GPT Light passback tags

In situations where ad requests to a third-party ad server should ultimately be filled by ads trafficked in your own DFP network, you should traffic passback tags in the third-party ad server. Passback tags are used when the ad tag on the publisher page isn't yours, but the publisher would like to send you the impression. In this case, the publisher traffics your passback tag, giving you an opportunity to serve or monetize on the impression.

GPT Light passback tags have been optimized for speed, while retaining most features of Standard GPT passbacks: compact syntax, key-value targeting, click tracking, page URL overrides, and category exclusions. Most restrictions that apply to Standard GPT passbacks also apply to GPT Light; for example, you cannot use single request architecture (SRA) or deliver companion ads. However, GPT Light tags are always asynchronous, while asynchronous mode is not supported for Standard GPT passbacks.

GPT Light tags use simpler, declarative syntax and typically have lower latency than Standard GPT passback tags, which may reduce discrepancies with third-party ad servers. However, GPT Light tags support only one size per ad unit. For multi-size ad requests, consider using Standard GPT passback syntax instead.

GPT Light should only be used for passback tags. For any other situations, please use Standard Google Publisher Tags (GPT).

Create a GPT Light passback

To generate a GPT Light passback tag:

  1. Sign in to DoubleClick for Publishers.
  2. Click Inventory and navigate to the ad unit for which you want to create an ad tag.
  3. Click Generate tags.
  4. Select Google Publisher Tag from the drop-down and click Continue.
  5. Under "Select Tag Options," check Create passback tag.
  6. Select GPT Light Syntax.
  7. (Optional) Make additional selections such as inventory unit size and inventory level unit targeting.
  8. Click Continue, then traffic the resulting tag as you would any third-party creative.


Examples of GPT Light passbacks

Here are a few examples that highlight the differences between regular GPT passbacks and GPT Light passbacks.

Basic example of a GPT Light passback tag

Standard GPT passback:

<script src="//www.googletagservices.com/tag/js/gpt.js">
 googletag.pubads().definePassback('/1234/MySite/MySection', [300, 250]).display();
</script>

GPT Light passback:

<div data-glade width="300" height="250" data-ad-unit-path="/1234/MySite/MySection"></div>
<script async="async" src="https://securepubads.g.doubleclick.net/static/glade.js"></script>

Example of a GPT Light passback tag with custom targeting

Here's how each type of GPT passback tag looks if you add custom targeting under "Select Tag Options."

Standard GPT passback:

<script src="//www.googletagservices.com/tag/js/gpt.js">
 googletag.pubads().definePassback('/1234/MySite/MySection', [300, 250]).setTargeting('sports', ['football', 'rugby']).setTargeting('teams', ['Real Madrid', 'Olympique Lyonnais']).display();
</script>

GPT Light passback:

<div data-glade width="300" height="250" data-ad-unit-path="/1234/MySite/MySection" data-json='{"targeting": {"sports": ["football", "rugby"], "teams": ["Real Madrid", "Olympique Lyonnais"]}}'></div>
<script async="async" src="https://securepubads.g.doubleclick.net/static/glade.js"></script>

Was this article helpful?
How can we improve it?