Gmail Image Caching Update Is Good News, Kind Of

I wrote late last year about changes to how Gmail handles images and the impact for marketers. My thanks to Jordan Cohen of Movable Ink for a heads up on a more recent change.

You may recall that Gmail started caching images in emails and doing so in a way that had a number of negative impacts for marketing messages, including loss of location, referrer, browser identification, cookies, timing, and total counts. These losses were mitigated by Gmail’s subsequent decision to load images by default, thereby increasing counts and improving how most marketing messages looked.

In a recent change, Gmail has started supporting cache control, a mechanism for websites to indicate how long it’s OK to keep content for without requesting an update. While this change does not address all of the capabilities that caching impacted, it does improve the situation for some of them. This is most important to companies that are dynamically serving image content, such as Movable Ink and LiveIntent.

One might reasonably wonder why the Google cache didn’t support cache control in the first place. Web caching has a bit of a checkered history and many websites effectively discourage caching of all their content, thereby undermining the reliability of cache control instructions. It’s also possible the Google developers were just being lazy – their original problem didn’t require cache control so they never implemented support for it. Whatever the reason, they have relented and now support the cache control header.

This means that suitably configured image servers will be able to ensure that every email open results in a new image being served. This is great if you wish to:

  • Track total opens (most email service providers)
  • Deliver dynamic content (Movable Ink, LiveIntent, etc.)

What has not changed are all the other things we lost when the cache was introduced. Gmail still hides:

  • The IP address, used to determine location
  • The user agent, used to determine browser and platform
  • The referrer string, used to determine email client and inbox placement
  • Cookies, used for data syncing and ad retargeting

The net result is that of the six pieces of data we originally lost, this change gives back two (timing and counts) but not the others. I do not expect to see the other four pieces of data coming back, ever. Though they have legitimate value and uses, each of them can also be used by nefarious actors to undermine privacy and in some cases security.

As a marketer, if you see a jump in your Gmail open rates, it does not mean you’ve found some great content, rather than Gmail’s total open rates are being recorded again. Which is good news, kind of.

Until next time.

Related reading