search engine optimization

Google "Universal" has placed increased emphasis on image results. Prior to Google Universal, users viewed search results one vertical at a time. Now days, users have options and access to search results from across key Google verticals all within their main search engine results pages. The Google internal video below illustrates in real-time how one eye-tracking study participant migrates between verticals within universal SERPs.

This migration between vertical search results may explain Google's introduction of Image Ads. In addition to ads, Google Images offers more options than ever before. Users can search for images by size or in a variety of content types including news, faces, clip art, line drawings and photo content categories. While a lot has changed when it comes to image optimization, users still enter text queries for Google to translate into image results. For that reason, linguistics is still critical when it comes to image optimization. Images are indexed in ways similar to text but have their own flavor of PageRank.

Before diving into the finer points of advanced image optimization, let's see if the basics are still valid. In terms of basic image optimization best practices, the rule of thumb is to provide as much descriptive information about images as is possible, but without "keyword stuffing" which could cause your site to be perceived as spam. Focus on including images with relevant textual context. Be sure to provide hypertextual clues about the subject matter of pages where images appear. With images like everything else the key concept is relevancy when it comes to search. Again, image results are based on textual queries.

Basic Image Optimization Best Practices:

- Informative filenames provide important signals about images and/or their subject matter. For that reason it's best to incorporate descriptive wording into image file names. (IE beach-dog.jpg instead of 1.jpg) (Hint - In some cases, image filenames may be used as the snippet in SERPs.)

- ALT attributes provide users and engines alike, with textual information about image subject matter. Engines rely heavily on the structure present in hypertext, especially where images are concerned. It's best to always incorporate short but descriptive ALT attributes when using images. ALT attributes assist engines in determining the most relevant result for image specific keyword queies.

Basic Image Optimization Case Study Experiment:

To determine whether or not images ranking for the query used in Google's eye tracking study above, follow basic image optimization best practices.

To examine the hypertext structure related to images appearing in Google Universal search results for the query [how to tie a tie] and record observations related to basic image optimization.

Image A -
ALT = "How to tie a necktie video"

Image B -
ALT =""

Image C -
ALT = "Tying the Tie"

Two of three images in this case study have descriptive file names in conjunction with descriptive ALT attributes. In addition to file names and ALT attributes, there seem to be other key factors to consider when optimizing images. As Google states, "other factors" do seem to come into play based on the results of this case study.

Advanced Image Optimization:

It's important to note that Google uses crawl caching proxy techniques to make images available in other services and that Google doesn't index images directly. As a result, there is no need to include images in your XML Sitemap. Either way, quality images start with quality pages. Quality pages contain few errors and load quickly. In order to decrease load time for pages, focus on template images that appear in every page (navigational images, logos and/or other). When possible, consider converting static GIFs into PNGs. When using GIFs, be certain palette sizes are correct based on the number of colors in the image. For JPEGs use a lossless tool to like Photoshop to remove unnecessary information from your file unless it's important for users and/or search (see Exif below). Always define image sizes via (X)HTML by size and not scale and be sure to include a favicon for branding, bookmarking and to avoid 404s. These steps will help decrease load time and increase page quality.

Once quality has been addressed from a technical perspective, it's important to be sure URL paths aren't blocked by JavaScript, Flash, Robots.txt and/or other obstacles. Images, like text can be scraped and republished multiple times. To assist engines in identifying your images as "source images", provide as much information as possible via hypertext, metadata and link structure in your pages. When using Flash, always provide alternative images with static URLs in GIF, JPEG and/or PNG format otherwise Google won't display images from Flash in SERPs. As a webmaster, you may have little control over external anchor text but you can control internal anchor text within your site. Be certain internal anchor text for images is descriptive and relevant. For some images (ie product image in category page, video, other) you may want to use thumbnails with unique ALT attributes and create detail level (product, video, other) pages. Provide quality images in a variety of sizes and formats including JPEG when possible and allow others to use your images.

Place high quality images high up and above the fold if possible. When necessary create unique static detail pages linked via thumbnail. Don't prevent other sites from using your images even if it means loosing bandwidth. Posting and tagging sample images at other sites can help get more eyes on your images, but be sure to include links to your site as the objective isn't to donate content to social media sites.

"Google analyzes the text on the page adjacent to the image, the image caption and dozens of other factors to determine the image content. Google also uses sophisticated algorithms to remove duplicates and ensure that the highest quality images are presented first in your results."


Always provide textual content in close proximity to images. Obviously, at this point it's worth pointing out again that image accessibility is of paramount importance. In addition, TITLE elements, captions and image titles in your pages can provide important clues for search engines. When it comes to search, more data is better data so consider taking the extra time to include as much as possible while avoiding techniques that could be detected as an attempt to spam search engines. Be sure to place your images near or above relevant text in pages and always include descriptive captions. Don't embed textual content within vector graphic formats other than .pdf as engines can't extract text from other image formats. Fresh images accessible in various sizes are a good idea but be sure to define image length and width information via hypertext as well as for users. Enable Google Image Labeler via "Enhanced Image Search" in Google webmaster tools. While PNG is more optimal than GIF in terms of load speed, certain circumstance may require image metadata contained in JPEG. Consider a file structure for images that denotes individual directories for thumbnails, art, drawings, photos and/or other. Be sure not to mix Adult images with images for general audiences.

The Future of Image Optimization:

In terms of the future of image optimization, all signs seem to point to Exif. The Exif file format for image metadata is one specification used by digital cameras and was developed nearly 10 years ago. Exif metadata includes information such as the date and time an image were captured. When set properly, modern digital cameras record the date and time images were captured and this information is recorded in image Exif metadata associated with image files. In addition to date and time, digital cameras record image metadata pertaining to the camera's manufacturer, model, orientation, aperture setting, shutter speed, focal length of image, meter mode, ISO speed information, a preview thumbnail and copyright information.

So, what does any of the information in Exif have to do with SEO for images you ask? Well, Exif can also be used to record information about where images were taken and whether or not they've been "photoshopped" for example. For images or universal queries related to news or specific geographic areas, Exif could easily be used as a quality signal. I asked Matt Cutts about Google's use of Exif last year and his reply was "I'm not sure, personally. I could imagine that any stuff embedded in an image file might be used, though." Currently both Panoramio and Picasa use Exif and I'd expect to see this trend rise as new GPS enabled devices enter the market.

For more great information, check out Peter Linsley's latest post on the Google Webmaster Central Blog...

While Google has condemned buying and selling links that pass PageRank, they've encouraged listing in paid directories like Yahoo for years. It seems that era may have come to an end earlier today. The following bullet points have been removed from Google's Webmaster Guidelines Webmaster Help Center*

  • "Have other relevant sites link to yours."
  • "Submit your site to relevant directories such as the Open Directory Project and Yahoo!, as well as to other industry-specific expert sites."

Does this recent move reflect a renewed emphasis on rooting out paid links passing PageRank and/or low quality links by Google?

*As mentioned, the bullet points above have been removed from the US version of Google's Webmaster Help Center. Other versions may not yet reflect this change.

UPDATE: Hat tip to Barry Schwartz who noticed John Honeck's post in Google Groups where Google's John Mueller comments on the change. Barry provides a full recap at and

Google recently modified how they show results and in doing so virtually crippled at least one popular "rank checking software" package. As "JohnMu" pointed out, Google has always been clear about using these kinds of tools. In fact for as long as I can remember, Google Webmaster Guidelines has clearly stated:

"Don't use unauthorized computer programs to submit pages, check rankings, etc. Such programs consume computing resources and violate our Terms of Service."

Bottom line, automated queries require resources without any potential for generating revenue. It has been said that Google's "ultimate selection criterion is cost per query, expressed as the sum of capital expense (with depreciation) and operating costs (host-ing, system administration, and repairs) divided by performance."

During Q4 2007, Google reported capital expenses of $678 million with operating costs of $1.43 billion. According to ComScore 17.6 billion "core searches" were conducted by Google during the same period. Using Google's formula and financial data along with ComScore's estimates, it appears as though Google's average cost per "core search query" was nearly $.12 during Q4 2007. Again, this is a rough estimate and a rounded total but, personally I was a little surprised by the number.

If 1 million sites run ranking reports on 100 keywords 12 times per year at $.12 per "core search query", it costs Google $144,000,000. Over a ten year period, that's more than a billion dollars. Given this data, it's easy to see why Google uses "algorithms and different techniques to block excessive automated queries and scraping, especially when someone is hitting Google quite hard." Matt Cutts suggests, contacting Google's Business Development Team about permission on sending automated queries to Google.

Now, I fully understand the importance of ranking reports when it comes to SEO clients. That said, there are folks out there abusing the system, running ranking reports on thousands of keywords daily.