There is an interesting trend of ISPs selling broadband products that are not always guaranteed to be at the same speed and quality as other customers.
Throttling customer speeds is not new to the industry. Some of the companies that with long-time data caps throttle data to slow speeds after a customer reaches the monthly allowance of usage. Most such ISPs offer an alternative for customers to buy extra broadband to maintain their normal speeds. Some of the companies that have had this practice include the high-orbit satellite providers, cell carriers providing hotspot plans, and a handful of others. Many companies with data plans don’t throttle speeds and just automatically bill more for going over the data cap.
I’ve noted this practice again in recent years from the big FWA cellular providers that sell home broadband using cellular spectrum. AT&T, T-Mobile, and Verizon have all reserved the right to throttle customers any time that the network gets too busy. For example, from the terms from T-Mobile, “During network congestion, some T-Mobile internet customers might notice slower speeds, including Home Internet customers” Home Internet is the FWA home broadband product.
I’ve been able to observe examples of them doing this. I’ve seen speed tests from customers using FWA that have speeds over 200 Mbps during the year who occasionally get throttled down to just a few Mbps. I think these customers are surprised every time this happens and probably don’t understand or remember that the throttling is a part of the terms they agreed to.
It’s easy to understand why cellular companies would throttle home broadband customer first – they are protecting their cellular customers. I’m sure all of the FWA providers are happy with the new revenues coming from FWA, but T-Mobile is not going to let the home broadband for 6.4 million FWA customers threaten the experience of 130 cellular customers.
Starlink also throttles certain customers. One of the features of Starlink’s Away plan for campers and hikers is that Starlink reserves the right to throttle data usage if the network gets too busy. It’s also easy to understand this. As the RV products becomes more successful, it’s not hard to imagine a lot of campers coming together at the same location wanting to connect to Starlink. That traffic alone could overload a particular satellite, but Starlink is also shielding its customers who live in the same region and who are paying full price.
Starlink also reserves the right to throttle customers who buy its new ‘Residential Lite’ product for $80 per month. Rather than mention throttling, Starlink calls it deprioritization, “This service plan will be deprioritized compared to Residential service during peak hours. This means speeds may be slower for Residential Lite service relative to Residential service when our network has the most users online”. This term is at the top of the company’s advertising for the product, and they want customers who want the lower rate to recognize what comes with the plan. The company is making it clear that there are trade-offs for getting the lower price.
I’ve been thinking about all of these plans and net neutrality. One of the key features of the national net neutrality plan was that ISPs couldn’t engage in paid prioritization, meaning that a customer could not be charged more to be guaranteed a better connection.
It’s not clear to me that this practice violates that principle. In the case of the FWA products, every customer buying the FWA product runs the risk of having data throttled – there is no other class of customers with higher priority unless it’s cellphone customers. Starlink is a little different in that customers can save money by agreeing to possibly be throttled. Is having customers agree save money by being deprioritized the same as charging somebody else more to get a better priority?
It certainly doesn’t matter at the federal level since the Courts recently killed the appeal to the FCC’s net neutrality case – and the FCC would have killed net neutrality anyway if the Courts didn’t do it. It is a more germane question in California which adopted a state net neutrality plan that largely mimics the federal rules.
Excellent observations. In any industry where capacity has a scarcity component the business will make efforts to manage that capacity for the best overall outcome. Capacity constraints exist in GPON fiber networks, often a 1:32 ratio, relying on the fact that not all users will be actively using their full bandwidth capacity simultaneously. Wireless companies manage Spectrum availability.
Duke Energy will provide a discounted rate for electricity if you agree to be throttled during peak hours. No one is guaranteed a lane on the highway at the speed limit – sometimes congestion occurs and you drive at a crawl, not at the speed limit. Every industry has to manage their capacity – this should be no surprise to anyone.
The only exception to unlimited capacity I’ve heard of is in the bible, 1 Kings 17:14 – the Bible tells the story of a widow whose oil and flour never ran out, a miracle attributed to God’s provision and a demonstration of His power, not just providing enough but continuing to do so until a specific event, the rain.
Deprioritization is, to me, different than throttling.
My definitions….
Throttling is putting a hard limit on a customer’s speed so they are at the same slow throughput whether the cell site is busy or almost totally idle. Deprioritization only matters when the site is busy–the deprioritized customer gets fewer chances per unit of time to send data compared to “regular” customers. So if the site is idle even a deprioritized customer can attain high speeds.
I got my company to deprioritize instead of throttle. Why limit a customer when there is no reason to do so?
…because products delivered to consumers need to have set expectations that can be met. Consumers in particular judge services based on the re-occuring slow periods so if you’ve made this a typical thing, that’s what sits in their minds.
throttling actually sets expectations, especially if the oversubscription ratio on a tower is appriopriate, and then can be reliably delivered. ‘wide open’ makes it MUCH more difficult.
deprioritization allows priority users to muscle out low-priority users and makes their service really poor. The lack of any throttling on priority users really sucks.
You can see this very very clearly today with deprioritizing cell plans that are basically useless during busy times. deprioritizing sucks for the consumer, and it sucks more and more the lower the income level when people are buying bargain basement mvno services that are bottom priority.
Cell companies are in a manupulation phase where they’ve all aggreed effectively to onboard customers and overwhelm systems so they can show the government how badly they need new spectrum and new rules. Even on the most premium plans you can feel the deprioritization at play.
The vast majority of consumers would be far far far better off getting a split ratio of capacity during busy times ie a throttle, A modern LTE eNB might have about 1Gbps capacity and if there are 200 active users. Giving each 5Mbps is so much better than the deprioritized Kbps that can’t load a page. Even putting in higher class services you might have some people getting 10Mbps of alotment while value plans get 2Mbps, that’s STILL far better that deprioritization.
These types of services would be tramatically more reliable and consistent if they were sold by Mbps instead of wide open with a deprioritization scheme that makes for effectively zero connectivity for many users.
What was done in the case I was talking about was more a “ratio of speeds” approach. When a cell became congested, a deprioritized user would get about 1/2 to 2/3 of the rate of what regular users were experiencing. So, no one got close to zero throughput unless the cell was so grossly overloaded by some unusual event that the situation was “more or less hopeless” and everyone was having problems..
NN paid priority was mainly concerned with CASPs (content & application service providers) paying for priority, e.g., Netflix getting higher priority than Hulu. I don’t recall that content-neutral inter-customer prioritization raised concerns of neutrality. It doesn’t seem fundamentally different from speed tiers, just more dynamic. It may, as you note, raise consumer protection concerns if the customer is confused about what this means. It also makes broadband labels less informative – there’s no really good way to see the difference between “you’ll get slightly lower speeds on your speed test once a month or so” to “forget about watching Netflix on Sunday nights.”