two separate tests gave nearly identical results. Zooming in, we can see both tests have the same ID.

Screenshot showing two tests with identical ID numbers

I expect it was an honest mistake on SpaceX’s end, but it’s strange.

Data Outage Affecting Some Mint Mobile Subscribers

Christian 3 Comments

A data outage has been affecting some Mint Mobile subscribers throughout the day. One of Mint’s co-founders, Rizwan Kassim, posted about the issue on Reddit:

An upstream error seems to have caused data provisioning errors for a number of subscribers.

It’s being worked, they don’t have a root cause yet, but I know this has been escalated. Down Detector showing issues on our carrier as well; not sure if it’s related or not.

I strongly think, but do not know, that this has nothing to do with the iOS 14 upgrade many of you installed today.

Based on reports I’ve read from Mint subscribers, the issue appears widespread geographically. I don’t know what proportion of Mint’s subscriber base is affected.

Kassim’s Reddit post suggests the issue may also be affecting T-Mobile. While Downdetector shows a slightly unusual level of issues associated with T-Mobile, I don’t think Downdetector’s data is consistent with a large-scale problem for T-Mobile subscribers.

I did a bit of my own digging for T-Mobile subscribers’ complaints about the networks’ performance today. I didn’t run into anything out of the ordinary.

Verizon Expands Connected Device Plans

Christian Leave a comment

Today, Verizon announced changes to its connected device plans. Verizon is now offering two add-on plans for mobile hotspots and tablets. Most subscribers on Verizon’s unlimited plans are eligible for the add-ons.

  • Unlimited – $20 per month
  • Unlimited Plus – $30 per month

Subscribers on the Do More Unlimited and Get More Unlimited plan can get a 50% discount on either plan.

Here’s a graphic from Verizon’s website:

While the graphic suggests there are two plans, I think its easier to make sense of the new offerings as four different plans:

  1. A $20 hotspot plan
  2. A $20 tablet plan
  3. A $30 hotspot plan
  4. A $30 tablet plan

Hotspot plans

The hotspot plans really shouldn’t be called “unlimited.” On the $20 plan, subscribers only have 15GB of regular-speed data. On the $30 plan, subscribers have 30GB of regular-speed data.

Verizon throttles customers that use all of their regular-speed data. I expect Verizon is sticking with its old policy of throttling to 600Kbps. At that speed, mobile hotspots lose most of their usefulness. I would find working on a laptop with a 600Kbps connection extremely frustrating.

Tablet plans

The tablet plans have limited allotments of “premium data” (15GB on the $20 plan and 30GB on the $30 plan). While subscribers have premium data available, they’ll have high-priority connections during periods of network congestion. After premium data allotments run out, subscribers may experience especially slow speeds if Verizon’s network becomes congested. Fortunately, congestion is rare in most areas.

5G

The Unlimited Plus plan includes unlimited 5G Ultra Wideband data on both tablets and hotspots.1 While that sounds great, it’s rarely meaningful in practice since Verizon’s 5G coverage is extremely limited. I’m guessing the unlimited 5G data will be dropped from the plan when Verizon’s 5G coverage expands and the 5G network becomes more congested.

Artificial Hotspot Limits

Christian Leave a comment

It’s common for cell phone plans to include limits on mobile hotspot data that are separate from limits on overall data use limit. E.g.,

  • Verizon’s Get More Unlimited offers unlimited regular data but caps mobile hotspot use at 30GB.
  • One of Mint Mobile’s plans comes with 35GB of regular data but caps mobile hotspot use at 5GB.

Recently, a Reddit user was confused about Mint’s policy and asked:

What’s the reason for the 5gb cap on the hotspot? I have a friend who this plan would be perfect for, however he tethers his iPad frequently to watch YouTube. Not sure what the big deal is since you could just switch the sim anyway.

Here’s how I responded:

My speculation:

Even though Mint allows 35GB of use, it knows the vast majority of subscribers won’t use that much data. If all subscribers used their full allotments, the plan would be much less profitable for Mint.

By restricting hotspot use, Mint reduces data use and (more importantly) dissuades some very heavy data users from ordering the plan in the first place.

I may not have that that exactly right. Mint’s arrangements with its host operator, T-Mobile, are not public knowledge. But the underlying logic is right. A gigabyte of mobile hotspot data isn’t more cost-intensive for a carrier than a gigabyte of on-device data.

Verizon Plans To Acquire Tracfone

Christian Leave a comment

This morning, Verizon announced plans to acquire Tracfone. The planned deal will involve an acquisition of the Tracfone brand and a bunch of subsidiary brands like Total Wireless, Straight Talk, and SafeLink.

At the moment, these brands have about 21 million subscribers. The deal is slated to be worth six or seven billion dollars (or about $300 per subscriber):1

The consideration for the transaction will include $3.125 billion in cash and $3.125 billion in Verizon common stock, subject to customary adjustments, at closing. The agreement also includes up to an additional $650 million in future cash consideration related to the achievement of certain performance measures and other commercial arrangements.

Along with the subscribers and brand names, Verizon is acquiring Tracfone’s roughly 850 employees and Tracfone’s retail presence in over 90,000 locations.2 Verizon expects the deal to close in the second half of 2021.

Reflections & open questions

Tracfone and Verizon will need to pass through some regulatory hoops before the deal is official. If the acquisition goes through, it will cause a massive shift in the industry. Tracfone’s user base makes up about 5% of the U.S. wireless market and a major share of the prepaid market.3

At this time, I’m guessing Verizon will continue to operate several Tracfone brands rather than consolidate Tracfone subscribers under the Verizon brand name.4 Years ago, a Verizon executive discussing Verizon’s lackluster number of prepaid subscribers stated the following:5

“Our retail prepaid is above market. We’re really not competitive in that environment for a whole host of reasons and it’s because we have to make sure that we don’t migrate our high-quality postpaid base over to a prepaid product…Quite honestly, we use the Tracfone brand as our prepaid product.

About 13 million of Tracfone’s subscribers already have service running over Verizon’s network.6 I don’t know what will happen to the 8 million subscribers on other networks. I’m guessing Verizon will try to transition most of them to the Verizon network, but Verizon may sell the subscribers to other carriers.

When the merger between Sprint and T-Mobile closed, I wrote:

I continue to think the merger is going to be bad for consumers over the long term.
I’m guessing the merger between Sprint and T-Mobile contributed to the viability of Verizon’s Tracfone acquisition. As with the merger, I’m not optimistic about the effects this new acquisition will have on consumers in the long term.

Unlimited Plans: A Race To The Bottom?

Christian Leave a comment

In the last few years, many cell phone carriers have released “unlimited” plans that actually have limits. Most of these plans are sufficient for the average person. Problems show up for a minority of cell phone users that are especially heavy data users.

Many people read my posts explaining the limits carriers place on their “unlimited” plans and react with a version of:

Ok, fine Chris. Sure these plans aren’t technically unlimited. But you’re being pedantic as hell. These plans are as-good-as-unlimited for 98% of people.

My pushback on “unlimited” plans isn’t about protecting heavy data users. In the long run, I’m worried that “unlimited” plans are part of a trend that will be harmful to a much larger group of people: light and moderate data users.

Huh?

This has happened before

For more than a decade, I’ve been following the portion of the web hosting industry that caters to personal websites and small-business websites. When I first started watching the industry, almost every web host offered a fixed number of gigabytes of bandwidth each month. Customers that wanted more bandwidth had to pay more. At some point, a few web hosts began offering “unlimited” bandwidth plans.

Of course, no web hosts actually offered unlimited bandwidth. Hosts put restrictions in their terms of service agreements that made it possible to shut down websites that hogged server resources. If Google had tried to host its infrastructure on a $10 per month “unlimited” plan, it would have been shut down instantly.1

Even though most websites are tiny and have modest resource demands, people running tiny websites tend to like the idea of having an unlimited plan. Since the internet has way more tiny websites than medium-sized websites, web hosts could allow some unprofitable, medium-sized clients to stick around. The hosting bills for tiny websites essentially subsidized some more popular websites.

Over a few years, it became clear that offering “unlimited” plans was a winning business strategy. Gradually, unlimited plans became the industry standard. Fixed-bandwidth plans faded away.2

Back to cellular

Fixed-data cell phone plans are fading in the U.S. market. Take a look at the websites of any of the Big 3 networks. Which plans do you see? Unlimited plans get the attention. Fixed-data plans still exist, but they’re buried.

It didn’t used to be this way. The move towards unlimited plans has been rapid and will probably continue until unlimited plans dominate the market. Unless regulatory bodies step in, I see only two ways this can play out in the long run. Both scenarios seem bad:

  • Unlimited plans without many restrictions become standard. Light data users essentially subsidize heavy data users.
  • “Unlimited” plans with significant restrictions become standard. We get a race to the bottom.

Racing to the bottom

When fixed-data plans dominated the market, customers were aware of the limitations they were likely to run into. Hell, plans were their limitations. A 5GB plan might have been named “The 5GB Plan”.

As unlimited plans have risen, limitations have been hidden from customers and tucked away in the fine print of legal documents. Plan names turned meaningless: “T-Mobile Magenta” and “Verizon Above Unlimited.”

Carriers place limits on their “unlimited” plans so they can compete on costs. Have you noticed the policies below on the rise?

  • Video throttling
  • Hotspot data throttled to slower speeds than regular data
  • Monthly hotspot allotments that have no relation to overall data allotments
  • Data transfer that’s restricted to sluggish speeds after subscribers use a certain amount of data

For network operators, it’s not important whether a gigabyte of data is used streaming video, loading web pages, or running a hotspot.3 All these policies have the same purpose: reducing subscribers’ data use.

Low-priority data is another common limitation thrown on plans. Subscribers with low-priority data will experience normal speeds when a network isn’t congested, but their speeds will turn sluggish when things get busy.

Limits aren’t the problem

I’m not broadly against limits. I’m against limits that confuse consumers. I’m against limits that aren’t explained clearly and prominently.

Unfortunately, unlimited plans attract the kinds of limits I oppose. At some level, it makes sense, at least from a business perspective. If a carrier downplays how serious the limits are on one of its plans, the plan will be more appealing to consumers.

Carriers throttling heavy data users to 128Kbps don’t make candid disclosures. Imagine what that would look like:

After 35GB per month of data use, download speeds will decrease to frustratingly slow speeds (around 128Kbps). You probably won’t want to use the internet at these speeds unless you really need to. But if you have to load a boarding pass or an email after you’re out of regular data, you should be able to with a bit of patience!

No. We get vague disclosures like:

Data speeds reduce after 35GB but data is unlimited.4

Plans with low-priority data will have fine print mentioning reduced speeds during congestion, but details will be sparse. Customers trying to figure out how common congestion is, where congestion tends to occur, or how much speeds are slowed aren’t going to find the information they’re looking for.

Hell, it’s not just regular consumers that get confused and misled. My favorite tech review site can’t sort out prioritization policies. Here’s a bit from Wirecutter:

A T-Mobile spokesperson confirmed that policy, saying that although postpaid and prepaid T-Mobile service have the same priority, Metro by T-Mobile and other resellers ‘may notice slower speeds in times of network congestion’…However, AT&T and Verizon told us that they don’t impose any such prioritization.

Perhaps the scariest part of the excerpt is not that Wirecutter is wrong, but that people speaking for AT&T and Verizon were wrong about their own companies’ policies.

Where will we end up?

If nothing changes, we’ll continue to see the low-cost side of the market (a) throw more limitations on plans and (b) bury limitations deeper. In my view, the problem isn’t evil carriers. It’s bad incentives. Maybe the FCC or the NAD (National Advertising Division) will jump in and change carriers’ incentives. I’m not too optimistic, though.

RootMetrics’ Early 2020 Report

Christian Leave a comment

Earlier this week, RootMetrics released its report on cellular networks’ performance in the U.S. in the first half of 2020. The most recent round of RootMetrics’ testing got screwed up by the pandemic, but the company did its best to publish something that followed the structure of its usual reports.

While I continue to think RootMetrics has the best methodology of any network evaluator, I’m not going to discuss the latest results in detail. For the most part, the results were as expected and similar to what RootMetrics found in its previous round of testing.

Highlights

As usual, Verizon was the big winner:

Verizon continued its run of excellence in our national testing, winning or sharing six out of seven awards.

AT&T showed solid performance and easily took the second spot for overall performance. AT&T managed to beat Verizon on RootMetrics’ speed score.

While Sprint is gradually disappearing, RootMetrics included the in its testing. Oddly enough, Sprint’s overall score beat T-Mobile’s score.

Mint Mobile’s Unlimited Plan Has Limits

Christian 24 Comments

Today, Mint Mobile launched an “unlimited” plan. Mint has officially joined the ranks of carriers like Google Fi, Altice, Total Wireless, Wing, Tello. What do all these carriers have in common? Each offers an allegedly “unlimited” plan that strictly limits how much data subscribers can use.

The rest of this post is a rant. To be clear, I think Mint’s new plan is great. I just hate seeing the cellular industry move towards a scenario where every carrier has to offer plans that are misleadingly labeled “unlimited” in order to remain competitive. If you’re looking for a level-headed overview of Mint’s new plan, see my previous post. If you’re looking for cynicism and entertainment, keep reading.

Limits

Mint’s unlimited plan has three major restrictions:

  • Subscribers can only use 35GB of full-speed data each month. After 35GB of data use, Mint throttles data to sluggish speeds.
  • Mint limits mobile hotspot use to 5GB per month.
  • Mint throttles video streaming to a maximum of 480p.

How slow are speeds after 35GB?

Mint screwed up its communications about the throttling it imposes after 35GB of data use. Pre-launch information I received said subscribers would be throttled to 128Kbps. The only specific speed I’ve found mentioned on Mint’s updated website is 64Kbps:

Mint Mobile’s ‘Unlimited Data’ plan comes with 35GB of high-speed data, which is slowed to 64 kbps thereafter and reset at the next billing cycle.

I expect Mint will clarify its policies by the end of the day. For the rest of this post, I’ll give Mint the benefit of the doubt and say the throttle is 128Kbps. In some sense, it doesn’t matter if the throttle is 128Kbps or 64Kbps. The internet will be almost unusable at either speed.

(Update: Mint clarified that users exceeding 35GB of data use will be throttled to 128Kbps)

Objections

But Chris! Mint lets you use unlimited data at 128Kbps! Sure, 128Kbps is slow as hell, but the plan is still unlimited!

No. At 128Kbps, a lot of things won’t work. Video won’t stream. Some web pages won’t load at all.

More importantly, a rate limit can’t coexist with unlimited data. If a full 128 kilobits is transferred every second for an entire month, only 41GB of data is used.1 There’s an absolute cap on Mint’s unlimited plan on data use of about 76GB (35GB + 41GB). Realistically, almost no subscribers will get much past 35GB of use in a month, since the internet will be so frustrating to use after the 35GB of full-speed data runs out.

But Chris! 35GB is practically unlimited! Almost everyone uses way less data than that!
Agreed! If you’re excited about the plan, this post probably shouldn’t dissuade you.

I don’t even fault Mint for calling the new plan “unlimited.” I’m impressed Mint managed to hold out so long while its competitors offered unlimited-but-not-really-unlimited plans.

My point is that consumers would be better off in the long run if carriers weren’t incentivized to mislabel plans.

Anyhow, if you’re interested in Mint’s new plan, go for it. It’s an awesome deal for $30 per month. Just realize it’s a 35GB plan.

Mint Mobile Launches An Unlimited Plan

Christian Leave a comment

Mint Mobile launched an unlimited plan this morning. It’s available for as little as $30 per month.

Plan terms

Like many unlimited plans offered by mobile virtual network operators (MVNOs), the plan isn’t actually “unlimited” in the mainstream sense of the word:

  • Subscribers can use 35GB of regular, full-speed data each month. After 35GB of data use, Mint throttles speeds to a sluggish 128Kbps.1
  • Mint caps mobile hotspot use at 5GB per month.

I’ll save my complaints about Mint misusing the word “unlimited” for a second post. 35GB of data and 5GB of mobile hotspot access will be sufficient for the vast majority of people.2

Like Mint’s old plans, the new plan includes unlimited minutes and texts. Calling to Canada and Mexico is also included at no charge.

Subscribers with 5G-capable devices will get access to 5G coverage from T-Mobile’s network. While T-Mobile’s 5G network is lackluster in terms of speeds, it leads the nation in 5G availability.

Pricing

With the new unlimited plan, Mint is continuing to price service based on how many months of service customers pay for upfront.

  • $30 per month – 12 months of service
  • $35 per month – 6 months of service
  • $40 per month – 3 months of service

The unlimited plan is eligible for the same introductory offer that Mint offers on its other plans. New customers can purchase three months of service at the rate Mint usually reserves for customers that purchase 12 months of service. I.e., three months of service on the unlimited plan costs $90 ($30 per month).

Reflections

Competitiveness

I’m glad to see Mint offering a plan for heavy data users with such a low price point. I expect the plan will be popular, especially among people who only need one or two lines of service. The new Mint plan should be competitive with other low-cost unlimited plans offered by carriers like Visible and Cricket. While I don’t think Mint’s new plan will make T-Mobile’s Essentials plan irrelevant, I’m ready to argue Mint’s plan is almost strictly the better option.

Pricing strategy

Interestingly, Mint has narrowed the distance between pricing tiers with the new plan. Mint’s 8GB plan costs $20 per month for customers that purchase a year of service upfront. The plan is 75% more expensive ($35 per month) for customers that purchase 3 months of service.3 Mint’s unlimited plan is only 33% more expensive for customers that opt for 3 months of service.4

In the past, I’ve wondered whether Mint’s pricing structure made volume discounts too aggressive. The large difference between monthly rates on three-month terms and twelve-month terms may have made the carrier unappealing to budget-sensitive consumers that could have been a good fit for Mint. Is it possible we’ll soon see Mint narrow the gap between pricing tiers on its old plans?

Mint’s new approach to pricing has a funny consequence. In some situations, Mint’s 12GB plan is now $5 per month more expensive than the 35GB (unlimited) plan.5

PCMag Releases 2020 Cellular Performance Report

Christian Leave a comment

PCMag just released its 2020 report on the performance of cellular networks.

  • Verizon took the top spot for overall performance.
  • AT&T came in a close second.
  • T-Mobile came in third place but led in 5G availability.

Differences from previous years’ tests

Due to logistical issues from the pandemic, PCMag altered its methodology:

Traditionally, we’d tour each city and then test rural areas between cities before moving on to the next one. But that involves flights, rental cars, and hotels, none of which we felt safe using this year. So we hired roughly two dozen drivers to each test their own cities, in their own cars, sleeping in their own beds, shipping the testing kits from place to place. The result is a nationwide, COVID-safe test, but without the rural data we usually provide.

PCMag also started placing more emphasis on 5G connections. I’m a big fan of how PCMag handled 5G performance in its scoring (emphasis mine):

We had separate sets of 4G and 5G phones running tests offset by 60 seconds from each other…We ended up choosing the best result from each of the two devices on the same network, no matter what G they were on…What people really want is a consistent broadband experience—they don’t care what the icon on their phone says.

Reservations

Most of my reservations last year still stand. Notably:

  • PCMag focuses on performance within cities, while the largest differences between networks’ performance tend to show up in less-populated areas.
  • Average speed metrics get more weight than I think is reasonable.
  • Scores on different metrics get aggregated in a problematic way.

Highlights

While my reservations are serious, they’re not relevant to the granular, city-specific results. If you live in a large metro area, PCMag’s scorecard for your city could be handy.

The 5G-availability data is interesting. Here’s each network’s overall score for 5G availability:

  • T-Mobile: 54%
  • AT&T: 38%
  • Verizon: 4%

I’m surprised how close AT&T came to T-Mobile. While Verizon’s 4% availability score isn’t impressive, it’s higher than I anticipated. Verizon has been getting berated for the horrible availability of its exclusively millimeter wave 5G. Since Verizon hasn’t rolled out any 5G in some cities, the overall result masks heterogeneity between cities. E.g., PCMag found 9% 5G availability for Verizon in Chicago.