Its worth considering the WhisperNet Tax in-depth because its a great example of what happens when a large shared system like Kindle’s WhisperNet suddenly gets freeloaders who use resources to an inordinate amount.
I’ll be very clear from the onset that I’m not saying freeloaders are wrong or ought to be punished – any stable system always attracts freeloaders and they often kill out each other and help the overall system. However, sometimes the system itself or the entity patrolling the system has to step in and remove or restrict freeloaders.
A central assumption we’ll be using is that Amazon has to pay 12 cents per MB of data transmitted over WhisperNet.
Starting Off: A Shared System in a state of harmony
Initially most people on WhisperNet were buying enough books, and not transferring or emailing too many documents.
If you consider that a typical Kindle user probably -
- Bought 2-4 paid books.
- Transferred free books and personal documents and surfed corresponding to 3 MB data transfer over WhisperNet.
We get a 36 cent per month WhisperNet charge. That’s something that’s reasonably covered.
Perhaps there’s a small amount of the Kindle’s price set apart for this. Much likelier is that a portion of the kindle edition book price is set apart for this WhisperNet usage.
The Inflection Point: More and more people and companies pushing the boundaries
People tend to push at the rules and limits of any system they’re in. With WhisperNet this meant companies saying why not let users do things which were -
- Borderline Freeloading i.e. download free books directly over WhisperNet (like FeedBooks allows for).
- Straight Freeloading i.e. email blog feeds to the Kindle (like KindleFeeder allows for).
There’s an easy rationalization here – a step by step argument to justify freeloading off the system and even blame it on the system i.e.
- Blogs are free on the Internet.
- It’s ridiculous that Amazon charges $1 or $2 for blogs.
- I paid $359 (or $399, or $250) for a Kindle. Blogs should be free.
- Surely bandwidth costs nothing, and a blog isn’t much anyways.
- Therefore it’s justified to use a means to get blogs for free.
- Besides my initial Kindle purchase and my book purchases should cover this.
I’ll reiterate that we’re not saying there’s anything wrong with freeloading. We’re here to discuss the effect of freeloading on a system, and not any moral sccenarios.
Breaking Point: A Switch to enforcing WhisperNet charges becomes necessary
At some point, the amount of transferring of free books and blog rss feeds becomes too much. We now have a set of users who are doing one or more of -
- Downloading dozens of free books over WhisperNet. A few MB of data transfer per month.
- Having 20-30 Blogs emailed to their Kindles. Easily another few MB of data transfer per month.
- Emailing larger and larger personal documents. Again, a few MB of data transfer per month.
All of this adds up – To the point where perhaps 5% of Kindle Owners caused the whole system to become unstable.
If Amazon were to actually release data on WhisperNet usage, it wouldn’t be surprising to see 5% of ‘top users’ who were accounting for 50% or higher of WhisperNet usage. Not only is ’top users’ usage not balanced out by their purchases, its likely that even other Kindle owners’ purchases weren’t balancing it out (not that they should).
Leaving Amazon with no option but to tax high Whispernet usage.
Why can’t Amazon just restrict Freeloading Services or Freeloading Users?
Its mostly a legal issue, with some PR considerations thrown in.
- Firstly – People tend to always place blame on external factors. No one is ever going to say – “I shouldn’t have been transferring 25 blogs to my Kindle wirelessly.” They’re going to go complain on 50 forums about how you have to pay $359 and still can’t get any use out of the device and create a PR nightmare.
- Secondly – Amazon can’t go after every small company offering a service that leeches off of WhisperNet.
There are now not only sites that offer free streaming of blogs to Kindles over the WhisperNet, they also offer premium subscriptions.
Consider the extent of freeloading where not only is a company encouraging leeching off a system, they are even offering ‘premium leeching’ and making money off of it.
Legally, its easy to attack Amazon via one of 2 means -
- It was behavior Amazon encouraged by not clamping down on it for the first 1.5 years of WhisperNet.
- Favoritism as it’d be rather difficult to go after every service, and the ones that were taken to court would point this out.
Add on the PR implications of ‘free vs paid’ and ‘david vs goliath’ and amazon’s hands were tied.
The only option was to tax excessive usage.
Amazon actually did the right thing
Rather than let freeloaders overtax WhisperNet, and kill it for everyone, this tax of 15 cents per MB of data transferred ensures that users who’re using WhisperNet a lot pay for it.
Applying a tax on Whispernet means, as Andrys pointed out, that someone who’s having free blogs streamed to their Kindle now will end up paying $4 or more per month. They’ll finally be paying for the extra bandwidth they’re using.
Amazon pays 12 cents per MB of data transferred and a 15 cent charge hardly creates a compelling revenue stream. My money is on the reason for the tax being the abuse that a small percentage of people were subjecting WhisperNet to.
Filed under: evolution, thoughts | Tagged: kindle whispernet, whispernet tax
It was always in the cards that Amazon would see usage of whispernet and they provided a means of collecting money for non-store use from the start though they didn’t start charging for it yet. The charges for individual items was also built in but files downloaded are likely larger than they envisioned.
The Kindle would not be $360 without whispernet costs and so the ‘freeloaders’ have already paid for some of that whispernet already. So they are not freeloaders.
I’ve defended the cost of the Kidnles on the basis that whispernet costs were calculated — else why should it be $360? And it’s the defining difference between the Kindle and other untis. And again the reason for the higher price.
Furthermore, Amazon put bookmarks on the K1 and K2 explicilly giving people a way to get to other sites such as BBC news and MSNBC. They, in other words, encourage the use.
They’re mobile sites. They cost money to go to but Amazon wants people to experiment with that and then give feedback on it.
They intend to make money from it someday when the browser becomes worth using. For now, only those bent on being able to access wireless when out will have the patience. I’m one of them.
RSS feeds are sent to the world for reading on one’s chosen devices, whether computer or phones or, now, the Kindle.
If it becomes too costly (and it’s my firm belief that whispernet is done on contract for a lower flat amount for sprint’s ‘unused’ bandwidth). I’m sure Amazon gets a very good deal. It is still costly, for sure, but we have paid for a good part of it already and Amazon leads us to other sites themselves with their bookmarks and gets people used to using it.
The 12c is based on an industry user’s inside info.
That would be an averaged estimate. I don’t doubt they probably are seeing higher usage than they expected.
People use the system as they’re encouraged to do.
If Amazon needs to start charges it will be because they encouraged web browsing on an experimental basis, as a revenue stream later. Their ads also blare no monthly charges and it’s naive to think people would not use it for anything but the store when their web browser bookmarks lead you to other places.
It’s a bit like credit cards – encourage people to use it and get used to it. Me, I like wireless for searching when I need to outside. It’s a huge feature for me.
I have 7 feeds come to me by the excellent Kindlefeeder while I pay Amazon plenty to have the special formatting and navigation for NY Times Latest News, New Yorker, Slate and even Newsweek (I already get Time, hard copy). and each Sunday NY Times. People who use it as I do are not freeloaders.
We use differnet avenues available to us. And Amazon will need to be straight with people but they are going to be torn because the do see it as a future revenue stream and encourage the exploration and use, with feedback to them. I see improvements over the last 2 months so they’re apparently working on it.
As for ‘profit’ – a place like Kindlefeeder takes money to run and a lot of programming labor. Choi is the most responsive programmer I’ve encountered. Servers chugging out aggregated fees do cost money to run. So does keeping things running.
they key sentence here is – People use the system as they’re encouraged to do.
Now with the fees Amazon is changing behavior.
Andrys – you fall in between freeloaders and users who have no idea all these other options exist.
The freeloaders are the people posting on forums wondering if there’s a way to tether the kindle to their computer and get free internet.
kindle feeder might be a good service – however, your being a kindlefeeder user means you are disregarding that the programmer is using bandwidth he’s not paying for to provide a service and earn goodwill.
I’ll say it again. The price chosen for the Kindles does factor in the wireless costs and Amazon is not so naive as to base estimated wireless costs on only the store, since they gave us the MENU to explore BBC News, MSNBC news and a few other sites in Real Time. No 1-minute visits expected.
If I use the wireless in an even less $Time$ spending way than using THEIR bookmarks to go just sit on BBC News or MSNBC news for an hour or so on end, I and Kindle owners who use Amazon’s Kindle web menu are just not ‘freeloader’ and is not a justifiable word for people who quickly download what is usually less than a 1-meg file.
It’s just too inflammatory. Amazon business heads would weigh what Kindle owners pay for from Amazon against what they cost in the few seconds from freely available feeds on this Net.
Amazon is there to make a profit, so they are charging for the email conversion and delivery to the Kindle. The response to this has been anything but severe, from the Kindle community. It was expected and we get the free.kindle.com method if we want. It’s a reasonable business decision and would have been made in any case once they knew what the costs were and what they could recoup or get a profit from.
Also, kindlefeeder has written Amazon to give it a chance to say No Go but has never received a response. It is actually a Plus that the Kindle is capable of this, as the common complaint is that one can receive only Amazon-controlled material to the Kindle. People who receive feeds are generally avid readers, and people who buy Kindles have money even if they’re careful with it.
Since I’m one of those who buy subscriptions from Amazon, in addition to books, they’re making money from me/us.
In the meantime, there are options at Kindlefeeder to download to the computer or to get a prc file downloaded to the Kindle (as you yourself was thinking of doing in response to a note here yesterday) FROM the kindlefeeder server, not going through Amazon’s conversion servers.
We could cut our feeds to short feeds if we want but you mentioned that exposure was more important the revenue at this point.
We’re talking a Send/Download that’s less than a minute that uses the Kindle’s wireless to receive.
And I am talking the reality of their encouraging us to go Sit on websites with their experimental browser, which is far more expensive for them.
Your Amazon-estimated costs figure should include that also instead of the few cents a month that you guessed they had considered. A business has to think of revenue streams – so someday the web-browsing and sitting on websites they’re encouraging Kindle owners to do will see some kind of tiered plan *IF* the browser ever gets really useable.
Again, there’s a cost that we paid for with the cost to us of the Kindle itself, which would not be that expensive without wireless access that has no monthly charges.
And Daniel is paying for his server usage and his own bandwidth in connection with the delivery. Amazon faces the reality of the Net and is dealing with it.
In no way would I encourage people to spend $360 on the K2 unless it had wireless AND we were able to use it for whatever amount of time spent outside the Amazon store.
No sane person would use this web browser for anything but short lookups. But people report using the webbrowser menu and visiting news sites when having to wait for their newspaper subscriptions to arrive.
As time goes on, we’ll know what the limits are, but at this point no one has a good fix on that. The assumptions used here, while quite interesting, don’t justify the word ‘freeloaders.’
When I did budgeting for all kinds of projects, I always figured the stated cost to include all the labor/transportation/delivery costs, everything I could think of, to justify the cost of our charges.
Tasaka’s figure gives us no basis for the calculation and is hearsay at this point.
Besides the other typos, substitute “feeds” for “fees” in that last sentence. Am rushing out.
Hi – I’m the creator of Kindlefeeder, which you mention. I think your main point is fair, but a few factual corrections are in order, and I’d like to offer my different perspective on the issue, as an independent software developer.
First, Calibre can’t be accused freeloading off Whispernet bandwidth. As far as I know, Calibre does not offer Whispernet as a content delivery option.
Second, while Feedbooks uses the Whispernet to deliver content, this is a form of Whispernet delivery that is not being subject to the new $.15 charge. The delivery method Feedbooks uses, last I checked, is a direct .mobi download via the Kindle web browser, as opposed to a delivery plus conversion via Amazon’s Kindle document delivery servers. Only the second form of delivery is being charged $.15 per use.
Third, my service, Kindlefeeder services a miniscule portion of the Kindle user base. The active users right now are probably in the few hundreds. Also, Kindlefeeder has always offered 3 methods of delivery, only one of which was based on the Whispernet delivery method that is now being charged. Now Kindlefeeder also offers a fourth .mobi download-based method, just like Feedbooks has been offering for a long time.
Fourth, I didn’t create Kindlefeeder to abuse and exploit a loophole in Amazon’s system. When I wrote the code and initially launched the service, I was under the impression that Amazon *was* charging $.10 per delivery, as Amazon announced the price to be back then. It was only later that I found out that Amazon wasn’t charging, and even then I had no idea how long it would be before they would start charging.
The truth was, I wasn’t afraid so much of the beginning of charges, but of Amazon simply backlisting Kindlefeeder from its document conversion service. I was spending two hundred or so hours coding software that could be rendered useless on Amazon’s whim.
But I went ahead anyway, because I believe that the Kindle is a platform and that the platform should be open. Yes, I was hoping to earning some revenue from the service (which is much smaller than you might think), but I also wanted to demonstrate that third-party, independent developers could make genuinely creative and useful contributions to the platform, which Amazon seems to be slow to recognize. The iPhone has become an application platform. Netflix has released an API. And everyone is better off. Why not the Kindle?
I even argued back then that Amazon should raise the charges on personal wireless content delivery if that was the price we would have to pay to let third party developers develop applications like Kindlefeeder and InstaPaper for Kindle users. Here is the proof: http://www.teleread.org/blog/2008/09/14/kindlefeeder-rss-manager-and-feeder/ (see my comments in the comment section).
So while you make a fair point in thinking of Whispernet bandwidth as a “resource” that shouldn’t be over-consumed, I think it’s equally valid to be frustrated at how closed the Kindle is as a platform, and to support and encourage the software developers who try and dare to add useful features to it.
Anyway, that’s my perspective. Thanks for letting me share it.
Daniel,
Recent releases of Calibre do support optional delivery through Amazon’s Whispernet service, and the feature works well.
Thanks for the correction, Steve. I wasn’t aware of that. I think it’s great that Calibre is offering the option.
Interesting post. One question though: You have quoted a data delivery cost to Amazon at 12 cents/megabyte as if this is known. Isn’t that number just an undocumented estimate from LibreDigital’s Guy Tasaka? I’m skeptical of this number as applied to Amazon given its volume and negotiating leverage.
My exact words were – A central assumption we’ll be using is that Amazon has to pay 12 cents per MB of data transmitted over WhisperNet.
There are no other figures available and the general high price of wireless data packages indicates that 12 cents per MB is not too ridiculous.
I am a relative newcomer to to the Kindle and didn’t have the opportunity to enjoy the free delivery of non-Amazon content, and I am a little embarrassed to say that I don’t pay the least attention to how many $0.10 items I buy in any given time period. Since it appears to apply solely to documents emailed to Amazon for conversion then sent, via Whispernet, to the Kindle. Personally, although I am a voracious reader, I can’t imagine emailing more than, say, 100 items for Whispernet delivery. I use my computer to send for conversion and load via usb most non-Amazon content and consider the stuff I do send via Whispernet (usually from my iPhone when away from home) well worth $0.10 a pop. Would I bother with the computer at home if not for the dime? I can’t say for certain as I’ve never had the opportunity, but, as much as I like to believe I would–probably not. Would I use a service to deliver free blogs to my Kindle? No. No way. Not ever. I also lived through the heyday of Napster and have never–not one time ever–pirated music. Or a movie. Or software. Or anything else.
Andrys, I get that you’re uncomfortable with being referred to a freeloader, and I get why you don’t see yourself as one. And, at most, you could be considered in the ‘borderline’ category, and I’m not sure you see why. It’s not that you’ve used Whispernet bandwidth for non-Amazon content. Not all blogs are available from Amazon–my personal favorite among them. So, I DO use the Kindle’s browser to access this blog daily and don’t consider it freeloading because I would gladly pay Amazon for this blog were they to offer it. I DON’T use anything other than the browser–any service–because what services like Kindlefeeder are doing are profiting from bandwidth Amazon provides so you can enjoy it’s offerings, and even a few things they don’t offer. It’s just wrong to charge for assisting someone obtaining for free someone else intends for sale. And when that happens things stop being free. I was a customer of AOL, back in the beginning. But I travelled a lot for business, and a few of the places I went had no AOL local numbers. Because they had local numbers in virtually every place imaginable at that time, I also had a FREE Juno account for places where AOL didn’t have access. Eventually all the free internet access dried up. Why? Because instead of using it infrequently as a backup service people started spending HOURS online on a free service. They tried all sorts of schemes at first to limit the freeloaders, but eventually what was a great thing for most people was spoiled by the people who are now “posting on forums wondering if there’s a way to tether the kindle to their computer and get free internet.” (I love that analogy, switch11! It is SO TRUE!)
The problem, as I see it, is that if it doesn’t feel wrong for you, deep inside, to take more than “your share,” then services like these will always end up h
having to add a “tax” to keep freeloaders from destroying it for everybody. I wonder whether you would do the same in a more visible venue. For instance, if they were giving away free sodas, would you take one and say, “thank you,” or would you fill up your backpack so you’d have free soda for the week?
As to the price being high to include the bandwidth, what have you been smokin’? Cellular bandwidth, right or wrong, is outrageously expensive and the Kindle is VERY reasonably priced considering it’s components. Amazon would not be making a huge profit on these unless you buy content form them. (This isn’t an unusual strategy, Apple uses it with iPods and iPhones, Sony, MS, and Nintendo do it with game consoles. They bank on you buying content.) It doesn’t entitle them to prevent you from aquiring content elsewhere, but it doesn’t entitle you to obtain other content at their expense, either. The content is where Amazon is making their profit. As much as $300~$500 may be to you, it DOES NOT represent a huge profit to Amazon, nor does it entitle anyone to unlimited free bandwidth.
Spooky, wow, this was an older thread.
The feeds are not meant to be ’sold’ – blog owners choose to offer feeds and people can get them for free or pay something to get them in a well-organized manner maybe.
1. Kindlefeeder uses prc files now, and their own server has downloads you can do from there, not causing Amazon servers any load.
2. A combo-feed download takes about one minute. So it costs Amazon about one minute.
3. Anyone browsing with the Kindle to read blogs online in readers or whatever will mean a lot of live reading that goes on for maybe hours, certainly more than one minute.
4. They will then cost Amazon much more than a one minute download via whispernet.
5. those of us who also pay Amazon for specially formatted blogs will download, this way, blogs we would not pay for but the blog owners allow the feeds.
6. Amazon gets a LOT of my money so I feel no guilt about using a one minute download once a day (at most — I usually skip a few days). Amazon factored in wireless use when they priced the Kindle in a way that many perceive as ‘high.’
7. Don’t know where you got the idea I felt the Kindle was too high. I usually say it incorporates 24/7 free wireless costs, and Jeff Bezos verified this again in mid-June when he told the Wired Conference he could have charged maybe $99 for the device and then recouped his expenses and hopefully a profit by charging the usual wireless charges of $60 per month. Or $720 a year + $99 = $799 instead of $360
(now $299).
8. I love to browse on the Kindle when away from home and sometimes I read on the Net when a Kindle blog links me to something I want to see. The bookmarks Amazon sets for us encourages us to do experimental browsing, as do the ‘Google’ and ‘wikipedia’ options on the status bar of almost every area that involves typing (or highlighting) a word.
– Andrys