The Infamous 'Up To' Broadband Qualifier

Oh yeah, my gigantic selection for broadband speeds. I can get RoadRunner, or I can get SBC. That's such a huge selection that they just drive themselves down to where the service is practically free. /sarcasm.

Now let's look at some textbooks on amazon.com. This is what happens when you have 100 people selling the same product, the price drops.
»www.amazon.com/gp/produc ··· -5372006

$110 new, $3 from one of the 68 people selling it used.

Verizon does that but only to give back what their crappy implementation takes away(PPPoE/A)

there needs to be some guarantees here even if your service is best effort. say you are a provider and have 2 tiers(10000/1000 and 5000/512).
-
now if i subscribe to the 10000/1000 tier, there should be a guarantee of "Up to 10000kbps down and 1000kbps up but at least 5001kbps down and 513kbps up" and if not then that would count as an 'outage' and said user would get a credit for that day on their cable bill.
-
this is exactly why i downgraded my service(cox communications) couldnt even give me the speeds one tier down from mine was capped at.

I have to wonder how much of this is a tempest in a teapot anyway. I'm sure there are bad apples, but I suspect the reputable ISPs give the vast majority of their customers a pretty good approximation of what they're paying for already.

Apparently I've been lucky in that by and large Verizon has delivered the speed promised. They set marketed rate lower then sync rate to accommodate and hide network overhead. That way when a customers run performance tests resulting speed is close to advertised speed.

First-Mile access is difficult to market because so many variables affect user perception. What I think most folks are really upset about is when network slowdowns or instability persist for days or weeks and the company hides behind the its only a "best effort" service.

At the same time there may be many factors outside the companies control that affect end-to-end performance.

/Tom

Bad analogy with the gas thing. The gas only has to go from the underground tank to the car and you are comparing overall quantity to speed. It doesn't match up.

Now the ISP's COULD have a speed test from different points on their network to your router. That and have regulations say that the speed test has to be withing 5% of what is advertised. That would make things a lot better. They can then say more certainly that it is or is not a problem inside your network or with your pc.

a truth in advertising clause is being pushed... about time...

No need to argue among ourselves, now.

The point of the discussion is that the phrase "up to x meg" shouldn't be part of the TOS at all. The Australians seem to want the speed accurately stated in the TOS, perhaps as a minimum speed guarantee so that people wouldn't be disappointed when their speeds never reach the currently advertised higher amounts. That way when their speeds go higher, they're all happy campers.

If I say that my annual salary is "up to $50 million", that doesn't make me a millionaire, although what I'm saying is technically true.

Am I missing something here? My Comcast HSI was advertised 6 down, 768 up. Most of the time I got that. My FIOS is 15 down, 2 up. I seem to always get that. Obviously if your node is oversold services, it can occasionally kill your peak speed when everyones on at the same time, and few websites will ever reach the peak speeds anyway. But if you can speedtest the advertised speeds, what's the big deal? I think the "up to" sham comes mainly from DSL. Time and time again I've had to explain to co-workers and friends that DSL speed is distance limited. It's obviously unclear the way it's marketed or they wouldn't be surprised when I explain it to them. And I can understand getting mad about subscribing to "up to 3Mbps" and getting only 400kbps because your way too far from the CO, especially when the guy getting near 3Mbps is paying the same amount as you. Cable and fiber aren't distance limited, so it's a different standard altogether. But, yes, DSL providers should get slapped for not being straightforward, and charging accordingly.

An excellent response!

Like you, I can empathize with folks that don't understand that a $30 - $80 Cable / xDSL connection cannot realistically perform the same way as a $300 - $5000 SLA'ed circuit does 24/7. To vilify a company for using the phrase "up to" simply isn't fair though. I've voiced my stance on this subject passionately here at BBR many times, so I'm not going to spout out my opinion regarding consumers that don't bother to research what they purchase again. I do want to say however, that I don't feel that advertising a clearly defined minimum line speed would be too much to ask. On a 1Mbps connection, why not at least advertise the bare minimum as 384kbps or 256kbps?

Speaking as a future ISP owner, I would take enough pride in my product to ensure that I could at least deliver a set amount of bandwidth reliably. If customers weren't consistently getting at least 50-85% of the speeds I qualify them for, I'd acknowledge that something needed attention.

Full speeds 24/7/365 though? Hell no. Get off my network or sign an SLA contract.

- Tate

Comcast is a profitable company, but their earnings come on a 9.28% profit margin- not exactly what I'd consider "obscene": »finance.yahoo.com/q/ks?s=CMCSAMy cable speeds in Minneapolis have increased, and there is only one cable provider and no fiber or citywide WiFi offerings. Qwest offers 7M/896k DSL though you have to be relatively close to a CO to be eligible.

I don't see the "sham"; most DSL providers provide multi-tiered service- they know that not everyone will be eligible for the highest rate they offer.

If a distance or line considerations mean that customer isn't getting close to the speeds they signed up for, then I'd agree that they shouldn't be held to any service contract.

As mentioned earlier, the sales office can't predict what speeds a given customer will get with 100% accuracy. There will be times when they qualify someone for 5M when their line only ends up supporting 3M. This isn't intentional dishonesty, but rather the fact that the facts on the ground don't always match up with those on paper.

Edit: Just ran three speed tests in a row on my 6M cable connection, they came in between 2.4-2.9M. Being a shared local loop, cable is typically more susceptable to speed fluctuations than DSL.

I agree, though from a marketing standpoint I'm not sure that would go over well. Tell people they're only guaranteed they'll get 384k on a 1M connection from the outset and they'll assume the worst, even if in reality they'll get close to 1M 90% of the time.

Profit is profit and when the CEOs get massive bonuses, the profits are big enough to afford it. Nice qualifier. Most people do not live that close to the CO and DSL is distance sensitive. The farther away, the slower the speed.

As for cable speeds, they were at 10Mbits, went down and are now going back up but 8M/768 here is an extra $10/month.

Say you have two competing providers with identical networks and service plans. One markets their plan as "up to 6M", and the other offers "3M guaranteed". Which one do you think people are more likely to sign up for?

I agree that the ambiguity can be frustrating, but given that building a shared-resource network that supports 3M 100% of the time is much more expensive to build than one that offers 6M 90% of the time, there isn't an easy solution.

Granted, all of us here know enough to run a line test and see what we are getting as far as speeds but there are lot of people who may just end up thinking that they are getting a certain speed when they are not.

You are correct, they aren't "misrepresenting" a damn thing.

"Most consumers won't understand what 'up to' means and then they are significantly disappointed when they don't achieve those speeds..."

Without even reading one more post in this thread, you can already see that statement is exactly right.

People are either truly stupid by not understanding what "up to" means, or they are using *selective comprehension*.

*Yawn*, who cares...let them change the wording to "7MB, guaranteed" instead of "up to 10MB" and people will think they've pulled one over on the ISP's when they get 8 or 9MB.

Problem solved.

I'd agree that CEO compensation is often far too high, but in Comcast's case represents a very small percentage of their revenue or profits. For reasons difficult to fathom, even CEOs of companies that lose money often get outrageous compensation packagesI know duder, that's why I mentioned it

I agree that it's a bad analogy, but to be fair to Karl (whom I've criticized more than once) he represented that analogy as the view of some users, evidenced by numerous posts right here on BBR.

A better analogy would be gas mileage. Car companies often advertise fuel efficiency ratings of "up to x miles per gallon", even if few drivers will actually see it since they aren't driving on an EPA test bed- there are simply too many variables involved to guarantee how many miles per gallon a given car will actually get.

EXACTLY. I wish some of the dunderheads would get that through their, um, err heads and stop trying to derail the issue with crap about TOS, bandwidth hogs, capping up, capping down or being an "informed buyer",

The issue is providers via marketing hype LEADING you to think one thing but in reality giving you something less than what they want you to believe. The analogy about the gas pump could not be any clearer.

If the filling stations started doing what broadband providers do I would bet there would be ZERO misunderstanding.

I didn't read every comment, but I didn't see this point mentioned.

Your connection to the internet might be over DSL and the server might be connected to the internet with cable, fractional T-1, ISDN, frame relay, dedicated analog modem, etc. They can't possible guarantee the speed between your computer and the server will be faster than the server is capable of serving. The server could be an overwhelmed ancient Sun Sparc box or something with a dying SCSI hard drive and defective RAM. There could be water dripping in the case, the CPU fan is clogged with dust and on and on...

If the speed is only guaranteed to the CO (or equivalent) it's useless. When I had DSL, that was the situation for me. The DSL modem had a perfect connection to the CO at the full advertised speed, but the connection from the CO to the internet peaked at about 1/2 of that. It was exactly what they advertised, but it was only 1/2 as fast at the same time. I don't know how good the Australian government is, but I bet things like this will escape them.

It all seems so simple in my minds eye... advertize "minimum" speeds, they'll be happy when they get more... why do they think they have advertize the MAX speed you'll get at 4:30 am on a sunday...

/whatever

AT&T includes the transport overhead of their internal network (Layer 2) in the speed claim.

I believe they use ATM so packet segmentation uses up a lot of additional bytes for headers. 53 byte packet, 48 payload and 5 byte header if my lousy memory is somewhere in the ballpark.

So for example "up to 3000 kbps" will NEVER exceed about 2500 kbps of data at layer 3 and above.

This is deceptive because transport overhead is easily calculated unlike burstiness effects and other statistical guesses. Although they declare their true meanings in the fine print, they should be required to say, "3000 is the cap, and we use 500 of that just to get you to the internet. So you really have 2500 to play with when things are otherwise at their best."

I'm getting 10.8 mbs from Charter Cable! (on a 10mbs line). Who'da thunk Charter would rise from their debt to compete with Verizon's FiOS? It can happen!

I think this is a good idea, but I see the benchmarks as being the difficulty in enforcement.

Where will your official speed tests be done to? 3rd parties could be used as an element of trust, who would have to set up test nodes at peering points, etc. It isn't like comm. companies would have a hard time monitoring bandwidth to different locations themselves to keep on top of it, either.

It is frustrating that the "up to" claims really have no bearing on what you could actually experience. If you think about it, comm. companies (especially those in areas with little competition) have no reason to keep your speeds up to snuff. There is currently too much subjectivity involved in it, same with transfer limits. The industry definitely needs to start providing clearly elaborated service levels for your monthly fee.

So go start your own company and compete....once you read your cheap textbooks.

Hob

"Speaking as a future ISP owner, I would take enough pride in my product to ensure that I could at least deliver a set amount of bandwidth reliably. If customers weren't consistently getting at least 50-85% of the speeds I qualify them for, I'd acknowledge that something needed attention."

You are going out of business baby...

Hob

The thing that bothers me is that "Up to 3000 Kbps" means anywhere between 0 and 3000 Kbps. That means that your service could be slower than dial-up and your ISP could claim that you are getting what you paid for.

They should sell their service using a minimum/maximum speed, so if you bought 1000/3000 Kbps, your speed would not drop below 1000 Kbps and never be above 3000 Kbps.

Would that be so hard to do?

Except that minimum speed could be 0bps. No one can guarantee what speeds are across the Internet. The best they can ever guarantee are speeds to a speedtest box in their main data center, or maybe as far down as the local CO/Hub.

Customers only care how fast they can download something. That's why "Up to Xmbps" should be acceptable. If people can't be made to understand what "up to" means, then that's not really the fault of people selling the service.

Heck, they could even do a special acceptance test for the customer, hook up a speedtest box at the CO end, go to the customer's house, and show them that they are getting all the way up to Xmbps. And tell them, "this is your max speed. Sign this that we have tested out that your speeds work UP TO Xmbps."

But no one can guarantee a minimum or maximum speed to any site on the Internet, and that sounds like what they want to happen. No one can say from one minute to the next how fast your download is going to be from one particular site out there, not to mention the millions of possible sites out there.