Tynt breaks copy and paste, but only if you let it

In late May, a John Gruber blog post on Daring Fireball titled, Tynt, the Copy/Paste Jerks, finally explained to me why pasting headlines into the spreadsheet of stories I want to talk about on Buzz Out Loud each morning had become a pain in the neck. Tynt makes a utility that lets publishers modify what's put onto a computer's clipboard when the user performs a copy action, breaking usual computer behavior and upending user expectations.

As an example of how Tynt "breaks" copy and paste, go to Wired.com and copy headline text from the site, for example, "Deep-Sea Vent Discovery Sets Hydrothermal Life's New Depth Record," what you'll get when you paste it is instead, this:

It's the extra two lines, a blank line and the "Read More" text, that annoys users like me who are trying to fill out spreadsheets or forms with headlines, and who want the source links elsewhere (off to the side, in my case). I know it sounds like a minor complaint, but as Gruber points out, "It's a bunch of user-hostile SEO BS...Everyone knows how copy and paste works. You select text. You copy. When you paste, what you get is exactly what you selected. The core product of the "copy/paste company" is a service that breaks copy and paste."

I agreed, and I put up a Twitter rant myself: "How to screw up cut-and-paste: http://bit.ly/cC34ok Daring Fireball on Tynt. Bonus: How to disable it."

My retweet of Gruber's post lead to a call from the Tynt marketing team, a meeting, and the eventual realization that the people at Tynt are not jerks, that they haven't broken the Internet, and that, in fact, they're sitting on a killer business model.

Heat map of copied text
Tynt's business model is not to annoy journalists. Nor is it primarily to provide the capability to let publishers change copying and pasting from their sites, although that is a very visible feature that some of Tynt's customers choose to enable. Rather, Tynt is a Web usage analytics company. Its value, which it will eventually (hopefully) make money from, is the data it gives to publishers on what is being copied from their Web sites.

This is the Bit.ly model, by the way. Bit.ly's feature is URL shortening. That's why it's popular among users. It's value is the analytics it gives publishers on how and where their URLs are being linked. That's why publishers like it.

Currently, Tynt is a free service for all. To get its tracking capabilities, publishers install a small javascript include in their site's main story template. The heavy lifting is handled by the Tynt servers. Currently, Tynt is collecting data from 560,000 sites, on 150 billion page loads, and 300 million copy actions a month, Tynt CEO Derek Ball told me.

Most sites that use Tynt don't interfere with copying and pasting content. But the copy analytics are still tracked. And some of the sites that do put links into the clipboard--even when the user hasn't explicitly copied said link--have their installation of Tynt set up so that the link only goes into the clipboard when a larger amount of text has been copied--not, as it is on Wired.com, when just a headline is snared.

Ball told me that Tynt reveals that 70 percent of text that's copied is pasted into e-mail, 25 percent into facebook, and the rest is hard to determine (about 1 percent, he thinks, is Twitter). Tynt knows this because when a Tynt-provided link does get pasted, it has a tracking tag appended to it that can trace the link, if it's ever clicked, back to the app where it is clicked from. (Tech note: Tynt tracking tags are prefaced with the # character, usually used for anchor links in a story, instead of the ?, which is what most sites use for their own internal tracking tags.)

Tynt sounds like a potential privacy nightmare, to be sure. Ball says, "We dump the IP addresses very fast." He doesn't want his company holding on to personal data. "We can't afford to keep it," he says. Whether he was referring to the legal and privacy mess of holding that data or the storage and indexing resources required to use it, I wasn't sure.

The data provided to the publishers looks rich and valuable. Publishers can see not only which stories are popular but what text inside them is most frequently copied. Ball says that some publishers use this information to change headline text on existing stories, to give them an extra boost.

Frankly, I'd love to see this information available to CNET editors. And apparently, Ball tells me, our parent company, CBS, is a user of Tynt. I don't know which of our sites use the service, but clearly we're using it in the quiet mode, gathering analytics only, not interfering with users' copy/paste activities.

Quietly, please
It's the quiet use of Tynt that makes the company so attractive as a business. Ball was clear with me that the feature that modifies copied text in the way that Wired.com uses it is optional. Some publishers like to turn on the automatic inclusion of links in their text so people who see them know where they came from, but many do not, settling to simply collect analytics on what's copied.

In the future, Ball says, Tynt may make money by providing a derivative on the analytics it's now giving out for free. "We might try to grow this to where we sell a data signal that nobody else can see," Ball said. He told me that Tynt can compare what's being copied in specific verticals in commerce. For example, Tynt could send alerts to automotive sites when data across its member sites on, say, shock absorbers, starts to be copied a lot.

I started my interview with Derek Ball disliking Tynt for the pain it was causing me in the Buzz Out Loud spreadsheet. But I left the meeting with a lot of respect for the the business. Now all we have to do is convince Wired, and other sites that mis-use Tynt, to change their settings.