One of the most difficult things about protecting your content on the Web is finding ways to encourage legitimate sharing of content without encouraging less desired use. We almost all want people to Tweet, link to and quote our content but we don’t want to encourage spammers and we always want to make sure that proper attribution is applied.

There have been a slew of services that have worked to make exactly that possible. Ranging from licensing solutions such as Creative Commons to technology-oriented ones such as Emebed Article and Tynt Insight (formerly known as Tracer).

However, there is another service, Clp.ly, which hopes to make a dent in the field.

Clp.ly is both a bookmarklet and a button/JavaScript embedded into websites that make embedding content (or “clipping” content) extremely easy and elegant.

The idea is to make it simple and attractive to correctly cite content from various sites, including both pulling quotes and snapping images. This raises the simple question “How well does it work?” The answer, overall, seems to be that it does the job very well.

The Bookmarklet

The most common way one would likely use Clp.ly is via a bookmarklet installed in the browser. The bookmarklet can be used one of two ways.

First, if you select text on the screen and then click the bookmarklet, you are given the option to embed just that text. The quote can either be shared directly via a link on clp.ly’s server or it can be easily embedded into a blog post or any page with HTML. Here is an example taken from this page.

In April, @BadAstronomer, Phil Plait, moderated a panel

From Geeky Pleasures (share this quote)

All in all, the embed has three elements, the first is the quote itself, which is framed with large quotemarks and formatted to stand out, the second is a direct link to the source page (albeit masked by a short url provided by Clp.ly) and finally a link to share the quote via Twitter, Facebook and other services.

The other method for using the bookmarklet is to simply click the link without selecting text and Clp.ly will instead direct you to a page that will allow you to embed a formatted screenshot of the page via a screen like this:

Basically, the tool lets you select the relevant part of the page by moving the slider up and down the left-hand column and resize the image you want to make using the box in the center. From there you’re given an embed code for your site and a direct URL for the clip.

Here is what a sample image embed looks like, once again using the site above.

Clipped from: Geeky Pleasures (share this clip)

As you can see, it is a fairly powerful clipping bookmarklet and, if you register for an account, you’ll get the added feature of having your clips stored and remembered in your profile. This can make it somewhat useful as a research and bookmarking tool though, admittedly, Diigo is probably a better choice for that function.

But as nice as the bookmarklet is, it is only half of the equation. Clp.ly also offers an easy way to embed the service into your site, making it available for everyone.

Clp.ly for Webmasters

If you register for an account, you can also add your Web sites to the service and that will let you embed Clp.ly’s functionality into your site for all to use, whether or not the user has the bookmarklet installed.

The button works much the same as the bookmarklet, having the same functions on click, but the JavaScript that powers it adds a new feature, copy detection.

When a user selects and copies text on your site, they are greeted with a small popup similar to the one above.

Though the content is copied as is to the clipboard, the user is prompted to either format the content for embedding or take a screen capture if they want. Compare this to Tynt which, when it adds attribution to copied text, does so stealthily and without warning to the user.

This turns the attribution protection into more of a value-add for the person doing the copying as the formatted version is more attractive than plain text.

One feature I could not get to work in my testing is that the system is also supposed to warn visitors when they are attempting to copy too much text. If a user copies more than 100 words, they are supposed to get a fair use warning with a link to the Wikipedia article on the subject (though I would recommend this link instead). Likewise, if they copy a very small amount, they’re given a chance to search for the term.

To those who were uneasy about the way Tynt stealthily added the attribution line to content protected by it, Clp.ly will likely seem like a good compromise, pasting the content to the clipboard as is but offering the user the chance to have it preformatted for inclusion.

Drawbacks and Limitations

To be clear, Clp.ly is not a perfect solution and it does lack features that other services offer. For example, the service can not share images in the way that Embed Anything can nor does it offer the robust tracking of Tynt (though it does monitor “trackbacks” that pass through the service).

Also, during my use of the service I ran into a few bugs. For one, I received none of the emails I should have gotten from Clp.ly and instead received warnings the messages could not be sent. This made password recovery impossible after I lost my temporary password for my first account.

Likewise, some of the links to various clip did not work well for me, taking me instead to my clips when I was logged in.

However, for the most part, Clp.ly is a solid and effective service. It may not have all the features one could want but likely has the features that one needs and strikes a good balance between forcing attribution down a user’s throat and simply hoping for the best.

In short, it is not an ideal solution but for many interested in this type of protection it may be the best available.

Bottom Line

There’s a lot of features I’d like to see out of Clp.ly including more robust stats, direct links to source material (no Clp.ly short url when not needed) and an invisible button (meaning it is only activated on the copying of text) but the service is still very powerful and works very well.

Though I don’t think it’s going to sway large media companies who use Tynt at this time, largely due to the small nature of the company and the lack of invisible tracking, it shows a great deal of potential and may be perfect for small-to-midsize bloggers and webmasters who want to do a little bit more to encourage good copying of their text.

So if you are interested in this kind of functionality, give Clp.ly a look, you may find it to be perfect for you.

If you are curious to see how it works, I’ve enabled it on this article so feel free to play around with it for a bit.

No trackbacks yet

Earlier this month, I reported on a new beta service by Tynt called Tracer. The service works by having webmasters add a line of JavaScript code to their site that allows Tracer to track when users select, copy and otherwise interact with text and images on the site.

Though not actually a copy protection system, it lets writers know which of their articles are being copied the most and also adds an attribution line to the clipboard of the person doing the copying, so works are correctly attributed once they are pasted (this does not work with images).

Though the service has obvious limitations, such as not being able to work with RSS feeds, it has generated some interest for readers of this site and I have a few updates on the service as well as some p broad statistics on how PT is being copied.

Bug Corrections/Issues

First, the bug issue that I reported on that caused page views to not be counted seems to have been fixed. Page views are reporting though the numbers don’t seem to jive very well with my stats elsewhere. However, this could be caused by a lot of issues and may or may not be a problem with Tracer.

The actual tracking of copies and selections seem to be working well. One does have to remember that the default view sets to seven days, meaning older copies and views scroll off the page the table day to day (this caused some confusion).

All in all, the bug fix seems to work and the tracking is functioning as advertised.

No-Attribution JavaScript

Tynt is also in the early stages of testing a version of the JavaScript that does not cause users to have an attribution line copied along with the text, meaning that Tracer simply tracks the content passively and doesn’t interfere. I’ve added the test script to my site to help with their testing and you can play with it if you wish (though it shouldn’t actually be visible to the user).

It’s been a huge help to me as I often copy from my own site to include in emails and other posts, making that feature somewhat annoying.

Some General Discoveries

With nearly a month of Tracer usage under my belt, I’ve made a few observations about what content is being copied on PT that may be applicable to other sites:

1. Pages Get Copied the Most: Static pages seem to be getting the lion’s share of the copying. The top three pages on PT are What is a Copyright, Stock Letters (which is intended specifically for copying) and The Limitations of Copyright. Though some posts did well too, pages clearly saw the most copying, though only a fraction of the most traffic.
2. No Correlation Between Selections, Copies and Page Views: The three statistics almost could not be more unrelated to one another. The How to Find Plagiarism page was the most visited and was third in the number of selections but saw almost no copying. Likewise, an older post about the role of schools in fighting plagiarism saw almost no traffic, but was the most copied post despite very few selections. The end result is that it is impossible to predict the amount of copying based on the amount of traffic.
3. Most Copies Generate No Traffic: Despite several thousand copies, only 20 page views were generated by Tracer. This isn’t wholly unexpected as many of the copies will not be used in Web pages and many will remove the Tracer link in favor of their own, but it seems to favor the idea of removing the attribution requirement and making Tracer a passive observer.

All in all, these are some interesting, if somewhat unexpected, results from Tracer and they make me want to follow the service’s metrics closer but also having me wishing for additional information, such as referral data and more specific information about what was copied.

Bottom Line

Though Tracer has only a limited usability as a content protection tools, Webmasters who are interested in seeing how their content is being used, especially those that work with static pages, should definitely consider trying it.

It definitely has its limitations, but can provide some great insight nonetheless.

Currently there are a lot of great products for tracking where your content appears on the Web. From veteran services like Copyscape to newcomers like FairShare, there are many compelling choices for tracking your content across the Web.

However, all of these services have a common limitation. Though they are great for picking up where your content is being used on other sites, they don’t and can’t measure how people are interacting with your content on your site. In short, they can’t tell you what people are copying and pasting on your home page nor can they help you ensure that your work is attributed.

But a new service called Tracer, brought to you by Tynt, can do exactly that. A simple JavaScript dropped into any Web page, Tracer can provide you with information about what text/images are being copied on your site and, in some cases, can help you track where it winds and even help ensure that attribution is affixed.

Though the system is in an early beta stage, it holds a great deal of promise and is worth looking at for Webmasters interested in understanding how their visitors are interacting with their work.

Not only can this provide valuable content protection information, but also valuable analytics information about popular articles and keywords on your site, in a way that search terms and just can’t manage.

What Tracer Does

Once you’ve registered for an account and been accepted into the beta, you are then given a single line of JavaScript code to place in your site’s theme. It can be put anywhere in your site’s HTML before the closing body tag. Once there, the script begins to work like other statistics packages, such as Google Analytics, but tracking a very different set of data.

(Note: Stats are from only a portion of one day and there seems to be a bug in my account preventing page views from showing.)

Currently, Tracer tracks four sets of data:

• Selections: This detects when a user has selected a portion of the page, such as highlighting a phrase or a sentence.
• Copies: This detects when a user actually copies an element on the page, whether it is a section of text or an image. Also breaks the text copies down between small copies, meaning less than seven words, and large ones.
• Linkbacks: Tracks where the content is pasted on the Web, more on this in a minute.
• Page Views: A standard statistic in that it tracks views of the individual pages of your site. Meant to be more of a “sanity check” and to aid in the calculation of some percentages (IE: Percent people who selected text on a page).

Tracer will let you take a look at the specific phrases that were copied, as well as show you break downs of the most popular keywords both in a list format and in a tag cloud.

The other element of the service that many will be excited about is that, when a copy of a section of text is made, Tracer automatically adds an attribution line. For example, if I copy the first paragraph from my previous column about FairShare entering public beta, it appears like this:

Attributor’s individual-targeted FairShare service, previously covered here, has announced today that the service is now open to the public, no beta keys are required.
Read more: “FairShare Enters Public Beta | PlagiarismToday” – http://www.plagiarismtoday.com/2009/03/04/fairshare-enters-public-beta/#ixzz08tla2ShX

Obviously, this is a raw paste without any formatting on my part. It also allows you to see that the URL has a “#” element added to the end of it, which allows Tracer to monitor incoming links and report those in the “linkback” column of the dashboard.

All in all, if you’re a Webmaster that is interested in seeing how users are interacting with your content, there are a lot of compelling reasons to give Tracer a try.

Potential Uses

When you first log into your Dashboard, you’re greeted with a very glossy animation showing thumbmails of your site floating amid the list if recent copied phrases. Though the animation is attractive and interesting, it is ultimately fairly useless, even though it allows you to click through to a highlighted version of your page showing the passage that was copied.

Instead, the real power of Tracer is in the boring, dry data buried within the body of the dashboard, including the list of top keywords, the tag cloud and the most popular URLs with a table of actions taken.

With this information, there are two different approaches users could take with the service:

1. Analytics: The most powerful use, for most bloggers and other creators, will likely be the analytics element of it. You can use Tracer to track what keywords are the most popular on your site and what pages are getting the highest level of interaction in terms of selection and copying. This can be used to make decisions about future content to create or choosing which pages to promote.
2. Copy Protection: Though Tracer doesn’t do anything to prevent copying, meaning it is not a DRM system, it can tell you what posts are being copied the most heavily. This sets those pages up for greater scrutiny, points to sites that are using your content with links, and tells you what images are the most popular among those who are seeking to copy photos. This can help you identify which content needs the most attention and also help you find some uses you might have missed otherwise.

For most, it seems likely that the analytical aspect of the service will be the most meaningful, especially with the service set up as it is. But those who are interested in seeing how their content is being used, it might be compelling as a supplement to other services.

Either way though, there are some serious limitations and problems that need to be weighed before using the service for either purpose, especially considering that some visitors may be unnerved by the way the service works.

Drawbacks and Concerns

Almost certainly long-time readers of this site have already begun to predict many of the problems with Tracer. By being a JavaScript system, it makes itself vulnerable to a series of limitations and issues that may be a deal-breaker for some webmasters.

Consider the following:

1. No RSS Support: Since Tracer is a JavaScript and most RSS readers do not support the language, it will not and can not work on RSS feeds. For bloggers that likely get much of their readership and interaction through RSS feeds, not to mention face the most serious content scraping issues via RSS, this can be a huge hole in the monitoring.
2. No Detection Without Attribution: Tracer can only detect uses of copied content if the copy remains wholly intact, complete with attribution. If the attribution is removed or modified, Tracer can’t see the link and will only detect the initial copy with no information about where it went.
3. User Discomfort: Though Web surfers have grown accustomed to being monitors by traffic packages such as Google Analytics, Tracer also tracks and modifies their copy/paste function. It does so without any warning that I could perceive (possibly could be just my browser) and many users may be put off by this. If you have any thoughts on this, please leave a comment as I’ll be very interested to hear.
4. Increased Load Times: Though, in my testing, Tracer always seemed to load fast on PT, you are still calling a JavaScript from a remote server. It is easy to see how this could increase the load time of a site, especially if something goes wrong on Tracer’s end.
5. Lack of Options: Though the feature set is still being fleshed out, there are a lot of options that aren’t available at this time, including the ability to switch off the automatic addition of attribution, changing the definition of a “short” copy, etc. At this time, the only “option” is the ability to change your password.

Though some of these issues may likely be addressed in future updates of the serivce, especially the lack of options and, possibly, the load time issue if Tynt decides to create plugins for Tracer, but the issues with RSS tracking and possible user discomfort will likely remain.

The question is going to be how are Webmasters going to use Tracer and will it justify the potential drawbacks?

Conclusions

Tracer, in my opinion, has a great deal of potential. Though it will never be a replacement for other content tracking tools such as FairShare and Copyscape, it may be a good supplement. Especially compelling is its ability to track image copying in addition to text copying and its ability to detect what is going on the site itself, rather than waiting for copies to show up in various search engines.

I have Tracer installed on Plagiarism Today currently. So feel free to play with it as a visitor and see what you think of it from that perspective (It will also help me test the back end). I may not keep it however, I am not very comfortable with the idea of messing with people’s copy/paste function without clear warning and since it can’t be switched off independently, I’ll likely remove it altogether in a few days until that option is added.

In the end, my thought on the service is that it holds a great deal of promise but, at this time, isn’t really mature enough to widely recommend, especially for bloggers. If you operate a non-RSS Web site or don’t see a lot of RSS scraping, such as with a store, photo gallery, etc. it may be great to get some “on the ground” information about what is going on when people visit your site. For RSS-heavy sites, the information will likely be incomplete and the risks may not justify the novelty of it.

For most, this will be far more useful as an analytic service rather than a copy detection/protection one. However, it is easy to see how, with some work, it could become a nice compliment to existing copy detection products.

If nothing else, Tynt has definitely created a unique product with Tracer and it has a lot of reasons for Webmasters to be hopeful. It may not be up to its potential right now, but it is a service well worth watching.