Visual search

About one in four Internet searches is a search for a specific image, although, sometimes words are not enough for the artificial intelligence to process the request. Visual search comes to the rescue.

By making it easier for customers to interact with the search engine and reducing the time spent, you receive a large influx of target audience.

We are going to tell you about the most famous cases of visual search in 2020-2021.

 Pinterest Lens

Unsurprisingly, Pinterest quickly introduced this feature to their app. The company released Lens add-on — a visual search tool that allows customers to take pictures of an item to find out where to buy it, search for similar products, or view ads on sales sites or other platforms. According to Marketing Land, they turned the phone camera into a search bar:

Since launching in beta, Pinterest Lens has learned to recognize over 2.5 billion home furnishings and fashion items, processed nearly 600 million searches in Pinterest apps and browser extensions, and has seen a 140% increase in searches since launch.

Pinterest continues to update its functionality:

  • They introduced Pincodes, a QR code-like technology that allows you to save the items you find and sort them into one folder while you shop or flip through your favorite magazines.
  • Pinterest launched Lens Your Look to generate images from saved clothes.
  • The company has partnered with brands such as Samsung to provide visual search on their latest smartphones, and Target so that customers can search their catalog using related products they see in the real world.
  • They have fully automated Shop the Look, a feature that helps users buy products from companies that use Pinterest, so customers can easily order online.
  • Pinterest has created a number of new tools to help companies sell their products, including modernizing their product catalogs.

The main search categories for Lens are:

— Fashion

— Home decor

— Art

— Food

— Products

— Animals

— Outfits

— Beauty

— Vehicles

— Travel (countries)

 Google Lens

Google Lens — is a visual search engine from Google that recognizes objects and details through a camera. Here is the functionality of the application, after taking a picture, you open the following parameters:

— Clothing and household goods: similar items and stores to buy them at.

— Barcodes: use the barcode to find information about a product, such as where to buy it.

— Business card: save phone number or address in your contact list.

— Book: leave and read reviews.

— Advertising brochure or billboard: add events to your calendar.

— Landmark or Building: find out historical facts, opening hours, and more.

— Painting in a Museum: read about artists and find out more.

— Plant or Animal: learn about species and breeds.

This app has been downloaded almost 1 billion times.


CamFind is another visual search mobile app that lets you search for anything on your phone by simply taking a picture.

A few months ago, CamFind introduced its version of the application for Google Glass, which allowed the owner of «smart» glasses from Google, looking at an object, to find out what it is. The app carried out object recognition thanks to the built-in visual search engine. For example, CamFind could tell that a person is facing Notre Dame Cathedral or a Newfoundland dog.

The previous version of the CamFind API was used by over 750 companies worldwide.

CamFind has raised nearly $5 million in another round of funding. At the same time, the startup announced the launch of a new version of the API called CloudSight.

 Bing Visual Search

Using the Bing Visual Search, you can easily search for a specific element in an image without having to go through all current searches.

You are doing a normal image search, and when you click on one of the results, the search icon appears. After you click on the «Visual Search» button, you can select the area of interest.

Bing will not only try to find the right images for you, but also display product pages whenever possible.

Amazon StyleSnap

StyleSnap is an artificial intelligence app that helps you shop with just a picture.

To begin with, all you have to do is click on the camera icon in the upper right corner of the Amazon app and select the StyleSnap option, and then just upload a picture or a screenshot of what you like.

StyleSnap displays recommendations for similar products on Amazon that match the given image. StyleSnap takes many points into account when providing recommendations, such as brand, price range, and customer reviews.

StyleSnap uses computer analysis to determine the quality of its products, so it can suggest different categories, such as, for example, linen shirts.

According to Social Media Today

— 62% of millennials are more interested in visual search capabilities than any other new technology.

— Image search increased by 19% of Internet searches in Google

— More than 600 million visual searches are made on Pinterest every month.

“Marketers can gain an edge over the competition by taking advantage of the 2021 visual search trend to reach customers and offer them the perfect product.


So how can you increase your customer base with visual search?

Name your images correctly and make sure they all have correct titles and alt tags. There is not a lot of data to read in visual search, so it is very important to know the basics of SEO.

The filenames of all your images should describe the image, and the titles and alt tags should contain your product keywords. Simple enough, but it is important to help search engines understand the content and context of your images so that they can index them correctly.

Submit images to XML image sitemaps. This step ensures that search engine crawlers can easily access your images and show them among search results.

In terms of products, make sure the images are displayed clearly and are not loaded randomly so that the AI of visual search makes them as easy as possible to process.

Make a contract with search engines. For example, consider advertising on Pinterest. As noted above, Pinterest image searches have very high conversion rates, and its users are already used to searching for images.

Add to that the fact that half of all Pinners made a purchase after viewing a Promoted Pin, so testing your campaign here becomes an easy task for fashion and home décor brands. Pinterest offers an image capture — «Shop the Look» pins where users can select specific parts of the image and be redirected to pages where a purchase is already in progress.

Let’s look at visual examples of brands

H&M created different “boards” on Pinterest  — collections on different topics: collaborations, for children, leather clothes, underwear, best gift for men, etc.

As soon as a client finds an H&M item using the app, Pinterest immediately offers collections on the requested topics. The button «subscribe» to the pin-board is also displayed, so that the client will always receive notifications about new pins (pictures), that is, about new arrivals.

It is also interesting to analyze what Instagram visual shopping opportunities are. So far, the company has not created a search by image, but allowed brands to mark in the photo — details, which, in turn, have a direct link to the product, and also immediately indicate its cost, so any visual search engine will be able to redirect the client to an Instagram page, where there is information about the desired purchase. And will also select similar products of the brand.

For example, the profile of Kyliecosmetics and KylieJenner (a beauty brand from a reality TV star). Kylie has created her own brand and advertises it on her Instagram page, and also created a separate account for her cosmetics. The subscriber just needs to click on the product, and the panel immediately shows the cost and similar products:

The future of Visual Search

It is only a matter of time before a similar program from Facebook enters the market. They acquired GrokStyle, an artificial intelligence startup providing visual search at IKEA. Given the value of visual search to e-commerce, the Facebook Marketplace seems like a natural place to introduce this technology.

Interaction with Instagram also seems inevitable. Instagram ads are already working, adding images to the product page. They are currently directly linked to the advertiser, but moving to an automated process where users can do this from any photo would be a powerful nexus.

Thus, the more visualized your brand on the Internet is, the more customers you can attract.