Whenever we tell someone new about Ghostery, the odds-on favorite for the the first question they ask is: “so how do you guys make money?” Web users have learned to be skeptical when it comes to an allegedly free product.
We think that being skeptical is smart, and Ghostery is made for exactly the type of user who would ask this question with one raised eyebrow and one foot out of the door.
The data donation feature in Ghostery is called “GhostRank,” which is opt-in and anonymous. We know how important it is for us to make clear exactly how what data is shared and how we make use of GhostRank data.
Ghostery is free to use. It’s supported only by optional donations of anonymous data from users who have opted into GhostRank.
Ghostery never shares any data about our users with advertising companies or anybody else. GhostRank data does not include any identifiable information about our users – period.
Ghostery is owned by Evidon, a for-profit company. Evidon sells GhostRank data to businesses to help them market to consumers more transparently, better manage their web properties, and comply with privacy standards around the world.
Now that we have that out there, here’s more detail on those points above:
Ghostery is free to use.
We don’t collect any data whatsoever unless you specifically check a box in Ghostery’s options saying it’s okay. Evidon and Ghostery use data on trackers, not people. If you don’t opt in, the product isn’t affected in any way. We have no plans or intentions to make any of Ghostery’s features dependent on GhostRank participation.
Ghostery does not share any data about our users.
GhostRank data does not include things like “John Doe is a Midwestern car enthusiast.” It includes things like “Ad Network X’s tracking code was encountered 50 times on this car site.” We hash IP addresses in our logs (we only keep those scrambled strings to help us do a basic user count), and we strip out query strings in the urls we collect in an effort to make sure there isn’t any user-specific information lurking there, either. Put simply, none of us could use GhostRank data to identify a user out on the web even if we were inclined to do so.
Ghostery is owned by Evidon, a for-profit company.
GhostRank data helps us to create solutions that help businesses. It enables site owners to detect all of the tracking technologies on their sites and maintain control of their data, and provides competitive intelligence for companies across the web.
Evidon and Ghostery have transparency at the core of our collective business model – it’s not a “main goal” or a “central value,” but the actual core of our company. All of our products increase visibility into the online advertising world, and those products are designed to meet the most rigorous standards of operational transparency
Skepticism is healthy, and beyond that, all of us at Ghostery applaud that approach. But we hope that users don’t let skepticism turn into fear, uncertainty, and doubt about a product that can help you, in a safe and uncompromising way, be a better skeptic. We hope you’ll opt into our data donation to help support the product, but we understand that requires a level of trust about what we do with the data. To that end, if you have any questions, please find us on our support forums, Twitter, Facebook, and/or via email. You can also email me personally at email@example.com.
APPENDIX: GhostRank Logging Breakdown
Several users have asked for specific information about what we collect. What follows is a breakdown of a typical GhostRank log URL gathered from my own browser this morning while I was reading an extremely interesting article about Khloe Kardashian’s workout habits. Remember, this only matters if you’ve opted-in. If you haven’t enabled GhostRank, you can skip this, because we collect nothing.
Here’s the raw GhostRank url:
And here’s a breakdown of the elements included (with a bit of cleaned up encoding for the sake of readability):
- this is simply the address of the GhostRank logging api
- “bug id” – the tracking company id (the naming convention is leftover from when we referred to trackers as “web bugs”). In this case, the tracking company is #951, a video player called “Taboola”)
- “app pattern id” – the id for the actual Taboola element we discovered (many companies have more than one type of technology they deploy, and we make the distinction whenever possible).
- “domain” – the URL for the page I was visited, the aforementioned heady reading about Khole’s workout.
- “source” – the url source for the element we discovered. In this case, it resolves to a thumbnail image of a heavily tattooed person.
- “blocking” – an indication of whether or not I blocked the tracking element. During this exercise this morning, I was not blocking.
- “blocking mode” – an indication of Ghostery’s blocking settings, to give us a better idea of what the user was trying to do (and to help us make sure Ghostery is blocking correctly)
- “bug selected” – a flag to tell us if the tracking element was selected for blocking. A combined look a this and the other two blocking indicators can show us how frequently users are making blocking exceptions.
- “bug version” – the version of our library of tracking elements. We’re adding to this all the time, and it helps us to be able to compare some data only to other data collected in the same version of the library.
- “latency” – the time (in milliseconds) it took the detected element to load in the page (if it wasn’t blocked).
- “above fold” – an indication of whether the element was visible when the page loaded, or if the user had to scroll to see it.
- “version” – Ghostery’s version number.
- “cache version” – we recently made some changes to how frequently Ghostery counts similar trackers, and this number indicates which version of that behavior was at work when the data was collected.
- “user agent” – the browser in which Ghostery is working. I was using Firefox.
- “random” – a random number to help us properly collect unique GhostRank records.
And that’s all. Nothing about me in there at all, really – and there would be no way for anybody at Ghostery or Evidon to know this particular GhostRank report belonged to me, except now I’ve been very explicit about my interest in Khloe Kardashian’s workout on our blog, which will probably make its way around the office pretty quickly.
There’s also a separate report for a complete web page on which trackers are present. It’s pretty similar:
“d” is domain, “l” is latency in milliseconds, and “ua” is the browser, just like above. “s” stands for “spots”, which counts the number of ad spots on the page.