Apr 20

Tim O'Reilly

Tim O'Reilly

Amazon Sues Alexaholic...Everyone Loses

Alan Graham published a great account of my discussion with Jeff Bezos about Alexaholic (now Statsaholic at the Web 2.0 Expo on Monday. Entitled Amazon sues Alexaholic...Everyone Loses, Alan's blog entry does a good job of summarizing the the conversation, with a lot of background on the controversy to boot.

Jeff and I spent most of the time discussing Amazon's web services like S3 and EC2, and just how seriously Amazon is taking the idea of the web as platform, but at the end, the subject turned to the lawsuit Amazon filed against Alexaholic. I pointed out that when Paul Rademacher launched, the first Google Maps mashup, Google didn't sue him, they hired him, and created an API to make it easier for others to follow in his footsteps. Given the additional visibility and utility that alexaholic/statsaholic has given to Alexa (which is owned by Amazon), I asked whether a similar approach wouldn't be better here.

While I largely agree with the sentiments that Alan laid out, to wit, that it would be far better for Amazon not to be suing Ron Hornbaker, the creator of statsaholic, I do think that it's important to clarify a few points.

First off, I didn't surprise Jeff with the question, as Alan intimated. Jeff and I talked about the issue at length backstage before the presentation. And while we still disagreed, I think that there's more substance to Amazon's position than you would get even from Alan's relatively thoughtful piece, let alone from rants like this one from Pete Cashmore. My regret is that I didn't ask the question earlier in the session, rather than right at the end, when we didn't have time left to get into the subject in more detail.

Here are a couple of the points that Jeff made to me backstage:

  • Unlike and other Google maps mashups, which add value by combining data from multiple sources, statsaholic competes directly with alexa, offering only their data.
  • Also unlike Google maps mashups, statsaholic originally used a name based on alexa's own, and has continued to do so despite requests to do otherwise. While Ron did change the name, he did retain the alexaholic domain, and still uses it to drive traffic to his site. (He originally just did a redirect. Now he has text there that points people to the new site.) I can see Jeff's point. I doubt very much that Google would have been so tolerant if had been called
  • Alexa did negotiate in good faith with Ron about acquiring the domain. Ron confirmed this with me when I talked to him about the situation on the phone. While neither Jeff nor Ron would tell me the price, what I was able to glean by inference suggests that it was a fairly substantial price for a domain name. Ron turned down that offer.
  • This is not a case of Amazon offering an API, and then turning on a developer when they build a successful site with it, as Pete Cashmore incorrectly asserts. The Alexa API has apparently never provided access to the Alexa trend graphs. As one commenter on Pete Cashmore's blog entry about this (link above) noted: "Alexa’s graphs are copyrighted, so simply hotlinking to them is illegal and using their bandwidth for statsaholic’s benefit. If he wants access to the data to make his own graphs, Alexa provides a web service to do that (at a small cost per query-$.15/1k queries). This pay-as-you-go service is like all of the other web services Amazon provides so I don’t really see how this is going against what Amazon’s platform is all about."
  • When I asked Jeff about Alexa's refusal to offer an API to directly generate the graphs, as Ron has asked them to do, he said "This may be a bad business decision, but it's their decision." (Note that the word "their" refers to the Alexa management team. I was intrigued to note that Jeff always referred to this as an Alexa decision. I don't think that this was just plausible deniability though. I've seen strong signs in a number of my interactions with Amazon over the past few years that Jeff has become a strong delegator, and very reluctant to over-ride the decisions of his managers.)

In short, I ended up with a lot more sympathy for Amazon's position than I expected to. If someone decided that they could make a nice business by re-using the content from O'Reilly books in a new, more accessible form, I'd probably be miffed too. I'd be dumb not to jump on the opportunity if I saw signs of uptake, but it would be my call as the copyright holder. And if they used my name in building their service to boot, and refused to stop when I asked them nicely and even offered them money, I might eventually have gotten pretty angry too.

It appears to me that both Alexa and Ron Hornbaker have been acting somewhat foolishly here. Alexa should have taken an early lead to bring Ron into their fold, as Google did with Paul Rademacher and other Google Maps hackers. They should have also learned more quickly from the user-interface innovations that he pioneered, rather than waiting so long to adopt them. And they should definitely be looking at how to use web services to syndicate their graphs more widely. With companies like hitwise and quantcast using traffic graphs as a key marketing tool, better syndication is a key competitive tool.

Meanwhile, Ron should have been more flexible and respectful. It is, after all, Alexa's data that he's using -- and only that data. He's trading off their name and the service that they built. The value that he added was in user interface and usability (the ability to compare multiple sites, rather than the simple pairwise comparison that Alexa itself provided.) When they made him a nice offer, he should have taken it. And he should have worked harder to build out differentiated value in the site, so it isn't just direct competition for Alexa's own offerings.

I don't imagine that Amazon will now be willing to roll back the clock and repeat their offer, but I'm hopeful that they and Ron will at least be able to come to some kind of settlement. Ron can't afford to go to court, and Amazon can't afford to suffer the kind of PR black eye that this will give them, whether undeserved or not.

tags: web 2.0  | comments: 39   | Sphere It

Previous  |  Next

0 TrackBacks

TrackBack URL for this entry:

Comments: 39

  Krish [04.20.07 02:45 PM]

Your explanation about Ron feeding off Alexa's data puzzles me. I may be missing something here but a question arises in my mind after reading the above explanation. You say Statsaholic is wrong because it is feeding off Alexa's data. Isn't Alexa taking the data of every other website in the world to build their business. If Alexa can do it, why not Statsaholic? Alexa didn't ask my permission before acquiring data about my website. If I have to take a stance like Amazon, I would consider Alexa's action as a clear violation of my privacy. I could even consider Alexa to be a stalker. When they have their business model based on the data of other people's websites, I find the Amazon's lawsuit a case of greed. Can someone explain to me if I am missing something here?

  Tim O'Reilly [04.20.07 03:20 PM]

Krish --

Alexa is taking data from people who've downloaded their toolbar (plus other sources), and who, in return for using that toolbar, have agreed for their aggregate surfing data to be used to drive the service. So that isn't taking people's data without their agreement.

If Alexa were taking your data, you might be right. (And in fact, there may be a lot of people who are contributing to some of the other services, like hitwise, without knowing it, since they get their data by agreement with ISPs, and the agreement with the user is probably buried deep (and unread) in the ISP's terms of service.

So the right analogy to think about for your data is this: you've got your krishwords blog, and let's assume for a minute that it was really popular, but had a bad interface. Someone decided that they could create a better interface, and recreated your blog postings every day. How would you feel? Or if someone decided to translate it into Chinese without your permission? In the latter case, you might well grant that permission (just as I've allowed many sites to do translations of my What is Web 2.0? article) in exchange for the additional visibility, but it would be your decision to allow it.

Alexa has invested significant amounts of data in building their services and database. They have a number of authorized ways to re-use that data. Ron is using the data in an unauthorized way. Amazon/Alexa should be the one to decide how they want to redistribute the data they've gathered.

Now, I'm not saying that Amazon/Alexa is right in their decision not to allow Ron's desired access -- but that's a business decision. It's "wrong" because I think that alexaholic has given Alexa additional visibility and prominence. It's a fan site, not a deliberate competitor.

But in the end, it has become a competitor, of sorts, in that people may choose to go to the statsaholic service rather than directly to alexa. But I don't really understand how that harms alexa -- perhaps because I don't understand their business model. It would seem that more visibility would be a good thing. After all, if I were to link here to an alexa graph, Alexa wouldn't mind at all. But Ron is providing them in bulk, with a clone of the services that Alexa itself provides.

(It's not exactly a clone, but close enough that I, for example, had completely switched from visiting alexa's own site in favor of statsaholic.)

Given that Ron turned down Alexa's financial offer, and instead tried to run a net PR campaign to to make them change their minds, it seems that Alexa got mad and decided to sue to stop him instead.

That's bad for Amazon, since it feeds the net PR campaign and paints them as the bad guy, but in the end, I bet most of the people railing against them would feel the same way, if in the same situation.

Much as I like Ron, and much as I like the service he built, and much as I wish that it weren't ending up this way, I have to say (reluctantly) that Amazon is in the right on this one, morally if not pragmatically.

As I suggested with the title of this post (shamelessly copying Alan's), this is a situation where "everyone loses."

  mrshl [04.20.07 04:19 PM]

Tim, I wonder what you think about services like Twitterverse and Twittervision. Or other products/blogs that reference a popular product. I can see a lot of common sense in Alexa's data claims, but I'm wary of their cybersquatting action. That kind of suit, if successful, would allow "hub" services to wait for smaller, innovative developers to gain traction, then chase the dependent service from the domain name as soon as the hub is able to add the features in-house.

This gives hubs a lot of power and discretion. Encouraging development when it suits them, then later absorbing the smaller services without actually paying for either the features or the allegedly infringing domain name.

I realize that Alexa offered money for the domain name in this service, and Statsaholic was pretty dumb to turn it down, since they're likely to lose it in any settlement. But i hope this case doesn't become bad precedent.

  Jose del Moral [04.20.07 05:03 PM]

I also think that Alexa's case is not the same as Google Maps'. Amazon has always thought that Alexa's future lies in providing business services. It also was Google's will, but they have somehow changed their mind by creating the MyMaps service.

  Krish [04.20.07 05:23 PM]

Thanks for the explanation Tim. Under the circumstance that a website owner has an agreement with Alexa, I have to agree that Ron is wrong in a way. However, I have seen Alexa tracking some of my sites for which I don't have an explicit agreement with Alexa. Thatz why I felt that Ron's actions are no different from Alexa's.

However, I do agree that Alexa's case is different from Google Maps because Amazon is not offering an API.

Thanks for the response.

  Tim O'Reilly [04.20.07 05:35 PM]

mrshl -- I agree that this sets a bad precedent -- but it's one that's always been waiting in the background for anyone who surfs on a bigger company's bow wave.

In the case of twitter and their admirers, I'd imagine that the benefits to twitter in terms of visibility outweigh any competition, whereas statsaholic actually provides the core service of alexa itself. If someone could use twittervision *instead* of twitter, that would be a closer parallel.

  Mark [04.20.07 05:47 PM]

Do I infer correctly that you would have sided with Odeon on this issue from 2004?

If not, why is that case different?

  TJ Mahony [04.20.07 05:50 PM]

FYI: is working with Statsaholic to incorporate Compete's data. Compete launched six months ago and released a free API and is working with several properties to leverage its audience measurement data for free.

There is more and better data than Alexa out there. We are working hard to let people know where it exists.


  Alan Graham [04.20.07 06:16 PM]


I'm going to have to sue you for taking the title to my blog post.


Actually...I am happy to share it since it is how I feel about this whole situation.

When I intimated that you surprised Jeff with the Alexa question, I think that's just how it came across to those of us in the audience...but I hope that it didn't come across to you that I was saying you were playing gotcha journalism...never got that vibe at all. It just seemed to come out of left field.

There are always two sides to every story and I tried to contact both Alexa and Amazon before writing my piece because I wanted to be fair...but I ran with what I had and I'm happy to give equal time.

But more important than that, what I'd like to see...and was my intent with the to now have calmer heads prevail and everyone work this out in the spirit of the community. As I said, I think Alexa has a good case...but in the court of public opinion...not so much. I think it looked really bad that they waited so long before they took issue with Ron...and that was after they made improvements that were inspired by his work.

Regardless...what I'd like to see happen here is for the lawsuit to go away...both parties find a way to move on...and I'd like to see Alexa come out of this with a new outlook and level of cooperation. I don't feel that Alexa is evil...I just think there was likely a better way to resolve this.

I think what this affords us is a good opportunity to break new ground and a path for resolving these issues in the future...we live in a world of mashups...and this certainly isn't the first we'll see of this. So both parties here have a chance to teach future parties how to resolve their differences better.

Thanks for the praise Tim...and nice piece.


Alan Graham

  Rogers Cadenhead [04.20.07 06:19 PM]

I'm a person who didn't find much use for Alexa until Alexaholic came along. Alexa's interface was too clunky.

When you read the history of Alexaholic/Statsaholic, Alexa's legal intimidation of Ron Hornbaker makes no sense at all. He's been evangelizing Alexa harder than anybody and working hand in glove with Alexa engineers. He gave up the name Alexaholic.

How in the world can Alexa not come up with a better solution than to sue him? The price to the company in goodwill will far outweigh any benefit they may derive by driving his venture off the web.

Alexaholic became popular because it was a better interface to Alexa data. They should be hiring from the guy or learning from his ideas, not giving their lawyers something to do.

  mrshl [04.20.07 06:40 PM]

"I think it looked really bad that they waited so long before they took issue with Ron...and that was after they made improvements that were inspired by his work."

I think this is why so many have objected. People are having a common sense reaction that's closely related to legal concepts such as adverse possession and estoppel. Essentially, there's no confusion or bitterness in the community at large when you aggressively defend your intellectual property in the first instance. But, sometimes its in your interest to encourage symbiotic effects even if a service is using your data or referencing your trademark. That's because it brings your service additional notoriety and perhaps enhances your own network effects. But if you wait passively and allow the weeds to grow, or even encourage their growth, it doesn't look good if you call in the lawyers only once you're able to duplicate the efforts of the innovative upstart.

Obviously, API developers face this risk all the time, and there I can't say I've got much sympathy. But the cybersquatting claims are dodgy for two reasons. First, your claim loses some of its equitable force once you've tolerated or encouraged the infringing use. Second, I'm not sure services like Alexaholic or Twitterverse are violating the Anti-cybersquatting law at all. Unlike sites that prey on mispellings or other kinds of consumer confusion, the names of these services clearly indicate that they are satellites of the original that add value and offer viable products of their own. While it's true Alexa eventually competed directly with Alexaholic, that wasn't initially the case. And, just as Alexa can envelop Alexaholic's services, so can Alexaholic broaden the original scope of their services.

In other words, services like Alexaholic aren't squatting at all. They are a dynamic entity capable of creating their own value in the long run. Their developing partnership with Compete seems to confirm this. So they're not squatting at all. This makes the "squatting" claim not nearly as viable as the data claims, in my humble, admittedly non-expert opinion.

  Tim O'Reilly [04.20.07 07:35 PM]

TJ -- you point to a good outcome for Ron (maybe not so good for Amazon). Ron gives up the alexaholic domain -- let Amazon have it -- and stops using their graphs, substituting someone else's, who wants the additional visibility.

Would Amazon drop their case if Ron did these things? I don't know, but I don't see why not.

  Nik Cubrilovic [04.20.07 07:45 PM]

Tim, just a point on the Alexaholic side that I think you are missing (while what you say is valid), and that is that nobody I know, nor nobody in this community of ours (bloggers, entrepreneurs) used Alexa until alexaholic came around.

While this might be hard to prove, I think we all know it is true. Techcrunch and other blogs never linked to Alexa but they did link to Alexaholic once it was released.

I think this nullifies the statement that Alexaholic traded off of the Alexa name - if anything Alexa would have had a click-through boost from the Alexaholic users. Can Amazon publish stats on how many visitors were referred back to Alexa from Alexaholic?

Ron might not have acted properly, but you have to remember that he isn't a large corporation - just a developer who built a site, at his own expense, for the benefit of all of us. It's not that he should have known better but he doesn't have the benefit of having large PR and legal teams at his disposal to help him out or to understand what you can and can't do. I guess what he thought he was doing was right, and most of us agreed.

Also, Ron was paying Alexa a fee each month to use the API, while the images might not have been part of this data licensing deal they still did carry Alexa branding and Ron didn't try to hide the source of the data. How this harms Alexa to the point where they must take out a lawsuit to seek 'damages' I will never know.

My take is that Alexaholic or Statsaholic were mutually beneficial to both Alexa and users. An argument can be made about the use of Alexa in the name but as Statsaholic the service is a good thing for all.

  Ron Hornbaker [04.20.07 07:52 PM]


Thank you for the even-handed commentary here. You're certainly right about both me and Amazon making mistakes in this ordeal, and I do wish I could turn back the clock on this one.

While their acquisition offer seemed good at first (about 2.5X earnings - I was making a bit from advertising on the site), I felt pressured to sign quicker than I liked, had some partnership issues to take care of on my end, took some time to clear those issues up and clear my head, and then that's when things started to go downhill. It's been kinda crazy since then, and I didn't want things to get this public and personal - but when you're an individual, and get backed into a corner and sued by a large corp, you do and say things just to try to survive.

Some have suggested that I encouraged you to confront Jeff Bezos at the conference, but you and I know that wasn't the case -- you called me Monday morning out of the blue, and wanted my side of the story in case it came up with Jeff. Coincidentally, Amazon and I were supposed to have a sit down to attempt a settlement on this on Monday, the same day as the conference, but something came up on their end and they cancelled.

I still wish Amazon and Alexa no ill will over this, and have been trying for almost a month now to hand over the domain they want and start paying them for the use of their graphs. I'm not "holding out" for a better offer from them or anything of the sort - I'd like to continue developing Statsaholic with more features, bringing in Compete and Quantcast data, but I was delaying doing that because I thought it would make Amazon even angrier to see their competition alongside.

Now my chances look bleak for this to have anything close to a happy ending. Amazon is understandably miffed at the extreme negative press this caught this week, and seems intent to carry through with the suit _and_ take both and from me. I'm very close to throwing in the towel on this one.

For what it's worth, I'm sorry for the trouble I've caused by resisting in such a public way. My wife and I have a baby boy (our first) on the way, due in just a few weeks, and my energy and thoughts are turning increasingly towards them. I might just need to give in to this and focus on what's really important in life for the next few months, instead of getting dragged even further down by this.

  Ron Hornbaker [04.20.07 09:07 PM]


That's a bug. Since Alexa doesn't publish traffic data on their own domain, you're seeing a shift of the graph lines - the two lines you see on that graph are for and

Checking Quantcast and Compete for traffic indicates Alexa gets more than Alexaholic, but during the last quarter of '06, Alexaholic was on a sharp trend up and was probably in contention.

  Tim O'Reilly [04.20.07 09:34 PM]

TJ -- you point to a good outcome for Ron (maybe not so good for Amazon). Ron gives up the alexaholic domain -- let Amazon have it -- and stops using their graphs, substituting someone else's, who wants the additional visibility.

Would Amazon drop their case if Ron did these things? I don't know, but I don't see why not.

  Tim O'Reilly [04.20.07 09:49 PM]

Mark -- I don't think I'm "siding" with Amazon here. I'm pointing out that it's not just Amazon being the big bad bully here, and that they do have their own justifications for why they are suing Ron. As I said on stage and repeated here, I think that's a dumb move for all the PR reasons -- Ron built a fan site, and as a number of people above have noted, it helped to build the popularity of Alexa, not tear it down. Ron has always credited Alexa, and has meant them well. However, Amazon has its issues, and when they weren't able to come to terms, made its business decision.

I'm sad if the bad press they got has hardened their position and made them angrier at Ron, as he suggests above. I would hope that they would come to an accomodation of some kind.

But specifically to the Odeon case: I'd make the same judgment as I make with Amazon here. They are within their rights and probably have their reasons, even though it puts them in a really bad light from a PR point of view, and thus it was probably a very dumb move. That being said, from the Odeon article you pointed to, the guy building the take-off site responded to the cease-and-desist, and didn't push back, as Ron did.

It's a tough call. I really wish it hadn't come to this. Ron sounds like he's bending over backwards to make nice with Amazon, and it's a real shame if they want to hold a grudge.

And I'll confirm that Ron didn't ask me to bring this up with Bezos on stage. It was an issue that I'd wanted to bring up myself, and I called Ron to verify facts before the meeting, just as I talked to Jeff in advance. (Ron did contact me to reach out to Jeff months ago, when the problem first arose, and I did so at the time, expressing my admiration for alexaholic, and asking Jeff to persuade the Alexa folks to look favorably on his site. But that obviously went nowhere.)

  Tim O'Reilly [04.20.07 09:53 PM]

mrshl -- I agree that the fact that Alexa seemed supportive for Alexaholic for a long time is the most troubling part about all this. If they'd let Ron know right away that they disapproved, that would have made their case much stronger. But the fact that they made approving noises on their blog, and generally seemed supportive, only to later change their minds, seems really unfair.

And Nic, I can't agree more with your point that alexaholic really put alexa on the map for a lot of us. In the long run, I believe Amazon has a lot to lose by shutting Ron's site down. I'd be very surprised if prominent bloggers continue to use Alexa data. I know that I immediately reached out to Hitwise as soon as Alexa started pressuring Ron.

  Cory Doctorow [04.21.07 05:13 AM]

I wonder about the graphs being copyrighted. IANAL, but how can this be so? After all, the graphs aren't creative works -- they're programatically derived representations of factual data. By definition, that seems to be an un-copyrightable subject matter. Only creative labor attracts a copyright. Phone books don't.

  adamsj [04.21.07 07:41 AM]

I found mrshl's comments about adverse possession quite interesting, but then I'm not a lawyer--and if I were a lawyer, I suspect I'd be working on putting them to the test.

  Rita Hornbaker [04.21.07 08:44 AM]

Is it surprising that Ron Hornbaker is a convicted extortionist?

  h [04.21.07 08:54 AM]

ur-cabal evolution?

'Small business' is still taking advantage of 2.0, which is OK, but to say that centi-billion dollar companies are 'revolutionary' or '2.0' is a joke; 2.0 is exposure to lawsuits from the tech-industry-pension-fund corps.

When 2.0ians got job offers from MSLive/Adobe-consolidated/Google-AOL/etc it became Marketing2k3./troll

  Teddy [04.21.07 10:08 AM]

Ron-- don't give up. You have created something cool, try to find a way to keep it going.

If you've got momentum with Compete and Quantcast, go with that and worry about alexa next. Figure out a way to end this, give up the alexaholic domain and settle with them so the service can stay up.

  Tim O'Reilly [04.21.07 10:18 AM]

Cory -- First off, this is data collected by Alexa via their toolbar, not public data, and that might be a factor. But secondly, even in the case of compilation copyright (e.g. phone books), you can't just copy someone else's compilation. You have to recreate it from the same source data. But IANAL, and I'd rather leave all that kind of speculation to the lawyers.

Meanwhile, I continue to think that it would be best for all of us were this to be settled. If Amazon doesn't want Ron using Alexa data, he should stop and give them the alexaholic domain. And he should relaunch statsaholic with other data, leaving alexa out of it. I imagine that in the long run, they will regret the decision to force him to stop using their data.

  Cory Doctorow [04.21.07 02:52 PM]

Hey, Tim -- neither of us are lawyers, but that's not how my understanding of a compilation copyright works. Compilation copyrights aren't attracted by phone-books -- rather, compilation copyrights are attracted by compilations made by creative selection, such as "The 100 greatest songs," or "My 10 favorite phone numbers." A list of all the names of all the people in the city isn't copyrighted, period -- which means that it is freely copyable, without any permission. Data sources are also irrelevant -- facts, no matter how they are discovered, and not copyrightable subject matters. The famous phone-book copyright case was settled for the plaintiff, granting him the right to copy phone books (and not by asking everyone in town what their phone number was).

  Douglas Karr [04.21.07 04:12 PM]

I've read through all of the details on this and I believe there's really only a single point: Alexa made the data available for external resources to use. That they did not have a comprehensive agreement beforehand on how this data should be utilized is their mistake.

Companies are learning that the semantic web is becoming more popular and APIs and Open Systems are garnering business, not hurting it. This is the beginning of the end of Alexa.


  Nik Cubrilovic [04.21.07 05:49 PM]

So if Statsaholic took the same data from Alexa that it has already licensed, and then created its own graphs, there is no issue?

  Tim O'Reilly [04.21.07 08:08 PM]

Nik -- I really don't have the answer to your question. I've never heard a reason from Amazon or Alexa as to why they object, just that they do. You or I can think all we like that statsaholic is good for Alexa, but they don't see it that way.

I'm hoping to talk directly to someone from Alexa to hear just what it is that they are looking for out of this situation.

  Michael R. Bernstein [04.21.07 09:40 PM]

Tim, you said: "Alexaís graphs are copyrighted, so simply hotlinking to them is illegal and using their bandwidth for statsaholicís benefit."

Huh. And yet, copying or hotlinking to Alexa graphs happens *all the time*. I think even you've done it on occasion Tim.

Has Alexa ever made an issue of their copyrights on these graphs to anyone else?

  Michael R. Bernstein [04.21.07 10:50 PM]

Actually, I shouldn't have attributed the quote to you, but to the commenter you were quoting in turn.

However, you were posing this as a possible justification for Alexa's stance without significant challenge, so I'd still like to know the answer to my question: Has Alexa ever made an issue of their copyrights on these graphs to anyone else?

  Tim O'Reilly [04.22.07 06:53 AM]

Michael, there's a big difference between showing an individual graph, or a few of them, in the course of making a point (that's what's called "fair use") and showing ALL of the graphs by building a competing service. Statsaholic is a complete substitute, not a fair use. It's like the difference between quoting from a book and republishing the entire thing in a new format.

Now, because I don't understand Alexa's business model, I don't know why it matters to them whether the graphs are served up by them or statsaholic, but apparently they do. (It would seem to me, as to most of my readers, that Alexa would like the additional exposure, but it is true that once you start using statsaholic, which is so much more usable than alexa's own site, you never go back there. He's thus depriving them of some of the deeper attention they might want.)

  Michael R. Bernstein [04.22.07 09:33 AM]

"It's like the difference between quoting from a book and republishing the entire thing in a new format."

I'm not sure you can stretch the analogy that far for a service. At no time has SA made a complete (or even substantial fraction) copy of all of Alexa's graphs. This is not to say your underlying point is invalid, just that I don't think we *have* a fair-use standard for services at this time.

"He's thus depriving them of some of the deeper attention they might want."

Is he? That assumes Alexa would have had that deeper attention if Alexaholic hadn't existed. This seems to me the same sort of 'every free download is a lost sale' fallacy that the large media companies have been promulgating for years.

  Tim O'Reilly [04.22.07 09:46 AM]

Michael -- Hmmm...

1. Seems to me that providing a means that gives complete access to any and all graphs is similar in function to providing a complete copy, even if the complete copy has not yet been made. But you definitely point to an interesting corner case that copyright law hasn't dealt with yet.

2. Re. your second point, I agree with you. I think alexaholic was and is GOOD for alexa. But Alexa doesn't think so, and are within their rights to ask Ron to take it down. They can be wrong on their business decision and right on their legal position. Just like the music companies: downloading free copies is illegal, and despite the fact that many of us think it makes more business sense to give copies away sometimes (see Piracy is Progressive Taxation), doesn't mean that it shouldn't be the decision of the copyright holder.

In my own investigations of the merit of free downloads of O'Reilly books, we see that sometimes it helps and sometimes it hurts -- there are a lot of complex circumstances that make us decide to go one way or the other. But ultimately, it's our decision and that of our authors, and while we want to listen to the marketplace telling us what kind of access they want, we don't support people who decide to take stuff over our objections.

Again, I think Alexa is making a profoundly wrong business decision, but I support their right to make it. And I think Ron will be much better off if he finds new partners who want to work with him, rather than getting into what was at first a PR pissing match, and now a legal battle he can't afford.

  Stans [04.22.07 07:54 PM]

Tim, good post. One point though. We also are in a business whose main (really only major) asset is data which we collect. We spend a lot of money and put a lot of effort into it. It's not exactly the sort of thing Alexa is doing, and distinctions could be made, but I think it's pretty close.

It turns out that copyright of data is a pretty dicey area. Fundamentally, facts cannot be copyrighted. What companies do who have databases is, first, claim a "compilation copyright" and a copyright on the edits and massaging of the data, the "enhancements." This is not something that tends to get upheld in court, however. Courts err on the side of putting factual material in the public domain.

Before the 1989 copyright revisions, there was a "sweat of the brow" doctrine that allowed facts that took effort to gather to be protected from copying, although you could gather them from scratch again yourself and use them. The database industry is trying to get this reinstated in the form of the Database and Collections of Information Misappropriation Act, but no luck yet.

Copyright is a great way to protect stuff, since if you register you can get statutory damages, and the DMCA lets you get it off web sites easily. But because of he copyright situation, database people tend to rely more on contract law. Places like Alexa (and us) will have "browsewrap" user agreements, contracts that prohibit copying, that in theory the user "agrees" to by using your site, which will have a Terms of Service link on it. You can see Alexa's browsewrap agreement on their site. Enforcing a contract is a bit less effective than enforcing copyright, and you have to jump the hurdle of whether your browsewrap contract can be enforced in the first place. It's weaker than clickwrap contracts, which are weaker than shrinkwrap contracts, which are weaker than an actual signed contract.

I'd like to see the lawsuit complaint: is it trademark based, or are they going to litigate the data use, and if so, on a copyright or contract basis? The pattern for companies with somewhat weak and flakey claims is to avoid litigation. A good example is Adobe, which has avoided litigation on its claim that it's typefaces are protected under copyright. Type designs are exempt from copyright, but Adobe claims that they are not type designs, but rather "copyrighted PostScript computer programs." They have not seen fit to defend this theory in court, however.

Alexa's data was crawled and is also being used in a wholesale manner by the startup, which recently got $1 million in funding. I suspect that the legal advisors to AboutUs told them not to worry too much about a copyright or contract lawsuit from Alexa.

  Michael R. Bernstein [04.22.07 08:09 PM]

1. My point wasn't that a complete copy had not yet been made, but that a complete copy will never be made. Thus, you can't make an argument that what was going on was 'republishing the entire thing' if the 'thing' is Alexa as a whole. OTOH, if the 'thing' is an individual graph, then what distinguishes Alexaholic from other folks grabbing the graphs is simply how many counts of copyright violation he should be accused of.

2. I agree that the decision as to whether to allow uses beyond fair use is the copyright holder's. However, any misleading rhetoric surrounding that decision (on either side) aught not be given a pass. Do you actually think that "He's thus depriving them of some of the deeper attention they might want.", or do you just think the folks at Alexa feel that way?

3. Cory, if you're still tracking this conversation, I think the graphs *are* subject to copyright, because the look and feel of the graph is arguably a creative element. That said, the data embodied in the graph (were it to be extracted) is *not* protected.

I think the (somewhat weak) comparable analogy in the phonebook case you cited would be a distinction between making photocopies of the phonebook, versus copying the data (names and phone numbers) out of the phone book.


  Michael R. Bernstein [04.23.07 12:25 PM]

So, looks like they are mostly focusing on the trademark claim. Except that they waited a long time since becoming aware of the violation before they tried doing anything about it, which sorta weakens their position.

Their data 'stealing' and 'theft' rhetoric is also kind of lame and misleading, IMO.

  Michael R. Bernstein [05.02.07 10:38 PM]

Here is a relevant blog post on the subject of the copyrightability of graphs:

Post A Comment:

 (please be patient, comments may take awhile to post)

Type the characters you see in the picture above.