eCommerce visual search: 9 smart optimization tips (+ 4 tools to use)

“The future of search will be about pictures rather than keywords.” — Pinterest CEO, Ben Silbermann

The primary purpose of visual search in eCommerce is to make life easier for your customers. 

Hence, before you start optimizing your site it’s important to understand how your customers are adopting visual search. 

Intent Lab’s research throws some light on the effect online visual search has on customers. Here are their key findings:

• 36% of customers have used visual search and 59% of them prefer visual search over text search.
• Mobile is also the preferred device for conducting a visual search with over 53% of customers using it primarily.
• In terms of shopping activities, product comparison is the action that visual search helps the most in—as per 41% of customers.
  

Thus, placing yourself in the customer’s shoes and analyzing the results they want to achieve through visual search can help align your visual product search optimization.

Google’s case study also reveals how eCommerce businesses taking a customer-first approach can get the best out of visual internet search.

In Google’s example, CCC Group boosted their conversion rates by 4x—in comparison to text search—after adopting visual search.

In the words of Michal Pachnik, the head of eCommerce campaigns and mobile apps at CCC Group: “We found that consumers who use visual search are more likely to add products to their basket and buy them than those using a traditional keyword search.”

The likes of Google and Pinterest Lens have made it even easier for shoppers to find their desired products and complete the transaction. 

Once you’re aware of what customers are looking for, you can optimize your site accordingly to cater to those needs through your visual search results. 

Takeaways: 

• Optimize your product inventory for visual search to get your products indexed by Google and Pinterest Lens
• Choose which site optimizations you’ll need to do based on customer visual search data for your site (more on this in the next section)
• Choose a visual search tool that can help you achieve your goals (find the list of tools here)

9 visual search optimization ideas for your store

1. Optimize your images for search 

You put images on your product pages so that your customers can view them. 

However, they’re able to see them only when they’re discoverable. Visual search helps customers find your product images easily. 

To come up in the visual search results, using multiple images is a great tactic. This is because the more the images and the better they are indexed by eCommerce search algorithms, the easier it will be for your customers to find them through visual search. 

If you haven’t been taking this too seriously yet, here’s why you should care:

• It’s good for the user experience. Customers want to see products from different angles. 
• Google considers it to be a ranking factor and puts a lot of emphasis on it. So, if you add multiple keyword variations into each image, it’ll help them rank better. 
  

As per research, brands use an average of 8 product images per page: 

The research also found out that only 1.5% of product photos had lifestyle or contextual elements in them. 

What this means: customers landing on the product photos will have lesser relatability. 

Let us explain. Imagine a customer using visual search to find a home décor product. Among 2 images—one is of the product against a white background and the other one is placed in a room inside a house. 

Which one do you think the customer will relate to more? No doubt the second one! 

Customers using visual search are usually those on the purchasing stage of the eCommerce funnel. Hence, using contextual, real-life images that help them visualize the product will help close the deal better. 

Tips for retailers: 

• Use multiple product images per page. An average of 8 images is good. Adding a real-life context can help conversions. 
• Organize your images properly, for example by URL. Using an image sitemap can help Google easily find your product images. 
• Add multiple keyword alternatives for each image to improve your image rankings on Google. 
  

2. Compress your images for speed

You already know the problem with big images: they take a long time to load, hampering the user experience. A site that loads slow overall hurts your SEO. Ultimately, this will affect your Google rankings and reduce the chance to come up in visual search results. 

The solution is image compression—but done the right way. 

Delving a bit on the technical side, there are several factors to consider: image formats, pixel dimensions, bit depth, etc. 

Working with Scalable Vector Graphics (SVG)—an XML-based visual for rendering two-dimensional graphics—is advantageous since it saves you from loading unnecessary data.

Again, pixel dimension is an important aspect. For smaller images, the file size won’t seem much but for larger images, it can shoot up exponentially. 

Here, take a look: 

See how the file size increases with higher pixels and dimensions? This will cause a problem in loading these images on devices without a lot of memory such as lower-end mobile devices. 

This is where bit depth comes in. By reducing the bit depth, you can win 50% compression savings. 

But this requires your discretion—there’s no one-size-fits-all solution. You’d already know about lossy and lossless compression. You can choose which option to use based on your requirements. 

For example, for images where you’d want intricate details you can go for a lossless optimization, and for the others use a lossy optimization.

Tips for retailers:

• Use vector formats wherever possible since they are adaptive to multiple devices and higher resolutions
• Minify your SVG files through a tool such as SVGO and configure your servers to allow GZIP compression for your SVG assets   
• Identify images that are wrongly sized with Google Lighthouse and use tools such as ImageMagick to resize them correctly.
  

3. Always go long-tail

Old thought: Keyword research impacts SEO. 

New thought: Keyword research also impacts visual search. 

This again boils down to Google rankings. The better keyword research you do, the better your product descriptions include the right search queries—a higher chance for your images to appear in visual search results. 

Goes without saying that the goldmine is the long-tail keyword. It’ll help you target the specific search queries that your customers will be using. 

For example, a broad keyword like face oil will have a high volume with lots of sites competing for it. In comparison, a niche keyword such as vitamin c rose oil will have a comparatively lower difficulty and will be easier to rank for. 

This is what Biossance uses for its product description. See how they use it along with related keywords such as squalane oil naturally into the copy. 

Tips for retailers: 

• While conducting keyword research, include long-tail keywords for your product descriptions
• Use search terms that have an image stack for them—this means users want to see images for that search query. Google attaches image stacks to keywords that are re-entered in image search because this means users were looking for images in the first place. 
• Use Google autocomplete effectively. As per Google’s guidelines, popularity and freshness are key factors for auto-complete. So hit the search term on image search and see which ones Google suggests—they’re sure to have volume. (Note: Don’t forget to search for them in incognito mode to get unbiased results.)
  

4. The format matters

High-quality product photography is definitely going to bring people to your site—or if they are already on your site, then stay on it. 

But they have one more function: product discovery. 

Visual search engines identify images by their colors, patterns, and shapes and match it to similar or complementary colors, patterns, and shapes. If the images are of poor quality, the visual search engines won’t be able to identify or match them. Hence, the chances of coming up on visual search results will be less. 

Choosing the right image format can help you maintain the right quality for your product images. You can choose between vector and raster graphics. 

Pros of vector format: They are ideal for images with geometric shapes such as icons or logos. They also work well with high-resolution screens. They offer sharp results for every resolution and zoom setting. 

Cons of vector format: They don’t do well for scenes with too much detail such as photos. 

Pros of raster format: They are zoom and resolution-independent—hence adaptive for most settings. They depict complicated scenes well. 

Cons of raster format: They become blurry and the pixels burst when they are scaled up. You may need to save them multiple times at different resolutions to get the optimum result. 

Verdict: Use vector images wherever possible. In case you’re using raster images, use responsive ones.  

Here’s a guide to the common raster image formats:

Using desktop-suited images on mobile devices can take up 2–4x more data, hence you need to optimize them before using them. 

Tips for retailers:

• Use vector image formats as much as possible. 
• For high-resolution product images, choose raster images after selecting the format that applies to you. Opt for WebP images over older raster formats since they reduce file size by 35% and lead to 10% faster page loads. 
• Try to make your images responsive by resizing them and saving multiple versions. You can automate the resizing process with a tool like sharp npm package or use a manual one like ImageMagick. 
  

5. Add image badges

Image badges take the future of visual search one step further. 

Visual search is the equivalent of a super long-tail keyword. Customers using visual search often look for very specific results. 

Image badges aid the process to a great extent. They help users find relevant content easily. 

From a backend perspective, when you add structured data to your images, Google adds image badges to those images—this helps users identify the type of image and decide which action to take. 

As per Google’s guidelines, adding the relevant markup to images can help in their identification. For example, if you add the recipe markup this is how it’ll come up in visual search results:

This is how images with a product markup come up: 

Following Google’s product markup guidelines, you can earn yourself a place in the rich results. It’ll also help Google to offer detailed product descriptions that can attract potential customers. 

Tip: Ensure you update the images with the latest product information such as price, availability, and customer reviews. This will help you to maintain the accuracy of your product information and help users find content that’s relevant and recent. 

You can use the product markup both for single product pages or aggregator pages for a single product. 

Tip for retailers:

• Add product markup to your product images so that Google can show it in visual search results when customers search for it.
• Make sure your product information is up to date. Include shipping details and price drop alerts too in the product markup. 
• Use the Structured Data Testing Tool to check for product markup errors. Further, you can analyze the performance of your markups with the Rich Cards report. 
  

6. Optimize your image names, captions, and alt text

Visual search engines factor in a lot of information to deliver the most relevant content to customers. Hence, aspects such as image name, alt tags, and captions all play a vital role in discoverability. 

Adding relevant, short, and keyword-rich names can help search engines find your images easily. This way, even if the bots don’t understand the image contents, they can still identify it through the name. 

Check out this product by ASOS. 

Instead of a generic name, it uses a specific name for the product that describes very well what the product is: Crooked Tongues hoodie with fruity supply character back print in green | ASOS

Alt tags again help bots find your image as well as customers using screen readers to get an idea of your image content. Use keywords carefully here—avoid stuffing. Keep it relevant and clear for the user. 

The same goes for captions. Keyword-rich captions help boost visual search results as well as offer customers detailed information on the product. 

Tips for retailers:

• Update image names, alt text, and captions for all images.
• Make your image names and captions relevant and keyword-rich.
• Avoid stuffing keywords in your alt text. Ensure they explain clearly what the image is about. 
  

7. Experiment with other visual elements

Visual search results aren’t limited to images only. It can incorporate several multimedia options such as videos. 

In spite of the opportunity to experiment with this visual format, only 28% of brands are actually using video. 

Imagine this: would a customer be more hooked by a picture of a sofa or a 360° view of it? Obviously, the latter. 

You can experiment with it like how Ethan Allen does. It uses 360° product videos to entice their customers. 

In fact, you can understand the future of visual search from the fact it was identified at CES 2019 as one of the 3 V’s—voice, visual, and video—that’ll impact marketers and retailers. 

As per Intent Lab’s report, videos are a great catalyst in the consumer journey leading to a purchase. In fact, they are 1.2x more effective in reducing browse abandonment by helping in product evaluation—this brought them closer to actually buying the product. 

In the eCommerce funnel, videos are a great tool in the comparison phase of the shopper journey since they helped customers weigh options before purchasing. 

How does all this connect to you as a retailer: think video catalogs, shoppable 360° videos, AR and VR experiences, etc. 

This is how BoConcept creates an engaging video catalog to enhance the customer journey. 

The catalog features a how-to video of unfolding a sleeper sofa—something that’ll be a consideration while comparing products. This will help a great deal in bringing down their barriers and moving them to purchase. 

Tips for retailers: 

• Explore with 360° videos on your product pages to offer an immersive experience to your customers
• Help customers explore your products with an AR/VR experience to offer them more control during consideration
• Use video catalogs to help bridge the gap between consideration and purchase for customers

8. Customize your own visual search

Visual search is going to be the next big thing. 

Gartner’s research finds that brands adopting visual search early will be increasing their eCommerce sales by 30% within 2021. 

Moreover, the global visual search market is set to surpass $14,727m by 2023—at an average CAGR of +9% within 2018–2023. 

This is a great time to grab the opportunity and redesign your website to include visual search elements—before the adoption reaches a saturation point. Doing this will give you a headstart and iron out any technical challenges you may face while your competitors are still deciding whether to go for it or not. 

There are several case studies of brands that have adopted visual search successfully. Let’s look at a few: 

Neiman Marcus

• Launched its visual search app Snap. Find. Shop. that enabled users to click a 3D image and search for similar products on their site. 
• The search results were 95% accurate and were well-received by customers.
  

ASOS

• Launched its own visual search tool. In the words of Richard Jones—head of product and UX at ASOS—“If you know what you want, you can quite simply get to what you’re looking for. But what we’re trying to find is more of that discovery use case – if you’re not quite sure what you want, or you’ve seen something that’s inspired you, visual search is designed to kickstart that discovery… It’s about getting as close as possible to giving you something that is visually similar.”
• Revolutionized the mobile experience. Presently records about 80% of traffic and 70% of orders from mobile devices with an average time spent of 80 minutes per month on the ASOS app. 

Tommy Hilfiger

• Launched the Runway Recognition app to enable users to purchase the fashion they loved as soon as they saw it on the runway or a store. Also, the Tommyland app helped enable signboard and billboard scanning. 
• The visual search results recorded 100% accuracy for 2D and video and 93% accuracy for 3D. 

Alt tag: Example of visual search by Tommy Hilfiger

Tips for retailers: 

• Integrate visual search into your eCommerce site and apps with tools such as Shopix. 
• If you don’t have the resources to create your own tool, rely on search engines such as Google Lens and Pinterest Lens. 
• Keep monitoring for accuracy of results. This will help build trust. In Performics’s study, 37% of customers hesitated to use visual search due to a lack of trust
  

9. Include visual search features in your chat

50% of customers are dissatisfied with their chatbot experiences. Their primary concerns are slow response speed (59%), inability to understand the context (51%), and low accuracy of solutions (44%). 

The solution: visual bots. 

The chance of a solution to the customer’s problem depends on two things—the customer’s ability to describe their problem and the bot’s ability to resolve it. Both have limitations of their own. The customer is limited by text to explain their problem and the bot is limited by specific set of words and phrases. 

Visual bots help solve these problems. These AI-driven tools don’t read the problem but see it. There are already takers in the eCommerce industry. 

Levi Strauss & Co. launched their AI-based chatbot Virtual Stylist both available on their website and Facebook Messenger. It offers personalized sizing recommendations to customers by asking relevant questions such as How would you like your jeans to fit through your hips and thighs? 

Tips for retailers:

• Ensure that the images used in the bot are on-brand because the bot ultimately be the face of the brand
• Since chatbots don’t open into full-page, use clear images with fewer details so they don’t look clumsy
• Stick to natural images and avoid using artificial images. Also, the humans used in the photos should target your audience group. For example, if you’re selling a top for 20-year-olds using a 50-year-old human in the picture won’t help 

4 visual search tools to transform your customer’s shopping experience

If you want to stay ahead of your competitors, it’s time to invest in a visual search marketing strategy. 

Here’s a list of 4 tools that can help your get started: 

Vviinn

Why to use: 

• Removes shopping barriers such as complex navigation or irrelevant text search and replaces it with a visual search that helps land more accurate results
• Offers better product recommendations and helps boost click-through rates. 
  

Features to look out for:

• Image search directly from the gallery or a social media link
• Offers personalized product recommendations based on material, color, and textures

Slyce.it

Why to use: 

• Offers visual search services based on categories—industry, use cases, and tech
  

Features to look out for:

• Object recognition to analyze user-generated images
• Print recognition to scan retailer catalogs, packaging, labels, etc. 
• Barcode and QR code scanning

Syte.ai

Why to use: 

• End-to-end product discovery platform that offers a complete shopping experience for customers
  

Features to look out for:

• A visual discovery suite to take customers right from browsing to purchase—with image search and complementary product recommendations
• A hyper-personalization suite to boost conversions with relevant recommendations based on customer behavioral data and AI
• A searchandising suite to improve navigation and product discovery by enriching product descriptions and enabling deep tagging

Vue.ai

Why to use: 

• A complete digital customer experience management platform offering end-to-end automation 
• Offers specific use cases for retailers such as creating personalized shopper journeys, AI styling and outfitting, email and push notification personalization, and personalized listing pages
  

Features to look out for:

• Content management solutions such as product tagging, taxonomy management, and image quality moderation
• Customer experience management solutions such as eCommerce site personalization including personalized search
• Retail automation solutions such as product photography automation, site merchandising, and virtual dressing room experiences

Have you given visual search a thought or are you already implementing it for your store?