"Man, the eyewitness, needs the picture." This quote is from the Italian universal genius Leonardo da Vinci. Looking today for his world-famous oil painting, Mona Lisa promptly smiles at us from the Knowledge Graph on Google's or Bing's search results page. Search engines seem to know exactly what we want to see when we type "Mona Lisa" into the search box. But where does it come from? How is it possible for search engines to process and understand images?
The simple answer is learning. Every day we feed the artificial intelligence (AI) of the search engines with information by providing uploaded images with descriptive data and additional information. It can be the file name of the image, the text in the vicinity of the image, or the metadata. The artificial intelligence reads the picture along with context information and compares it with the existing database. Based on the new information, the database can be confirmed or expanded. In addition, the artificial intelligence is now quite good in reading content directly from a picture. If you want to put the Google artificial intelligence to the test, you can do it with Google Vision.
We can help the search engine on the jumps
Even though search engines have meanwhile independently translated image content into text in order to understand a picture, we can still provide assistance. By accurately describing and labeling our images, we explain to Google, Bing & Co., what the picture is about. At the same time, we increase the likelihood that the search engine will return our image as a result of a matching query.
There are several factors that have varying degrees of influence on the image rankings. Let's take a closer look at seven key factors:
- Alt text: The alternative text (short text alt) is used to output content when an image file can not be loaded. In the alt text, the image content should therefore be described as accurately as possible, in the optimal case using the relevant keywords. It is also very helpful for visitors with visual impairments when the contents of a website are read out by so-called screen readers and thus provides barrier-free accessibility. Search engines, especially Google, use the alt tag to better interpret the image content if the filename does not immediately show this information. Thus, essential information can be transmitted, which also contribute to a positive listing within the image search.
- Title Tag: Images can also be tagged with a tag tag that provides search engines with more information about the image. But for the users of the title tag is helpful, as it, as soon as you move the mouse over the image shows more details. So, if here information about the picture is transported in a short and precious way, this is also an advantage.
- Filename: In the past, the file name was one of the most important ranking factors for pictures. It was usually recommended to give a meaningful file name. Although this factor has meanwhile lost in importance, it still remains advisable to assign speaking file names and to select them in such a way that they describe the contents of the respective picture as exactly as possible. As a separator between terms, a hyphen is preferable to an underscore.
- File size: Where we are in the user experience - this will also suffer if the images have a poor quality. On the other hand, a high resolution of images ensures that users perceive our site as professional, which positively influences the perception of the site or even the purchase decision. A counterpart of high-resolution images is again the file size and associated loading time The larger picture, the higher the loading speed of the entire page. Since loading time affects the placement of a page in organic results, compressing image files is very important for balancing quality and file size. In addition, images should always be provided in the optimal dimension, ie for the representation of the maximum number of pixels.
- Context: From an SEO point of view, a page is optimally content if it has a clear keyword focus. Ideally, images supplement the content of the pages on which they are placed. Not only the user, but also the search engine is happy about relevant pictures. This will be made easy in the thematic classification of the page, when pictures, texts and other elements treat the same topic.
- Markups: With the help of structured data like schema.org, the contents of a page can be better read and understood by search engines. These data not only facilitate crawling, but also the thematic assignment. Images should also be provided with structured data, especially on product and recipe pages.
- Image sitemaps: Another way to provide search engines with more information about images is through so-called image sitemap. These are Sitemaps that list paths to images instead of HTML pages. Once the Image Sitemap has been maintained, the search engine records and indexes all relevant image content on a website. By means of tags, further features such as the geographical location can be transmitted to each image.
The content of a picture should always be in line with the message that a website or individual page wants to meet (keyword: relevance). Using file names and alt texts, we provide the search engine with important information to understand our image. Thus we increase the chance to rank with our picture with an appropriate search query.
Technical tricks like image compression, markups and image sitemaps complete the image SEO. If we perform all these tasks, in the future, when entering an appropriate search term, we may be able to smile at a picture of our own website. At the end of the day, not only search engines but above all the users thank us for excellent contents in image and text.