Google's New Image Format: WebP

Last night, Google released a new image format: WebP. Designed to compete with JPEG, they are claiming approximately 40% reduction in file size without, according to the post, "perceptibly compromising visual quality". Since images still make up almost two-thirds of web traffic, this is potentially a major step towards speeding up the web.

Here's a side-by-side from their comparison image gallery:

JPEG: 1,175,642 bytes
WEBP: 864,134 bytes (26.50% reduction)

I can see differences between the two when they're next to each other, but nothing I think I would notice seeing them independently. And I don't think I would be able to identify which was the WEBP in a blind taste test, so to speak.

The biggest obstacle right now is simply that none of the major browsers are ready to support the new file format. Microsoft is making a similar play, backing the JPEG XR format, and if history is a good indicator Apple will be a long-term holdout, so it's not guaranteed the format will be widely adopted. On the other hand, Google is one of the net's juggernauts, so throwing their weight into this might be enough to make it happen (although see Buzz and Wave). If one other major player, such as Facebook or Twitter, picks up on the new format, it will probably be a lock.

A few thoughts on how this will impact our industry in the short term:

  1. Architecturally-focused pages can be very image intensive; as we've all seen, it's common on such pages for the text to load quickly and images to take much longer. If site navigation or important parts of the message are in those images, that can negatively impact a user's experience with your site. Smaller image sizes mean quicker downloads, mean better experiences with image-heavy sites.
  2. Sitting in my office with the DSL running straight into my laptop, it's easy to forget that not everyone has a reliable high-speed connection. About a third of Americans still do not have access to high-speed connections, especially outside of urban areas. Better image compression will make it easier for these net users to access your page.
  3. This will be a huge boon for mobile! While I'm in the office my phone connects via wifi and downloads images about as fast as my computer, but when I'm on the road and downloading over 3G, or using the airport's free - but slow and crowded - wifi on phone or laptop, download rates crawl. Better image compression will improve your reach with mobile users, which is becoming a key marketing issue. Plus, smaller images means you can store more on your mobile device, making it easier to bring your entire photo library with you on the road. 
  4. When I send photos to editors, usually I start by sending a collection of thumbnails; they pick their favorites, and I send full-sized images using ftp or a file transfer site like To fit the thumbnails comfortably through email, image size usually gets reduced along with image quality, meaning the editor is looking at a bunch of 2" square images. For the same file size, WebP could potentially let me send larger images via email, making it easier for editors (or prospects!) to see details in the compressed photos. 
  5. There is a potential SEO benefit; if Google image search becomes optimized for WebP, images in that format will return higher in search results.

  1. We don't usually think about it, but most file formats start out proprietary; it blew my mind when I met the man who created MP3. After a while the format becomes so ubiquitous that it doesn't matter as much where it came from. Short term, however, this will be a proprietary technology which means many web users may not be able to view them properly. If you are concerned about universal accessibility, I would recommend you move slowly in adopting WebP.
  2. Google has managed to maintain a great deal of trust as they have grown, something most large corporations have trouble managing, and they have done a good job of acting in a manner worthy of that trust. Still, between search, Gmail, YouTube, Maps, and Docs alone, they have their hands in a huge percentage of my data; handing over all of my images might be too much.
  3. Given how many image formats are already available, adding another increases the chaos. Sorting our photo library is already tricky, and it is not uncommon to find the same image in JPEG, TIFF, DNEG and PSD formats on our hard drive. We just added a new 1.5 TB drive, so I'm not too worried about storage space (it should last us through the end of the year, I hope) but keeping track of which version is color corrected, which is cropped, which is the original, which has up-to-date metadata, etc. can be a headache. It's unlikely we will go to the trouble of converting our entire existent library, so there is a compelling reason to stay with our current formats.
My initial gut reaction to this news was that Google is overreaching again, and this will become another failed program (which, to their credit, never stops them from attempting the next innovation). As I think more about it, though, if it is successfully adopted it could have a beneficial impact on the way the net works. For now I'm holding out until there is confirmation of support from browsers besides Chrome, but this is definitely a story to watch.

[H/T to ReadWriteWeb for the tip]