What does this even mean? Google provides the leased line to the store? The store obtains an Internet connection the normal way, but then tunnels to Google? Or is Google's level of involvement that they have paid for an ad in the SSID?
Google Fiber is a nice business opportunity for them, but that requires infrastructure. This comes with none of that. It's not even a precursor to Google Fiber rollout. Either they are getting a similar-quality leased line to _one_ place, or it's just an ad. And even if the Starbucks uplink is "Google Fiber", the wireless connection itself will be a huge bottleneck, as it always is when you have more than around five devices communicating at once.
> Or is Google's level of involvement that they have paid for an ad in the SSID?
They're paying for the increased speeds in exchange for the ad, yes. It seems pretty clear to me that Google wants to brand itself with the word 'fast'. This goes as far back as the original Chrome commercials. I think it's rather smart.
I've actually never thought of AT&T as 'mediocre' because of a Starbucks relationship. I view them as 'mediocre' due to a decade of terrible, unreliable, and spotty cellular service. It's amazing what service we'll put up with just because all of our friends and family are using that same provider.
What I'm saying is that I think they are using Level 3 for the whole stack, not just circuits. I think L3 is doing the captive portal and everything else. I could be wrong though.
Do you really think Starbucks would give up its captive portal? Have you seen it? It's full of Starbucks branding and content. I don't see that aspect disappearing with the new provider.
Sounds straightforward to me -- Google is subsidizing the installation of faster, free wifi at all Starbucks. If you are asking what Google gets out of it, probably good PR, their name in the network SSID, and of course, the more people use the internet (on average) the more money google makes.
> the more people use the internet (on average) the more money google makes.
People tend to forget about these indirect effects - changes in habits, behaviours, economies. Anything leading to "more internet" or "more computer" in our lives will, almost certainly, lead to more revenues for Google.
Also, something I didn't think about, but if this push forces other companies to follow suit, and makes free wifi more standard, then this indirect effect could be pretty big.
"you’ll know your new network is ready to go when you can log in to the “Google Starbucks” SSID."
LOL either the real one will be up or someone running a MITM attack.
It is an interesting area that has not been tackled by a startup. I have no idea how they'd tackle it. But it would be nice to know when I connect to some random network if I'm on the real thing or just some spoofed SSID. Yeah yeah connect to your VPN... but another layer of security, or a reduced level of threat, never hurt anything. Maybe a free VPN would be the point of the startup, "Don't trust your internet access? Connect to us!"
Also this would have been a cool place/idea to roll out guaranteed working IPv6 connectivity. This may very well be the case, although unpromoted.
This is more of a standards problem then a start up problem. Minimally, you need to get something standardized using public key Identification.
Another option is to deploy something like secure in band wireless pairing[1] which prevents anyone trying to spoof an access point while you are in range of the existing access point. Essentially, the access point uses transmit/not transmit as an encoding mechanism for secure announcements. Anyone attempting to jam that can't because forging an announcement requires you to not transmit at key points
[1] http://people.csail.mit.edu/nabeel/TEP.pdf
That's no different from how it is today, except the SSID will be slightly different. You have no way to tell whether you're logging into the official Starbucks WiFi or some black hat's stand-in.
There are quite a few VPN products out there advertised specifically to protect you while connecting to untrusted public WiFi networks.
In order for that to work, I think you necessarily need the cooperation of the network owner, and an informed end user. Given that, it seems like the technical solution is basic public key cryptography. Give the access point a public key (possible the same key it uses for https), then all the user has to do is validate that the access point is using the appropriate certificate.
You could provide software that automatically validates networks when you connect, and/or provides a more user friendly way to do so.
The biggest problem I see is that their is no way to automatically distinguish between a MITM, and the vendor simply not participating.
"informed end user." Or it could be a free automatic feature of Android 4.4. The almighty GOOG has the unusual situational advantage of being the same guy doing the WIFI and doing the phone/tablet OS. The only place AAPL has them beat is they sell the wifi access point devices too.
They could release an ITMS app for iDevices, I suppose.
My relatively new android phone is running 2.3.7, so somehow shoehorning it into an android app would be more useful than adding to a OS I probably won't have access to for years.
Its going to need to be cross platform for the laptop users anyway.
Note that as a startup idea you don't need to be .mil grade and encrypt and verify every packet. Something as simple as a notification pop up along the lines of "Holy Cow you are in Great Danger!" would be more useful than the present nothing.
I don't know how all VPN systems work, but if you are using ssh -D I believe that your first connection creates keys for your computer pair in ~/.ssh/authorized_keys and then future connections are fairly secure regardless of who is eavesdropping.
What is Starbucks' incentive for making wifi faster? Do they really want to encourage the people who sit there all day on laptops?
Will the Google wifi require a login via web browser before given internet access? My phone is always auto-connecting to attwifi but it ends up paralyzing my phone because nothing works until I go to the browser and agree to their terms.
Starbucks derives much (if not most) of their success from product positioning. It's no secret their coffee is just meh in comparison to any other coffee shop. People frequent Starbucks because they love the brand. A lot of things comprise a brand. The logo, the music, the smell, the brown tiled floor, the atmosphere, and yes, the wifi they offer.
If Starbucks partners with Google it isn't really a play to make the free-loaders more comfortable from an operations standpoint, it is about improving their brand image. Once you know they have faster wifi, it will be one more aspect of the brand that you are drawn to, even if it isn't to use the wifi. It may sound absurd, but I believe this is more of a marketing play than actually improving the customer experience.
Starbucks coffee is actually pretty darn good. Sure, if you live in SF, Seattle, Portland, or New York City, you have access to niche coffee shops that can really do better. But objectively speaking, Starbucks coffee is really about 90% as good. And for the rest of the country it _really_ raised the quality of coffee. 20 years ago you got bitter swill anywhere except the above trailblazing cities. Today you can get a properly pulled espresso with perfectly microfrothed whole milk anywhere in the country, which is a pretty mean feat. In a decade or two we've hit Italian levels of coffee quality.
There may be objective qualities that have statistical correlations to people's "good/bad" assessments, but that doesn't stop "good" and "bad" from being subjective.
Ehh, nearly every city I've been to with a population over ~20k has had noticeably better coffee in at least one local spot. It's not just the major cities. But yes, comparing the average cup of coffee before/after Starbucks got big is a night-and-day comparison.
I just wish everyone wouldn't try to taste like Starbucks, now. They can't do it half as well as Starbucks does, so you just get worse everything.
I agree. My favorite is Philz, but even compared to that I would say Starbucks is at least 90% good. I would even go further to say that their Americano is sometimes better than a lot of what those hipster-famous coffee shops serve (but it's subjective because i don't like tangy flavor)
I don't think I've ever lived in a city that doesn't offer at least one better option. To me Starbucks is just a safe bet. If I'm in a new town and in a rush its a guaranteed commodity. Their mixed drinks are pretty good, but that's because they are so loaded down with sugar and syrups you can't taste the coffee. I usually avoid them since I'm not one for sweets. Their straight black drip is below bar. It used to be laughably bad but a couple years ago they made a serious effort to improve it.
To me they are like Wendy's. Yeah, not the best, but not bad. I can do better at home or if I know the area... but that's not always the case.
I usually avoid being subjective, but Starbucks coffee is pretty bad, and most of their barista are hardly knowledgeable about coffee. Once, I asked for an espresso to stay, and it took them 5 long minutes to find an espresso cup. (I can't handle drinking espresso in an americano cup to go.) And talk about grinding.
Anyway, I live in NYC and as you said, there are plenty of local coffee shops with awesome coffee (Gregory's, Kaffee 1668, grounded), so I can't complain. But it'd be terrific if Google could support these better coffee shops instead of just following the masses, because they provide a much better experience.
Interestingly, Italians, from what I can tell, are not "coffee snobs" at all. They drink espresso, cappuccino, macchiato, etc... and just expect it to be pretty good, but I've never seen people discuss it much, like you might with wine.
I don't drink the stuff at all, so have no opinion on the quality anywhere.
"My phone is always auto-connecting to attwifi but it ends up paralyzing my phone because nothing works until I go to the browser and agree to their terms."
"Say hello to my little friend", which neatly encapsulates how I feel about stupid web login screens on wifi and exactly how I feel each time this app bypasses them for me:
aka "Wifi web login". Can be a bit touchy with some but I've not found one yet that it doesn't eventually work on. Once configured, it just transparently sits in the background, at work, or restaurants, or the library, or pretty much anywhere.
I would imagine there exists an iDevice work-alike.
Unfortunately for us iDevice users, invisible background tasks like that are a no-go. Apple simply won't let you ship an app that sits around and invisibly logs you in like that.
Fortunately, iOS itself is smart enough to notice when your network has a login page like that, and will display that page on-screen automatically so you can click through it. Not as smooth as that Android app (which I'm now jealous of), but at least things don't invisibly fail until you figure out what's going on.
It's likely going to get much, much worse. Try to sign in to Starbucks wifi in Italy & much of continental Europe.
That's the future, sadly - Italy for one requires you to register with your passport or other ID in order to access free wiki. It's an anti terrorism measure. Every hotel you stay at requires you to sign on individually, in a manner connected to you personally, every single time.
> Do they really want to encourage the people who sit there all day on laptops?
Presumably they do. Sure there are arguments against it, but on the pro side: if people are going to spend a day on their laptop, do you want them buying your drinks every X minutes/hours, or do you want them elsewhere?
If you have their combination loyalty/prepaid card and make more than 5 purchases on it a year, you get free refills on brewed coffee during a visit (so long as you bought the first cup with the card).
So they actually don't mind if you come in and order the cheapest cup of coffee they have, camp on a table all day, and keep coming back to the counter for free refills. They just care that you do it often, and that you minimize their credit card fees when you do it.
They pretty much treat that refills as equal as actual paid refills ($0.50) and they won't even know you are getting free refill until they scan the card, too.
Of course, if they know you enough, they'll eventually figure it out by looking at your face.
Only store I had a problem was in Nacogdoches, TX -- somehow, they didn't allow refills -- citing some health law disallowing it , so I guess it can vary by store.
Since almost all stores are owned by Starbucks (some stores in airports and malls are not corp. stores), shouldn't they have a clear, firm wide policy?
For all the store I've visited before, other than indirect operation stores, pretty much have very consistent policies regarding pretty much every aspect of their operations.
The one store I was treated differently (which was a direct operation store) was little surprising to me, but might have been something to do with bizarre laws in their local jurisdiction, in which case they wouldn't have much choice.
It could have been better if they could accommodate me somehow, though...
I've spent hours in a few different Starbucks in the south bay before. My observation has been that most people either get their order and leave, or stick around for an hour+. The number of people who sit down just long enough to eat a muffin or something is pretty small.
Why is that argument stronger? You think Starbucks hasn't run the numbers/experiments to test whether a warm body on a laptop is better than the same body ducking in and out? I would bet money they have done a frightening amount of analysis on this question.
Also, to your specific point - when people choose restaurants they often go where there's a crowd (the food must be good here), not to an empty restaurant. The situation is a little different because Starbucks is so pervasive, but don't discount the customer draw of a bustling store full of people.
Over the years just kind of casually observing the placement and layout of the stores, it seems that they are maximized to be easy to access by the morning rush crowds as well as the lunch and after work rushes. Also, a pronounced increase in drive-through locations.
I could imagine that the majority of a location's revenue comes from walk-in/out business during peak times, and the tables could stay full with little rotation all day without appreciable impact on revenue, while solidifying the chain as a go-to "third place", especially for the business class.
i'm guessing there are very few people who go to starbucks, see the seating area is full, and leave. If they're already there, they're going to buy something. If it is busy they might order and then go to a nearby park or something, but once they're at the door they're pretty much a guaranteed sale.
I hope that Google finds a smart workaround for the need to login on a web-browser - it is an annoyance every time I walk by a Starbucks.
There are a lot of subtle ways of regulating how long customers stay. In recent years, I've seen Starbucks replace cushy chairs with stools and rigid chairs that most people wouldn't want to spend more than 20 minutes in. I've seen more than a few restaurants block off wall sockets. And music can be effective at dissuading loiterers without them being too aware of it - even innocuous coffeehouse stuff can be fatiguing after so long.
Yup. When I travel, I look for a Starbucks when I need to get online (or want to get a coffee and space out on Facebook for a little while). That can't be an altogether bad thing to have people associate with your brand.
(1) conflates services and infrastructure, which is sloppy / oversimplified.
Google mostly gets from 1 to 5 by getting folks to use its services, not via its infrastructure directly. The infrastructure is primarily there to ensure that folks use the net more overall (and by extension, its services.)
That's not necessarily a bad thing: in return for a service you want to use, you're being shown ads targeted at you. To ensure you consume those services, they're made easy to access.
I wouldn't be too surprised if Google analyzed their traffic and discovered that a lot of people were getting frustrated while logged into Google properties due to slow, crappy connections in Starbuckses, and decided to solve that problem once and for all.
Starbucks is, from Google's perspective, a major ISP with somewhat poor service that's very concentrated geographically. That last part means it's very cheap to just swoop in and spend the money to solve the problem definitively.
I have no idea if this is actually the case, but it wouldn't surprise me if Google has decided that they're losing $BIGNUM from bad connections at Starbucks, and that it'll only cost $BIGNUM/2 to solve it with brute force.
> Google mostly gets from 1 to 5 by getting folks to use its services, not via its infrastructure directly.
That may be the case mainly in the past. But with technology evolving, I do think it's tempting to go further and extract valuable data from network traffic as well, which in turn can be used to create an even more precise picture of the user.
Perhaps. But that's speculation about the future, which needs to be separated from the current facts. Don't blame people for things they haven't done yet.
Well on the flip side they've been removing outlets and quietly replacing workspace-ready tables with shorter cafe tables that are hard to work for a long period of time on. So while the wifi would be "faster" the starbucks itself isn't conducive to a work session that would last longer than an hour or two
My experience is probably not typical (I move around a lot and rarely go to the same Starbucks more than a few times) but I'm only about 70% sure that any given Starbucks will have a connection at all. So I only go in if it's really convenient or I have no other options. That's to say nothing of the speed or reliability - both of which are typically mediocre to terrible.
If I knew with certainty that every Starbucks had a fast and reliable connection, I'd probably patronize them 50-100% more than I currently do. Hell, I'd be on my way out of this hotel in search of a Starbucks right now if I knew that!
If you were nervous about Google knowing all of your search history, e-mail, and youtube activity, how do you feel about them being able to snoop on literally every packet you send and receive? Also as a bonus they can now track you by your MAC address.
"Google has long invested in helping the Internet grow stronger, including projects to make Internet access speedier, more affordable, and more widely available."
Speeding up the wifi at Starbucks seems like it's about a 20% fit with this mission statement. Ie - if this mission statement was really your goal there'd be a LOT more options that Starbucks. Why try and paint this as anything more than a mutually beneficial commercial agreement between two parties?
To be fair, they have to start somewhere and Starbucks are extraordinarily widespread. Don't forget about all of the other things Google are doing to get Internet access to everyone they possibly can (for example: http://www.google.com/loon/)
Starbucks and the Barnes and Noble cafes near me both use the same incredibly slow AT&T wifi. The most I can pull down when the place is completely empty is ~ 150kbs, so I'm assuming they have a T1 circuit for internet.
To make it worse, the router they use appears to not do any traffic-shaping at all. So with just one person streaming video or downloading app/OS updates, it becomes excruciatingly slow. It's kind of rude to other guests when you use the wifi to catch up on the last 4 episodes of Breaking Bad or torrent something. There's also the wifi congestion when the store is very busy but what can you do about that?
With that said, I'm curious as to what the up to 10x faster than before means. If they offer a 10/15mbs connection, that would be a huge improvement.
All the independent coffee places in town either offer wifi backed by 20-30mbs cable or 6mbs DSL, so they are much better destinations if you know you're going to need decent internet access. Nothing like running 'bundle install' and after 45 minutes you just pack it up and leave because you're getting dial-up speeds.
If you are in the ad targeting business, what demographic is going to produces greater yields - kids with no income or people who blindly overpay for shit coffee?
a) Google gives free access to public libraries in cities that get Google Fiber, and,
b) AT&T WiFi services (nee: Wayport) also has McDonalds as a client.
Right?
With this announcement, you can be sure that McDonalds is considering their options. McDonalds competitors may well approach Google, too.
Getting better WiFi into a large chunk of public libraries requires making arrangements with thousands of cities and towns. Getting better WiFi into almost all Starbucks locations in the US requires just one deal with Starbucks. To me, that makes it seem perfectly natural for Google to focus on Starbucks and to reserve their efforts with public libraries for places where they have existing relationships like Google Fiber cities.
Local prop taxes pay most of the library budget here, and why yes, they do have free wifi. And it rocks.
I'm usually logged into ancestry and google docs while I do genealogy stuff with the microfilm machines. Although as more "stuff" goes online over the years, there's less need to use the microfilm machines. I like the historical map collection and the local history reference books. They also have almost a century of phone books, which is interesting.
Facebook's experiments with wifi in Bay Area coffee shops encourage you to both login w your FB acct and check in to the location publicly on FB timeline.
It's interesting that this is (apparently) a big draw for people to come Starbucks. I kind of wonder what the big use case is.
I assume most people showing up at a Starbucks have smart phones with data plans. So it's probably not for phones, but tablets and laptops. But then I also assume the speed is not that great as the Wifi will be super heavily congested. So you are probably not getting high bandwidth and a lot of stuff will probably be blocked or throttled anyway. Which means most people could probably just tether to their phone and get the same performance.
So then it's for a) those people who can't tether or b) those who can't be bothered tethering?
These days I've virtually stopped using any kind of free public Wifi because I don't trust the security and my data plan is enough for most of the things I want to do anyway.
You've hit the nail on the head. Tethering is somewhat of a black art in the US, as carriers have started clamping down on tethering users in order to charge them $20 or more to access the same bits via another device.
Most Android users don't even have the standard tethering options here, as carriers mandate that be removed and replaced with their own tethering apps that enforce this restriction and protect their revenue stream.
And besides that, most users aren't technical enough to understand that you can access the Internet from your laptop through your smartphone.
And besides that, most carriers have limited data plans, so getting lots of work done over a tethered connection is not exactly feasible.
I'm not sure that people having to do their homework at places like Starbucks and McDonald is something to be proud of.
Edit: I'm of course refering to the linked WSJ article. It's of course good that the option exists, but sad that the poor basically has the same access to the internet as someone hit by a hurricane.
Living in a college town, you'd be amazed how many people meet up at Coffee Shops to work. Personally, I'm like that as well. Something about getting out of the house and dropping $2 on coffee makes me force myself to work without distraction; it helps get you in the productive mindset.
Plus its a great way to retain customers; I'm more likely to go to my coffee shop of choice because of the hours I've sat there. Convincing me to 'move' shops needs a drastic measure (like putting Google in their store).
..now if my Starbucks would only have more outlets...
I hope you don't mean for the students themselves. I live in NYC and know at least two people who don't have a connection at home... I'm sure there are thousands more.
Most (if not all) of our public libraries have Wi-Fi, but at any given location you're more likely to be near a Starbucks than a library.
Thank god. As someone who routinely visits coffee shops to help maintain some semblance of sanity as I work from home most of the time, entering a nearly full Starbucks almost always means the internet will be sluggish. Hopefully this means I can actually stream music and work at the same time now.
Err, usually the slowest part of connection is WiFi. I never use wifi if I want to get reasonable throughput or ping. I don't even live where the WiFi band would be unreasonably congested. (connection 1/1 Gbps FTTH)
I'm what you would consider a "Knowledge Worker" and I haven't connected an ethernet cable to my laptop for Internet Access in 18+ months. With 802.11ac now available, the slowest part of a connection is almost always going to be the speed of your internet connection.
I just came back from Singapore, and, between our Corporate Office down in the PSA building, the WiFi at the St. Regis, the various hotspots I had to work with on a client site, tethering to my iPhone, and the 2 mbit/s proxy on the client site - I can tell you I would have paid upwards of $20/day to get a solid 20 mbit/s connection if I could have - I never once saw a sustained connection of > 10mbit/s - my WiFi was never close to being the slowest part of the connection.
Public Wifi is near-universally terrible. Take WWDC this year. On the wired download stations you could get 300-600 MBit downloads. On WiFi you were lucky to connect at all. Heck, my experience is that wherever WiFi is available, 3G is usually both more reliable and faster.
At home, living alternatively in Japan and Sweden with 100 MBps connections in apartment buildings (=everyone has their own WiFi access point trying to squeeze in 100 MBit/s of BitTorrent uploads), WiFi is a continual pain in the butt and is often hampering my download speeds. I can't wait for 802.11ac, but I just got this fully-decked-out laptop last year and am not looking forward to spending another $3000
How many days were you in Singapore? Probably the best way for a traveler to get a solid unwired connection would be to buy a new phone (unless you're lucky enough to already have a device with compatible LTE bands) and then enjoy the wonders of prepaid LTE and tether ($7/week/GB).
Maybe it is just because most people haven't upgraded yet, but I have a 4G connection over 90% of the time on my phone (with good latency) and I can't remember a speed test under 10 Mbps down (20-40 typical). It's good enough that it takes my home WiFi (backed 200 Mbps fiber) working near-perfectly on the 5GHz band to beat it.
I'm not sure having a synchronous 1Gbps connection really counts as "usually" in most of the world, but I guess in that case you'd want to avoid wifi to get the best out of it.
For the rest of us, wifi is far from the bottleneck. I'm currently working from a holiday home in the country where I'm sharing 2Mbps down with 4 other houses.
I'm not sure having a synchronous 1Gbps connection really counts as "usually" in most of the world, but I guess in that case you'd want to avoid wifi to get the best out of it.
You must be American. :)
Seriously, most of Europe has had gigabit fiber, flying cars, and robot maids for nearly a decade now and in South Korea they're slowly phasing out the government in favor of a benevolent AI.
I guess that would make a difference if you wanted to transfer a few hundred gigs of video from an SSD array while at Starbucks, but for the rest of the world I think the improved speed of this free WiFi is a good thing.
This is part of a bigger plan to turn Google into a mobile carrier. 18 months from now, you will be able to ditch your T-mobile, Verizon, Sprint subscription and connect to Google Wifi nationwide using their Starbucks mesh.
>If you’re in a Google Fiber city, we’re hoping to get you a connection that’s up to 100x faster.
I look forward to seeing how this new competition begins driving Comcast and the other larger services to improve. In SF, if I had the choice to switch from Comcast to any other company that could offer similar speeds I would. Seems like Google Fiber makes Comcast and ATT's speeds look laughable.
None of their tax was unpaid - they paid exactly what they were legally obliged to pay. (Note that I don't agree that they were paying a morally satisfying amount of tax but ...)
Blaming Starbucks for a broken tax regime where the government bends over backwards to create obliging tax rates and structures for big companies is dumb.
I bet this doesn't please AT&T one bit. I always thought Starbucks internet connection speed have been considerably faster. It always felt like I was surfing in the internet on DSL circa 2001.
Well, the only way to avoid that appears to be to go off grid. Perhaps some coffee shops will begin to advertise lack of wifi, and perhaps even relocate to areas without any mobile phone coverage?
Google Fiber is a nice business opportunity for them, but that requires infrastructure. This comes with none of that. It's not even a precursor to Google Fiber rollout. Either they are getting a similar-quality leased line to _one_ place, or it's just an ad. And even if the Starbucks uplink is "Google Fiber", the wireless connection itself will be a huge bottleneck, as it always is when you have more than around five devices communicating at once.