When Amazon rolled out StyleSnap at its 2019 re:MARS conference, the internet giant officially joined the growing ranks of retailers transforming the way consumers find what they want on the web.
Visual search tools like StyleSnap or Pinterest’s Lens harness the power of advanced neural networks and machine learning, enabling search engines to recognize colors and patterns in user-uploaded images and return relevant search results.
In a visual search, the image is the query. The technology behind it isn’t exactly new; Pinterest launched Lens in 2017, and Facebook has been using artificial intelligence (AI) to recognize human faces in user-uploaded photos for some time.
Companies like Amazon aren’t stopping at still-image recognition, however. Today, Amazon’s AI can even recognize items in user-uploaded videos, speeding up retail’s progression toward video catalogs, shoppable 360-degree clips, and point-shoot-buy technology.
Ecommerce and retail are changing fast, but the changes make sense. Humans are visual creatures, and visual search represents a natural evolution as AI gets closer to developing human-level thinking. While the internet in its earliest days was composed entirely of text, today’s web is truly a visual landscape, and text-based searches are often inadequate for navigating this landscape. That’s why 62 percent of Gen Z and millennial consumers in one study reported wanting visual search more than any other emerging technology.
With visual search, users visiting your website looking for a jacket or pair of shoes they love no longer need to describe the clothes to a search engine using text—they can just drop in a picture.
The simplicity and effectiveness of visual search has major implications for ecommerce, but it will also impact the travel industry, food industry, and others.
Imagine vacationing in another country, for example, and wondering about the history behind your favorite landmark. Just snap a picture on your phone, drop it into a visual search engine, and instantly retrieve all the information you need. Or maybe you want to recreate an amazing dinner you had on a night out. Use a picture of your meal as a query, and a visual search engine can direct you to a page with all the ingredients and a step-by-step recipe.
The possible uses for the technology in various industries are endless, but one thing is certain: If you’re not optimizing your ecommerce site for visual search, you’ll be left behind as the trend continues on an upward swing.
Preparing your site for visual search
Visual search is here to stay, and image-based search technology is improving as more people use it. Moving forward, ecommerce websites that aren’t optimized for visual search will be leaving revenue on the table—or, worse, losing profitability.
To avoid missing out on the next search revolution, follow these eight guidelines:
1. Add images to your sitemap
You want visual search engines to be able to easily scan your site for images. Make sure to create a large image library of well-organized original photos and include those on your sitemap.
Here’s an example of a typical sitemap, organizing the pages of a website by URL:
2. Pay attention to sizing
Images make up 21 percent of the average website’s total weight, but less is more when it comes to image sizing.
When exporting an image, choose the best file type for your needs. JPEGs are generally the easiest to adjust, while others like JPEG XR and WebP aren’t universally supported by all browsers. Always aim for high-quality images that don’t compromise page load speeds.
When exporting an image, you can adjust quality, proportions, and other factors to control overall size.
3. Give each image a name
Remember, search engine crawlers don’t just scan your website copy; they crawl image file names as well.
According to Google, file names and alt text work best when they’re short but descriptive. If a search engine can’t visually recognize an image on your site, a unique, descriptive name using relevant keywords will ensure it still gets returned in search results.
When saving an image, type in a distinct name that alerts crawlers and helps you stay organized.
4. Don’t forget a caption
Captions give search engines context when images can’t be identified right away. Unlike an image name, captions are only tied to the point where you choose to place an image in your post. You can use the same image in different posts or in different places within one post and include a unique caption each time.
As with image names, use SEO keywords in your captions to eliminate room for error when your page is scanned.
The box for adding captions can be found below “Title” and “URL” on your WordPress page.
5. Always include alt tags
Alt text makes the web more accessible to users relying on screen readers or browsers that block images. By including an alt tag with every image, you ensure that users with visual impairments are still able to appreciate the imagery on your site.
In terms of search, alt tags play a similar role as names and captions but are used by the search engine when a browser is unable to render the image. Always write your alt tags with the user in mind, not just SEO. Too many keywords will cause the search engine to mistake your content for spam.
When uploading a picture, alt tags may autofill. Be sure to include your own text to help optimize your page.
6. Optimize for text search, too
You want to be seen and heard by search engines, so don’t ignore traditional text-based search. Instead, maximize your site’s visibility by optimizing for both types of searches.
Use full-text search to retrieve a single document from a full-text database.
7. Stay active on social media
Regularly posting images on social media can be great for visibility—just make sure to optimize images based on the platform you’re using.
For instance, Instagram pictures are organized in a grid, so each piece of uploaded content should visually complement those around it.
Make sure to account for ideal posting ratios and time, as posting too much can actually decrease engagement per post.
8. Consider chatbot integrations
Chatbot technology continues to evolve, and some chatbots now offer visual search capabilities. This allows for streamlined interactions with users based on visual inputs and can significantly enhance your customer service.
Chatbots can ask questions and provide real-time answers based on user input.
The rise of visual search promises to completely transform the way consumers navigate the web. As more major retailers begin to implement their own AI-powered search tools, the technology will improve and proliferate rapidly.
Start preparing your site now, and you’ll make your vision of increased sales and profits a reality.
Hope Horner is CEO and founder of Lemonlight Video Production, a company that produces branded video content at scale. Hope is a three-time entrepreneur who has been featured in Inc., Entrepreneur, Forbes, and other publications highlighting her successes in the Silicon Beach community over the past decade.
