Jeff Croft

I’m a product designer in Seattle, WA. I lead Design at a stealthy startup. I recently worked at Simply Measured, and previously co-founded Lendle.

Some of my past clients include Facebook, Microsoft, Yahoo!, and the University of Washington.

I’ve authored two books on web and interactive design and spoken at dozens of conferences around the world.

But seriously, who gives a shit?

Blog entry // 04.18.2009 // 7:22 PM // 32 Comments

On Twitshirt

On Thursday, my friends at Airbag launched Twitshirt, a simple but awesome service that lets you print the text of a Twitter posting on a tee-shirt. On Friday, the app was gone, replaced by a message stating, “we’re reversing the polarity.”

Why? Because some people — and notably some people who turn out to be very influential in the Twitter community — had a problem with Twitshirt. Apparently, they felt like it was “stealing their intellectual property.”

That seems fair enough at first glance. But, it turns out Airbag anticipated this could be an issue, and had several measures in place to deal with it. First, they provided an opt-out. If you didn’t want your tweets to be available as tee-shirts, they made it painless for you to say so. Second, they attributed the original author of the tweet in the tee-shirt design. There was no confusion here — it was very clear who wrote the message. And finally — here’s the kicker — they actually paid the original author of the tweet royalties!

But still, some people had a problem with it. And again, that’s fair enough. But here’s the thing I don’t understand: some of these same loud, influential people who objected to Twitshirt are huge fans and users of Favrd, a site which showcases great tweets (based on people’s use of the “favorite” feature), and is supported by advertisements. Favrd doesn’t allow you to opt-out if you don’t want your tweets shown (at least not that I can see), and it doesn’t pay you anything for having a tweet featured. It does, however, attribute the original author (as well it should).

Personally, I love both of these projects. I think they’re great, and I think they showcase what kind of awesome things can be done with great content and open APIs to it. But I’m really, really baffled about about someone could be okay with Favrd making money off their tweets, but not okay with Twitshirt doing the same, especially in light of the fact that Twitshirt provides opt-out and pays royalties. Maybe someone can, in the comments, explain to me what I’m missing. Why did Twitshirt cause a shitstorm, but everyone loves Favrd? Maybe there’s a good reason — but I can’t find it.

Comments

  1. 001 // Adam Kolff // 04.18.2009 // 9:43 PM

    Actually, Favrd’s attraction is that it allows you to “opt-in,” and as a result you only appear in the database if you submit your Twitter information. Dean Allen designed it specifically so Favrd could filter out the mess of not-so-interesting tweets that are floating around Twitter.

    I don’t think it makes a massive difference either way, however. Just clarifying.

  2. 002 // Ryan // 04.18.2009 // 10:01 PM

    Isn’t Favrd’s opt-in (which, like many Twitter 3rd party apps, also requires you to input your password) so your favorites are included in the tally? Don’t people appear on the Favrd list without an “opt-in?”

    It may be you’re already on the list of watched favoriters, but if not, feel free to add yourself to the list by entering your Twitter ID and password below”

  3. 003 // Marten Veldthuis // 04.19.2009 // 12:35 AM

    I suspect the emotional difference between Favr’d and Twitshirt is that the latter produces tangible goods, and you’ll end up with random other people wearing your tweets. Favr’d stays on the web, and links back to your twitter account.

  4. 004 // Justin Lilly // 04.19.2009 // 12:53 AM

    I think another point is that the shirt features a single author. I am buying @jcroft’s latest tweet unless he is A) Aware of the site and B) opts out.

    On Favrd, however, its a bit different in that you’re making money generically. You’re not making money off of @jcroft’s work, you’re getting it based on what’s popular in the community. The service isn’t singling anything out.

    It might also be worth pointing out that its a change in medium. I can easily follow the link back to @jcroft’s other awesome tweets, less so with the shirt (which leads to less of a community feeling with it).

    I could definitely see not wanting to put my tweets on tshirts, assuming I used twitter like the folks of YLNT and other actual writers.

  5. 005 // Martijn van der Ven // 04.19.2009 // 1:26 AM

    Isn’t this the whole Twitter licensing problem?

    It is not actually clear what license should apply to tweets. Twitter could just as easy write down in their terms that every tweet will be public domain. The problem lies in the fact that all they say is that they themselves will not claim ownership without telling who does.

    Misters Clarke and Suda started tweetCC to fill this gap and I think Twitshirt could actually use their service. People who have already licensed their tweets in a way that would allow you to print them on a shirt should automatically be opted in.

    If we can get Twitter services to start using the tweetCC database maybe people will start to look into licensing their tweets.

  6. 006 // Lode // 04.19.2009 // 2:26 AM

    These kind of services, just like commercial e-mail, should always be opt-in. It’s not fair to assume everybody knows about your service as an excuse to use your IP.

    Having to provide your Twitter credentials to opt-out is downright baffling. (Even though Twitter’s oAuth implementation is still in beta is not an excuse at all not to use it in this case. Besides, there are other ways to verify someone’s the owner of a Twitter account without asking them for their password. (eg. creating a bot that sends a verification code via a direct message) Airbag just took the easy, lazy way here.)

    Cameron Kenley Hunt made a pretty nice analogy that should be pretty easy to understand a company like Airbag that’s used to their IP being ripped of.

    Another example would be Moo.com selling posters, cards, stickers, … based on Flickr photos, without regarding the license applied to the photo, requiring the user to pony up their Yahoo! password to opt-out..

    The fact that they took the site down so fast speaks for them, but this was a beginners mistake, and Airbag should have known better. Even though tweets are just 140 characters long, selling a t-shirt based on something someone else created, even though you compensate the original author (a fixed, non-negotionable amount), is not fair use. <a href=”http://mrgan.tumblr.com/post/96959065/not-yours.

    One way Airbag could have launched with a nice back catalogue of tweets would have been to contact some of the “Favrd superstars” and ask them first, or integrate with tweetCC and only use Tweets of people who’ve explicitly licensed their IP for third party use.

  7. 007 // Jeff Croft // 04.19.2009 // 2:34 AM

    Adam and Ryan: Sorry, but you guys are misinformed. Favrd is not opt-in at all. You can opt-in to having your favorites counted, but you can’t opt in (or out) of having your tweets shown.

    Marten and Justin: I think you may be right that a lot of people’s objection here boils down to the difference between a physical and virtual product, but frankly, I think that’s silly.

    Martijn: Agreed.

    These kind of services, just like commercial e-mail, should always be opt-in. It’s not fair to assume everybody knows about your service as an excuse to use your IP.

    Agreed, but that’s not really answering my question. The question is: why are some people okay with Favrd, but not okay with Twitshirt?

    I understand the perspective of those who have a problem with both sites, and those who have a problem with neither site (like me). I don’t understand the perspective of those who have a problem with one site, but not the other. That is what I’m trying to comprehend.

  8. 008 // Chris Wallace // 04.19.2009 // 7:49 AM

    Well, licensing issues aside, this could potentially be messy for Airbag if a tweet that can potentially harm someone’s reputation makes it onto 10,000 shirts without their permission.

    Who gets sued when Ashton Kutcher’s drunken tweets make it onto a shirt? What about when Oprah’s account is hacked and starts tweeting obscene things? What’s to stop someone from putting these on a shirt and proudly displaying them for all to see?

    I’m not against twitshirt being opt-out, but rather playing devil’s advocate a little.

  9. 009 // Mike D. // 04.19.2009 // 10:03 AM

    Marten is correct. I feel the same way.

    I suspect the only way this is going to work (and in fact, what they mean by reversing the polarity) is to set it up in one of two ways, or both:

    1. Once someone requests a t-shirt, twitshirt direct messages you and asks for your permission, probably with an “always allow” option as well. If you agree, t-shirt is sold and Greg’s Scotch fund increases by a few dollars.

    2. They try to pre-market themselves to Twitterers so that as many people opt-in as possible. With enough word-of-mouth and other awareness vehicles, they have a pretty good stable of people after awhile.

  10. 010 // Jeff Croft // 04.19.2009 // 10:34 AM

    Who gets sued when Ashton Kutcher’s drunken tweets make it onto a shirt? What about when Oprah’s account is hacked and starts tweeting obscene things? What’s to stop someone from putting these on a shirt and proudly displaying them for all to see?

    Chris: These are all really great questions, but I’d say they apply equally to Favrd, just as they do to Twitshirt. Don’t they?

    Mike: I do think Marten is probably right that the difference between a tangible product and a virtual one is the stickler here, for many people. What I’m trying to understand is: why?

    I thought we were mostly to the point, especially in our community, that virtual products (products made of pixels or code), had real, true value, just like that of a physical product. But, I guess not.

    As for your number one option, I don’t think that’s possible. because Twitter’s platform doesn’t let you send a direct message to someone unless you follow them. So, @twitshirt couldn’t send you a DM unless you followed them, which you wouldn’t, if you’d not already heard of the service. Here’s the way I think it should work, in light of the controversy (credit to some of these ideas goes to @Trammell, whom I was discussing this with a couple days ago):

    1. I try to buy a shirt. The shirt has a baseline cost of, say, $15.
    2. Twitshirt fires off an @ reply to the Tweet’s author (since they can’t do DM). It says something to the effect of “So and so wants to buy a shirt with your tweet printed on it. Click here to approve or deny the request.”
    3. The Tweet’s author clicks through, and is presented with a page that not only lets them approve or deny the request, but also let’s them set their price. So, you could say your cut was $4, which would make the total for the shirt $19. The page also includes checkboxes like, “Always let people print tee shirts of my tweets,” and “Never let people print tee shirts of my tweets.” It would be extra-sweet if it also let the author send the $4 to charity, instead of the author’s PayPal. Finally, it verifies the author is who they say they are (preferably by using Twitter’s beta OAuth implementation instead of asking for the author’s username and password).
    4. The shirt is either printed, or not. The author (or her charity of choice) is paid.

    I still like the fully-open way they had it originally better, but this is the best alternative I’ve come up with (again, along with Mark Trammell).

  11. 011 // Rob L. // 04.19.2009 // 11 AM

    Getting back to the original question: “why are some people okay with Favrd, but not okay with Twitshirt?” — I suspect that folks had the sense that Favrd, earning money off advertising alone, will be lucky to just recoup their hosting expenses, but Twitshirts has the potential to make some real money. Not sure if that’s actually true or not, but I think that was a big part of the perception that drove the reaction.

  12. 012 // Jeff Croft // 04.19.2009 // 12:18 PM

    Rob: You may be right, but if so, people are making a lot of assumptions about how much money there is to be made, here. I have no idea how much Favrd makes, and I also don’t know how much Airbag would have made — but at $20/shirt, I can’t imagine it was much. They’ve got hosting expenses, too. But, I do think you may be right that the perception that Airbag was going to make real money and Favrd does not probably has something to do with it. Good point.

  13. 013 // Chris Wallace // 04.19.2009 // 4:36 PM

    What stops people from retweeting the tweet they wanted and having it printed and keeping the money for themselves?

    The reason the CafePress model works so well is because the product includes artwork that can’t easily be duplicated. Unless you want a tweet from a celeb or friend (which should be the majority) you could essentially cheat the system and “steal” tweets.

    Maybe there could be a “duplicate” tweet system somewhere in the process that stops you with an error: “It looks like the tweet you selected was a retweet. We will send an authorization request to the original author before your tweet can be printed.”

    Again, just playing devil’s advocate, trying to come up with other situations where the system could fail.

  14. 014 // Cameron Hunt // 04.19.2009 // 9:48 PM

    It’s not a matter of who’s making money. It’s a matter of fair use.

    Fair use is defined in law:

    … fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright.

    Favrd can make an argument (even a bad one) that they use copyrighted material for teaching, scholarship or research. Twitshirt can’t. They are using it to sell a product. Favrd is an online publication which documents favorited tweets in order for review.

    Just like a newspaper can quote a book passage for a review or news article alongside advertisements, Favrd can collect copyrighted material. That same passage from a book, even attributed, cannot be printed on a t-shirt without permission.

  15. 015 // Jeff Croft // 04.19.2009 // 9:53 PM

    Favrd can make an argument (even a bad one) that they use copyrighted material for teaching, scholarship or research. Twitshirt can’t.

    This is an interesting perspective, even if we all know it would be bullshit for Favrd to actually ever use that defense (which they haven’t, and I don’t expect they will). Both of these products were created with the exact same intention: to scratch and itch the creator had (Dean wanted to know who favorited his tweets, Greg wants to be able to put funny stuff on a tee-shirt), and with the hopes of making a buck or two along the way. We all know that, and even if Favrd may be able to use that defense successfully in a court of law, those of us with any sort of bullshit detector would be smelling quite a nasty stench as they did it. :)

    That having been said, from a legal perspective, you may be right, and if you are, that’s the first really valid reason why one site is “okay” and the other isn’t. Thanks, Cameron.

  16. 016 // Ryan // 04.20.2009 // 5:55 AM

    Fair Use only applies to works that meet the requirements of Copyright. So, before you can invoke Fair Use, you should probably research whether or not 140 characters are copyright-able.

    You might be surprised.

  17. 017 // Jim Renaud // 04.20.2009 // 7:47 AM

    Ryan is correct. Most copyright lawyers agree that 140 characters does not meet the “original works of authorship” that is required to be copyrighted.

    Also, opt-in, opt-out issue: What is to stop me from re-tweeting somebody and making a t-shirt out of it (other than it being a dick move)? What is to stop my re-tweet from being featured on Favrd than the original? I think this another reason why 140 characters falls into that non-authorship area.

    Also, I think the medium issue strikes me as weird. Would I rather be featured on Favrd which sees it’s fair share of people or would I rather have somebody commit & spend $20+ on a Twitshirt and give you the best recommendation in person, then take a shot of the shirt, post it on Flickr and then Twitter about it? I’d take the personal shirt any day! Oh yeah, and I get some cash for it.

    Also in this day and age a Twitshirt is virtually clickable. I see an awesome Twitshirt, I follow that guy on my iPhone with Tweetie in 30 seconds.

    Please do not turn this comment into a t-shirt.

  18. 018 // Cameron Hunt // 04.20.2009 // 9:10 AM

    I haven’t seen anything to suggest tweets are anything less than intellectual property. Do you have any source for the lawyers who believe 140-characters to be un-copyrightable?

    That seems very wrong. So, are haikus un-copyrightable too? That’s a dangerous path to follow just because some lawyers have an opinion on the length of copyrightable material, while the law (as I’ve read it) doesn’t have any restriction on length.

    Also, Jeff, Favrd is indirectly making money while Twitshirt is directly selling a product that wouldn’t be sold without a specific tweet. It’s a different thing, Favrd being a website and Twitshirt being a product.

    For the sake of argument: people visit Favrd because it’s useful and interesting. Sometimes I visit just for a good laugh. Twitshirts is a little different, they’re essentially selling a single tweet for money. The t-shirt doesn’t sell itself, the tweet sells it. On Favrd, it’s the community and the collection of favorites that sells it, not one individual tweet or tweets.

    Twitshirt sells a specific shirt because of a specific tweet, Favrd gets ad impressions for the usefulness of the documentation of the community. I do think it’s different, even if the intentions of both creators were to scratch an itch they needed.

  19. 019 // Billy // 04.20.2009 // 9:19 AM

    I’d argue that what Favrd does is closer to aggregation than republishing. Something akin to a specialized feed reader.

    Which I guess leads us to this question: Would you equate what Twitshirt did with what Google Reader does?

    Put another way: If I took this blog post and published it in a book, would you expect me to ask permission first? How about if I’m displaying it as part of a “river of news” in NetNewsWire?

    You could also make the argument that when you sign up for Twitter, you allow your content to be made available via the API, and therefore displayed in various ways on other sites. What Twitshirt did went well beyond that.

  20. 020 // IANAL // 04.20.2009 // 10:34 AM

    I haven’t seen anything to suggest tweets are anything less than intellectual property. Do you have any source for the lawyers who believe 140-characters to be un-copyrightable?

    Possibly. Just as importantly, though: do you have any to suggest that they are?

    The discussion should revolve around whether or not tweets constitute “short phrases,” and from that, whether or not they are covered by the rights accorded individual works of authorship. If they are short phrases, then the whole “tweets == IP” claim is more complicated than some would suggest.

    It should also be stipulated that none of us are lawyers, and that this seems (at best) to be murky waters for those that are.

  21. 021 // Ryan // 04.20.2009 // 12:05 PM

    I haven’t seen anything to suggest tweets are anything less than intellectual property.

    If you’re going to make the argument that they are covered under copyright, the burden is on you to do a little reading and figure it out. Simply stating that they are or should be or that it’s wrong if they aren’t isn’t really productive.

    I’m not arguing one way or the other, I’m just saying it’s not as simple as many make it out to be.

  22. 022 // Ryan // 04.20.2009 // 12:31 PM

    Gah, I should be more careful with conveying tone in comments.

    Cameron, that was not a frontal attack on you, but on the general commentary about copyright-ability of twitter messages. Basically, people shouldn’t always assume something without looking it up first.

  23. 023 // Jeff Croft // 04.20.2009 // 12:38 PM

    Put another way: If I took this blog post and published it in a book, would you expect me to ask permission first? How about if I’m displaying it as part of a “river of news” in NetNewsWire?

    There’s a world of difference between a 140-character answer to the question “What are you doing right now?” and a thoughtfully written blog post. You know it, I know it, everyone knows it. So, I’m sorry, but this again sets off my bullshit detector.

    That having been said, there are some people who write creative, often funny, thoughtful, original things on Twitter. I love to read them. And frankly, they’re probably the most likely to be put on a shirt, too. And, I can sort of see their perspective about not wanting their creative writing used without permission.

    With all due respect to the writers of those 140 character masterpieces, I can’t help but ask: is a social networking site with a publicly-accessible API really the best place to be hosting your life’s work? Seriously. If you really believe your witty little remarks hold so much value, then you owe it to yourself to control when and where they show up, not to mention how much you get paid for it.

    If I had one of these people in a room by himself, I would offer him the following friendly advice: Dude, if this really annoys you so much, make your Twitter feed private. Or, publish your little wits on your own personal blogs, where you have more control.

    Should Twitter protect creative writers on their service from having their precious writing used outside of the Twitter context? Maybe. I don’t know. But they don’t. We all know it. So, if you consider your tweets little gems of gold, do yourself a favor and don’t make them public. Be smart, people.

  24. 024 // Ben // 04.24.2009 // 12:43 AM

    I have to disagree Jeff. You say

    I thought we were mostly to the point, especially in our community, that virtual products (products made of pixels or code), had real, true value, just like that of a physical product. But, I guess not.

    I don’t think you’ll find many people in our community who would disagree with you there. Which makes this seem like a bit of a contradiction…

    There’s a world of difference between a 140-character answer to the question “What are you doing right now?” and a thoughtfully written blog post. You know it, I know it, everyone knows it. So, I’m sorry, but this again sets off my bullshit detector.

    Is a straight-to-DVD movie that took 100 people a year to create worth more as Intellectual Property than a hit music recording that took one person a week to write and record, simply because more effort went into its creation? Of course not, we all know that’s not how IP is valued.

    Do I get to take your blog post, or Flickr photo, or deviantart illustration, or even your PHD if you so choose to put it online, and put it in a book/tshirt/whatever and make money off that without your permission? I’d hope not.

    Does a band who puts a song up for free download lose any claim to copyright simply because they chose to distribute that content on the web for free?

    Twitter is simply another way to distribute content. Saying “you put your content up somewhere other people can easily access it for free, bad luck if it gets used in a way you don’t like without your permission” seems contrary to the idea that we can claim ownership of IP.

    I don’t actually mind the idea of Twitshirt (if it’s run as an opt-in service rather than opt-out-if-you-happen-to-know-the-site-exists), and I’m also unconvinced that a 140 character message is copyrightable IP, but the problem here is we don’t know either way.

    Anyway back to the “why Favrd but not Twitshirt” argument, I think this from the Twitter Terms provides as good a reason as any:

    The Twitter service makes it possible to post images and text hosted on Twitter to outside websites. This use is accepted (and even encouraged!). However, pages on other websites which display data hosted on Twitter.com must provide a link back to Twitter.

    Favrd does this, a tshirt doesn’t.

  25. 025 // Billee D. // 04.25.2009 // 5:13 PM

    I honestly cannot see the damage that could be caused by allowing folks to wear a t-shirt that sported one of your tweets. That just doesn’t make sense to me, because personally I would be flattered.

    That truly sucks. Jeff. I just bookmarked the site the other day. I was excited and was looking forward to emblazoning myself with a few tweets of wisdom like @zeldman’s tweet about spec work. What a way to ruin the fun, eh?

    Thanks for the post.

  26. 026 // Ben Rowe // 04.29.2009 // 5:16 AM

    Last weekend, prior to hearing about Twitshirt, I had a similar idea: To release a new T-Shirt each week, based on Tweets that were submitted. My favourite submission would be released on a T-shirt for $20.

    I acted quickly, registered a domain name (www.tweetmyshirt.com) and opened up a spreadshirt store.

    I wasn’t until that next day that I came accross Twitshirt, and this post. To be honest, I really didn’t give much thought to this whole issue.

    Now I’m thinking, would this be an issue if they (or, in this case, I) just ask permission before printing, and split the profits?

    What would you think about that?

  27. 027 // Erik Vorhes // 05.07.2009 // 8:11 AM

    Hate to be a dick about it, Ben from comment 24, but a T-Shirt isn’t covered in the Twitter terms of use you cite.

    But even if it were, a Twitshirt provides a reference (or more!) back to the original tweet, which—wait for it—resides on Twitter.

  28. 028 // Dirk Stoop // 05.19.2009 // 12:45 PM
    1. Favrd is syndication, twitshirt is derivation.

      I make software products and sell them online. Generally, I am delighted to see excerpts of our marketing copy (and screenshots, or any other content from our product websites) in spots that syndicate worthwhile content—online, dead-tree, any format. As long as they credit us adequately.

      On the other hand, if I’d run into a cafepress shop offering mugs, mouse pads, etc. with our artwork, product names, payoffs, or anything else taken from our product websites on them, I’d figure out how to make it stop asap. Royalties or no royalties, credits or no credits.

    2. Content has a value, regardless.

      I like twitshirt and think it’s a neat idea. That’s probably because I’m in the business of software, not the business of writing.

  29. 029 // Ryan // 05.25.2009 // 12:26 AM

    What’s next? Your walking down the street and you happen to say something quote worthy.. and then a month later you see it on a T-Shirt? What the hell are you going to do? Sue? Yeah, probably.. lmao.

    Anyways, I like the idea. Opt-in service can be approved upon greatly as well to make everyone happy (see 11).

    & I don’t know about anyone else, but I can definitely see JCroft making a couple of $20’s off of this one:

    I think I accidentially “sexted” last night. -@jcroft

  30. 030 // FemaleEDPillz // 06.25.2009 // 5:30 AM

    Hello ;) Thank you for this website! Here is mine http://motrin.wikidot.com/

  31. 031 // Sweety // 11.04.2009 // 12:34 AM

    I find it useless, but my girlfriend often Twitters. At her Birthday, i give her a T-Shirt with her Picture on the front and the Twitter Logo on the back. She wears it often^^

  32. 032 // Marry Horses // 01.27.2010 // 7:18 AM

    Isn’t Favrd’s opt-in (which, like many Twitter 3rd party apps, also requires you to input your password) so your favorites are included in the tally? Don’t people appear on the Favrd list without an “opt-in?”

Tags for this entry
Links in this entry