HN Explorer

Starlink is reducing your monthly service fee

130 points by iam-TJ ↗ HN
I've just received a very pleasant surprise in email. Subscription in the UK was £89 per calendar month:

  Effective 24/08/2022, Starlink is reducing your monthly service fee to £75. 

  The price reduction factors in your local market conditions and is meant to reflect parity in purchasing power across our customers.

  No action is needed from you, the price reduction will be automatically reflected on invoices generated after 24/08/2022.

  Thank you for being an early customer and for your continued support of Starlink!

70 comments

[ 4.5 ms ] story [ 129 ms ] thread
ohiovr
I got fiber installed last week! WOOT
comboy
In Poland it went from 449zl down to 230zl ($48). Impressive.
andrew_lvov
In Ukraine it went down to $60. Looks like we're richer than you at this moment :-D
rasz
Probably because majority (out of 12K?) of terminals were financed by State Department (usaid)
badpun
Wow, that's great. As a Pole considering GTFO of the city and into the super rural area with no fiber and probably poor-to-no 4G coverage, that's great news. However, I'm a bit hestitant at the moment betting my life (can't/won't live without Internet) on Elon's side project.
tyingq
Adjustment for the pound falling versus the dollar I guess?
Danieru
That would justify an increase. Instead Starlink is thus charging much less now than when the Pound was stronger.
tjpnz
Cost of living crisis might get some of their customers to look at competing options a few months from now when energy bills really start to bite.
huevosabio
Got a similar email for my service in Mexico. Genuinely surprised and joyful with this company.

Not only does it work with minimal effort almost anywhere but they also reduce prices without going through call center hell.

No way Comcast or Telmex would do this.

__sy__
While on the topic of Starlink, I wanted to share this cool Youtube video [1] I watched tonight on how it works (PCBA design, beam forming, orbital velocity, up/downlink bandwidth as a result...etc). It's a bit long but has some good details!

[1] https://www.youtube.com/watch?v=qs2QcycggWU&ab_channel=Branc...

elaus
Incredibly interesting and very well made. Hit the sweet spot for me in regards of technical details.
thinkingkong
Did any EU regulation or competiton get announced recently? Because it seems limited to that region, ironically.
happyopossum
Commenters here from Brazil and Mexico would be surprised to hear of their recent additions the EU.
gambiting
People in this thread are reporting the same in Mexico and Brazil, so I don't think you're right.
mrtksn
Probably that's not the case, it looks like they introduced localised pricing similar to the one in the App Store(or Netflix, Amazon Prime etc.).

If you think about it, their costs are fixed and they should optimize for revenue and the sweet spot for the num subscribers and price is different at every market.

JumpCrisscross
> their costs are fixed

Their costs are in dollars. Ceteris paribus, this is coming out of Starlink's margin. (They may also be anticipating a compensating bump in volumes. Or they found some cost savings.)

mrtksn
It doesn't matter in what currency their costs are, there is a curve of num_subscribers v.s. payment_per_subscriber and they need to be at the curve's maxima and that maxima would be different for each market depending on the purchasing power of the customers.

I.e. you are better off if you to have 150 customers who pay 75GBP instead of having 100 who pay 89GBP. You already sent the satellites and you already employ the people who upkeep the infrastructure, so it doesn't matter what currency you use, you always want to have the greatest revenue.

JumpCrisscross
> doesn't matter in what currency their costs are, there is a curve of num_subscribers v.s. payment_per_subscriber and they need to be at the curve's maxima and that maxima would be different for each market depending on the purchasing power of the customers

You're describing a monopolist pricing model's revenue curve [1]. Let's complete the model. On the same graph imagine the marginal cost of providing the service (e.g. billing, customer support, peak bandwidth rationing et cetera). Subtract it from the revenue curve. You now have a profit curve. For SpaceX, if we dollar denominate then a rising dollar means the revenue curve moves down; if we local-currency denominate, a rising dollar means the cost curve moves up. Either way, symmetry. (Note: this works even if you assume zero cost elasticity, i.e. a horizontal cost curve.)

Thus, whether SpaceX is buying growth by losing money, hopefully in the short term; buying growth by making less money; or, as you suggest, increasing the quantity demanded on a unit-economically positive game is dependent on the exchange rate.

> already sent the satellites and you already employ the people who upkeep the infrastructure

These systems have marginal costs. Also, last I checked, Starlink has a pre-order backlog.

(My guess: churn is expensive for Starlink. Moreso than most telecoms. Buying goodwill and loyalty amidst rising costs of living is a smart move, despite it likely eating into margin. That or they're about to announce a massive deployment ramp-up.)

[1] https://blog.cambridgecoaching.com/how-to-determine-the-opti...

mrtksn
Sure the strength of the USD is a factor that moves the local price when you bill in dollars but I believe Starlink has a "tech company" financials, by that I mean the cost of servicing the costumer is minimal and most of the money spent is on building the product. Why? Because these are low touch businesses, they don't need to scale their workforce linearly with the demand. It's bit like selling software, you spend money developing it then you spend very little on delivery and support.

AFAIK Starlink's bandwidth bottleneck is on customer per area given, which means they are saturated in some hotsopts like SV or maybe London but not saturated in less dense areas.

Besides, they are building infrastructure all the time and even if there's a backlog today they might have seen some softness in the demand that can lead to losing out on competitors down the road.

It's very unlikely that the price change is altruistic.

JumpCrisscross
> unlikely that the price change is altruistic

Nobody suggested this. My original point was this is coming out of margins. It’s not boosting short-term profits. And it’s not aided by the strengthening dollar.

robga
> AFAIK Starlink's bandwidth bottleneck is on customer per area given, which means they are saturated in some hotsopts like SV or maybe London but not saturated in less dense areas.

Maybe. But just looked. If I order now, Starlink promises to install at my central London location in 1-2 weeks.

s1artibartfast
>You're describing a monopolist pricing model's revenue curve

It also describes any product or company that doesn't sell at fixed price and optimizes revenue.

You set your price to maximize total revenue, whether that is positive or negative.

JumpCrisscross
> It also describes any product or company that doesn't sell at fixed price and optimizes revenue

It's the name of a model. Monopoly is a simplifying assumption as one need not consider the dynamics of price changes in a competitive system.

> set your price to maximize total revenue, whether that is positive or negative

Not always! Scaling costs aren't always linear. And service quality isn't inelastic. The revenue-maximizing price may be one at which you're losing money or delivering a shoddy product that will degrade your customer base in the long run.

iam-TJ
Are all Starlink costs in dollars? Capital maybe, but operating costs?

In each legal domain (country ~ territory) Starlink operates in it has to install and operate one or more ground stations and back-haul into exchanges.

I realise there is a deal with Google for some of that - most likely it varies with the country, data-center locations, and exchange facilities.

But there are presumably ongoing costs to employ and support engineers to watch over the ground station physical assets and ensure back-up systems are always available?

woliveirajr
Brazil: down from R$500 ($100) to R$230($46)

Setup fee+hardware from $1000 to $400.

Price of local fiber: $30 for 600Mb/s

unixhero
The point is not to compete with fiber.
jandrese
If you have local fiber you should use that instead. You won't be sharing spectrum with everybody else in the area. Starlink is for people who are underserved by terrestrial broadband.
yuppiepuppie
Uh what? Starlink is for anyone who wants to buy it. I don’t think Elon would say that you shouldn’t use it, under any circumstance.
mrjin
Sure, you can buy it. But should you buy it? Especially when there are far better and much cheaper alternatives.
yuppiepuppie
Thats up to each individual to decide. But say you shouldnt buy because there are others that would need it more seems to go against the basic principles of supply/demand
snapplebobapple
He is just saying if you have local fiber and its priced similarly or cheaper it is illogical to get starlink, and he is right in almost all instances. This isnt a formal discussion,you kinda need to cut people slack and give benefit of the doubt where there is ambiguity or ask for clarification before going attack dog.
stazz1
assuming one is stationary
aaasss333
Why would you pay more for a slower connection though (both bandwidth and ping)? And probably also less reliable.
badpun
Starlink seems to have serious limitations. Apart from potential congestion in your area (leading to limited bandwidth), as the satelites are not geostationary (and hence move) you need to have unobstructed view of a big portion of the sky. If you have trees relatively near you, it can become a problem and connection will be dropped every x of minutes.

Having said that, it's still an amazing technology. I hope they will be able to iron out these issues.

ghaff
>I hope they will be able to iron out these issues.

Well, some, like the number of satellites are mostly an economic question. How much satellite is a given user willing to pay for?

Others, like objects blocking the antenna's view of the sky are more inherent.

Starlink is mostly not likely to be a great option for most people if they have decent terrestrial broadband options.

KennyBlanken
The comments here showering Starlink with praise, acting like they're doing this to be a Bro, are bizarre.

Starlink didn't drop prices out of the goodness of their hearts.

They dropped prices because they had to. Why is debatable, but signs point to the network crumbling apart, causing them to lose customers.

It hasn't been able to scale to handle the customer load, even in rural areas: https://www.pcmag.com/news/starlinks-massive-growth-results-...

"Hundreds of miles south near Leonard, Texas, John Lawyer has encountered download rates on his Starlink dish that can dive as low as 1Mbps, especially during the evenings."

That's a town of 2,000, in a county of 35,000.

New customers are probably also put off by the signup costs and problems with "Dishy" - thermal shutdowns, a permanently attached cable for power/data, and being fairly fragile.

wahnfrieden
Unscalable hype?
ohiovr
Huges was probably great an eon ago.
mynameisvlad
Hughes was never and will never be able to be low latency with their current setup. You’re fighting against physics at that point due to the location of the satellites compared to Starlink. They also have always had exceptionally restrictive data caps.

So I wouldn’t say they were ever “great”. Just realistic.

ohiovr
Good to know. Geostationary latency is a bitch.
ghaff
And data caps. Where my brother has a house in rural coastal Maine, the only two options further down his road (~1 Mbps DSL went as far as his house) were essentially Hughes or Verizon hotspot. Everyone who looked into it basically concluded that they were better than no Internet but you really couldn't do things like video streaming to any significant degree.

Starlink is a game-changer for somewhere like this that had really marginal Internet and basically someone couldn't have worked from there. (People did go into the local small city and do Internet from the cafe or library but that obviously doesn't work for regular full-time work.)

jandrese
If there are too many customers it doesn't make sense to lower prices to convince them to leave.
onion2k
If everything is working properly and the network is over capacity then it would make sense to keep prices high and accept the loss of customers.

If the network isn't working properly, and it should be able to cope with the number of users, then lowering prices to retain customers until you can fix the technical issues makes sense.

herendin
>The comments here showering Starlink with praise, acting like they're doing this to be a Bro, are bizarre.

Can you link to the comments that said that?

I can just see one comment saying Starlink is better than its competitors, which might fit that description. All the other comments seem like neutral personal reports of the price reduction in various places, or comments questioning the reasons for this, or even complaining about it

beeboop
Don't concern yourself with the musk derangement syndrome crowd, they basically only exist to make shit up
scarface74
If the network is crumbling because of congestion, wouldn’t the logical thing to do be to raise prices?
fomine3
IMO they shouldn't cut price much because their available bandwidth is strictly limited by satellites that is very costly to launch (though SpaceX is really great to cut costs, it's still satellites) and not unlimitedly expandable, unlike fibers. Keep price high (but affordable) so who really needed can use service at higher speed, and make profit. Perhaps different pricing per area is good way.
bottlepalm
It’s well known than cells in certain populated areas are at max capacity with degraded performance. The network isn’t ‘crumbling’ by any stretch. They are launching new satellites almost weekly to increase capacity. Starlink is still early days with version 2 around the corner.
elaus
Why would they reduce the sign-up costs if they couldn't even handle the current customers?

I understand wanting to keep current customers by reducing the price when they are unhappy with the service. But incentivizing even more people to use the service seems to contradict your thesis.

noncoml
I just hope they don’t go bankrupt. With 250,000 customers they are making at best $25m per year. Can see how that’s enough to sustain the whole thing.
lebski88
I think you forgot to * 12 for the months.

250,000 customers is approximately $300m a year.

250,000 * 100 dollars a month (approx - maybe less now) * 12

spech
Sames goes for Germany with 80€ and the German Telekom still isn't able to provide me a decent connection although fiber is installed in almost all adjacent streets...
gitgud
I have never seen a company inform all users that they will automatically pay less for the same service...
timbit42
Canadian ISP TekSavvy has done this numerous times. They don't offer satellite internet though.
thenickdude
AWS does this periodically for storage and bandwidth pricing
iam-TJ
I should have pointed out in the UK the cost includes 20% Value Added Tax (VAT) so the actual charge has reduced from £74.10 to £62.50 (US$87.44 to US$73.75 @ £1=US$1.18 on 2022-08-25).
anelson
I got this too for my Starlink unit in Ukraine. Dropped to $60 from $100. A welcome change although I for some reason assumed it was specifically for Ukraine. Looks like it’s in a number of markets.
jroesner
I can imagine there’s a direct connection to Apples announcement on Sep 7th. I don’t recall all the details, but read a report, that Apple bought a partially defunct satellite network, that has only very few, but therefore geostationary satellites which come with bandwidth licenses that exceed Starlink’s by a lot. Might be totally wrong, but did anybody else hear that rumor?
Kye
The latency is probably as bad as any other stationary satellite network. Apple might have bought it to ensure connectivity with its interests across the globe, not to productize it.
BenjiWiebe
The latency will be bad unless they use faster than light communication.
Kye
This is in a context comparing them to satellites that fly low enough to get latencies comparable to terrestrial.