Note: the following documentation applies to content created in the old ThingLink image editor. If you'd like access to API for content from our new editor, please contact us at email@example.com or submit a request for Technical Support with a description of your use case.
Sometimes it is necessary to tie your Thinglink performance to your existing tracking tools. We offer three basic integration mechanisms:
- Tracking pixels
- Click tracking
These options are only available to our corporate account holders.
While in the editor, select settings beneath the image and choose the advanced tab. Find "Tracking Pixel" and place your ad-network provided tracking pixel URL there. The URL should be different for each image.
Most tracking URLs employ a "cache buster" to ensure that the image is loaded every single time from the network. You can accomplish the same thing by placing the word "RANDOM" (in all caps) somewhere in the URL.
Note that if the tracking pixel link is invalid, we may reject it and not track it at all.
IMPORTANT: We validate the tracking pixel by fetching it once. This means that this view will show up as a view on the dashboard of the system you got the tracking pixel from, even if there are no actual views of the image yet. If it does not, there may be something wrong with your integration. Please always check that the views are visible on your dashboard! The requirements for the pixel are:
- It must be online and reachable by our servers
- It must be an exactly 1x1 size image
- A cache buster text must exist somewhere in the URL (e.g. either "[timestamp]" or "RANDOM" - without the quotes obviously).
Simply place the click tracking URL you have been given as the "link" in the tag editor. We will automatically dereference that link and figure out which rich media tag should be used.
The tracking URL is hidden from viewers for our Premium customers.
Thinglink's power comes from strong user engagement and the fact that the user does not have to leave the site to view the content. Therefore, traditional click tracking is usually not sufficient to truly appreciate the performance of Thinglink. If you want to track the hover and click events yourself in a more complicated manner than just tracking pixels and links allow, you may want to check out the Embed Code API for callbacks you can use on your own site. Some coding may be required.
Questions and answers
The tracking pixel counts differ from the ones on your site?
This is simply because we measure slightly different things: Our view count is dependent on when our script talks to the server; where as the tracking pixel is displayed only after the response has been received and processed. The browser can break the processing in between these two events, the user can click elsewhere before the tracking pixel has loaded but after the tags have been fetched, the network may be slow, the user may have tracking pixel blockers, etc.
Our impressions don't match your view counts
If you're using Thinglink for Ads, then you should note that we do not track "impressions". A typical impression count denotes how many times the unit was injected in the web pages, whereas our views count how many times the interactivity has been started. If your web page has lots of content to load (especially if you use an iframe embed inside an iframe inside an iframe), Thinglink API calls may execute pretty much as the last thing on the page. We try to not interfere with the page loading times, so we let your assets load first.
If your users jump from page to page fast, it is entirely possible that the user jumps away from the page after the impression has been counted, but before we have had time to send our API call, resulting in the case where the impression count can be bigger than our views count for the same period.
In other words, just showing the base image is not a view, the tags need to be loaded as well.
- Do not embed Thinglink iframes inside iframes; try using our regular embed code
- Perform basic performance analysis of the page to see when our API calls are being made if you suspect problems
- Thinglink works better for native advertising than for banner advertising
- Our views are not the same thing as impressions; they are only incremented when the interactivity in the image is available to the user.
- For impressions, you should probably rely on the numbers from your ad platform. However, if the impressions and view numbers are wildly different, you know that your users are leaving the pages before interactivity has time to kick in. In such cases, you should either consider reducing the amount of content on the pages, or consider alternate placements for Thinglink.
The click counts we track differ from the ones on your site?
While we do strive for accuracy here, there are also a number of reasons why this might be occurring: We track the user click events - and if the network or the browser fails before it actually goes to that link, we can count a click but you might not. Also, there are quite a few bots out there who are quite adept at following all the links on the site without actually clicking them. Most click trackers are fairly good at weeding these out, but sometimes one can slip through the cracks.
Can I use a single click tracker for the entire image?
Short answer: Only if you use a single tag.
Long answer: Click tracking URLs are unique per destination link. If you want to use multiple destination links, you will need to use multiple click tracking URLs, and manually add the clicks together.
There may be external providers which allow you to aggregate the clicks from multiple links to a single number, in which case you can put in multiple links but have a single number as a result. However, this is something you will need to raise with your tracking provider.
Thinglink's technology allows you to have multiple links per image, and therefore the traditional "single click per image" -tracking systems used by traditional ad providers does not convert well to our system. Therefore, the easiest way is to simply take a look at your Thinglink dashboard and see the aggregated clicks per image. You can also pull that information from our API.
Please sign in to leave a comment.