Editor’s Note: As 2013 comes to a close, we’re pleased to share our top email columns of the year. This article was originally published December 9, 2013.
The dust has barely settled from the furor around Gmail’s tabbed inbox when another change comes down the pike. I noticed late last week that Gmail has started caching images, though there hadn’t been an announcement from Google. I wasn’t the only one who noticed, either.
There is no doubt there will be cries that Google has, once again, killed email marketing. But to understand what impact this may have requires that we start by understanding what has changed.
Email messages are, by and large, pretty static. Once they’re sent, they’re fixed. The one exception is the images, which are downloaded when the recipient reads the message. As that happens, a treasure trove of information is shared with the image host. Many enterprising companies including ESPs, MovableInk, ReturnPath, Litmus and AudiencePoint use that data in a variety of ways.
By switching to caching the images, Google has caused that source of information to largely dry up.
There are six pieces of information that marketers often rely on that the Google cache impacts:
The Google cache hides the source Internet Protocol (IP) address of the reader. That address can be used to determine the location of the reader. Gmail users now all appear to reside in Google datacenters.
Often, this is used for nothing more than drawing pretty maps showing readership around the world. But some companies also use it to provide more targeted information, like an image with the address of the outlet that is closest to you right now.
The referrer indicates what requested the image be downloaded. This data is heavily used by reports that tell you which folder your email ended up in and how many recipients were using the Gmail web interface versus a native client or the mobile app.
The user-agent string tells a site a lot about the program that is downloading the image. Browser version, operating system and device are often deduced from this. The impact of not having this is two-fold.
Statistics on email client and device usage will be unavailable for Gmail users, but also any tool that customizes images for the audience will be unable to function properly. So delivering a mobile-optimized image for a mobile device versus a desktop will not be viable (responsive design notwithstanding).
Cookie synchronization for ad serving and retargeting may be undermined. Advertisers are increasingly performing cookie synchronization between ad data management platforms and their CRM databases. The benefit is more relevant ad targeting and more timely and relevant email. Gmail may become a black hole for this activity.
Tools that deliver images dynamically based on time (think time-limited offers) will be negatively impacted. Once the recipient sees the image, it will not be downloaded again and so will not be able to change.
Since images are cached they will, at most, be downloaded once per recipient. Total opens will fall. Tools that use multiple image downloads to infer read times and repeat reading will be negatively impacted.
One other impact of the Google image cache is that it restricts images to 10MB. Any image larger than that produces an error. While not an issue for most, this can be significant for animated images.
If you’re wondering why Google would make these changes, I can think of a couple of reasons.
The first is privacy. Unlike most other webmail providers, Google has always hidden the submitting IP address of emails. This change also masks the IP address of readers, as well as other information about them that might be considered sensitive (browser information could be used for exploits, for example).
The second is performance. Google’s cache service is fast, but some senders’ image delivery isn’t. This can provide a better experience for Gmail users.
There are other possible reasons, but most of them require ownership of a tin-foil hat so I won’t get into them right now.
As for what to do about it, talk to your email provider. For most email marketing, the impact will be on reporting and analytics rather than functionality. If your ESP provides gross or total open rate numbers, these will likely shrink. Providers that collect and report on statistics around location, device etc., are already working to exclude or adjust for the Gmail data and see if there are other ways to obtain the same information.
The larger impact though is if you’re delivering images dynamically based on device, time or location. In these instances, you will certainly need to make changes. The most likely outcome is that Gmail users will get a degraded experience, rather than the improved experience Google was aiming for.