Why Choose UsAbout UsClients and TestimonialsCareersLocations and City Guides
ServicesSoftware DevelopmentWeb DevelopmentMobile App DevelopmentSaaS DevelopmentCloud ServicesQA and TestingUI/UX DesignDesign MarkupHire ResourcesCorporate TrainingDigital MarketingData and AnalyticsCloud Telephony
SolutionsAI and ML SolutionsAI Marketing SolutionsCRM Sales AutomationCybersecurity and CloudStartup SolutionsTechnology Services
IndustriesHealthcareEducationBFSISaaSManufacturingE-commerceTravelEV SolutionsSupply ChainAgricultureEntertainment
Free ToolsAI Token CounterAI Cost CalculatorPassword Strength CheckerWebsite SEO AnalyzerMeta Tag GeneratorSchema Markup GeneratorAI Marketing ROI CalculatorUTM Link BuilderQR Code Generator
BlogContact Let's Talk
SEO and Organic Growth, illustrative photo

Using a Visual Search Tool for Shopping and Product Discovery

April 18, 2017Updated November 5, 2026By Piyush Kumar9 min readSEO and Organic Growth
In this article
  1. Key takeaways
  2. What Is a Visual Search Tool?
  3. Similar Items in 2017: Where It Started
  4. How Image-Based Product Discovery Works
  5. Where Visual Search Shows Up Today
  6. How to Optimise Product Images for Visual Search
  7. Visual Search vs Keyword Search
  8. Which Products Gain Most
  9. Common Mistakes to Avoid
  10. Conclusion
  11. Frequently Asked Questions

A visual search tool lets a shopper search with a photo instead of words: point a phone at a handbag, or upload a screenshot, and get similar products with prices and the stores that sell them. This guide explains how image-based product discovery works, where it appears today, and how online stores can prepare product images and data.

Editor's note: this article was first published on 18 April 2017, a week after Google announced "Similar items" in Google Images, a feature that found shoppable products inside lifestyle photos. That launch is kept below as history. The rest of the guide covers visual product discovery as it works now, through Google Lens and Google Images.

What Is a Visual Search Tool?

Ordinary search starts with words. A shopper types "tan leather handbag" and hopes the results match the picture in their head. Visual search turns that around: the shopper starts with an image, from a friend's photo, a magazine, a screenshot or a camera view, and asks the search engine to identify what is in it and find similar things to buy.

This suits the way people shop for many products. They often cannot describe exactly what they want, but they recognise it the moment they see it. A tool such as Google Lens detects objects in the image, such as shoes, sunglasses, furniture or a lamp, and returns matching or similar products. For a retailer, that is a route into your catalogue that does not depend on the shopper guessing the right words.

A photographer kneels to shoot handbags, shoes and sunglasses on white paper as an artisan adjusts one.
Clean, well-lit product shots on a plain background are the easiest for visual search to recognise and match.

Similar Items in 2017: Where It Started

Google announced the feature this post was first written about on 10 April 2017. Its Webmaster Central blog post said Similar items used machine vision to identify products in lifestyle photos and show matching products, with information such as price and availability. At launch it covered handbags, sunglasses and shoes, with other apparel and home and garden categories planned, and it worked on mobile browsers and the Google Search app for Android.

To be eligible, retailers needed schema.org product markup on their product pages, with name, image, price and currency, and availability. That requirement is the lasting lesson: the system could only show a product it understood, and it understood products through clean images plus structured data. Both still matter today.

How Image-Based Product Discovery Works

Behind the simple experience sits a chain of steps. Knowing them helps you prepare for each one.

  • Recognition: the system analyses the image and picks out the distinct products in it, even in a busy lifestyle photo.
  • Matching: it compares those objects with a large index of product images and listings to find visually similar items.
  • Enrichment: matched products are joined with product data such as price, availability, brand and seller, so they can be shown as shoppable results.
  • Presentation: the shopper sees similar items with prices and the sites that sell them, and can tap through to buy.

The cleaner your product photos and the more complete your product data, the easier it is for the system to recognise, match and show your items. Our guide to how visual search works and how to optimise for it covers the technology in more depth.

Where Visual Search Shows Up Today

What began as a Google Images feature has spread across Google's products. In October 2024, Google said Lens was used for nearly 20 billion visual searches every month, that 20 percent of all Lens searches were shopping-related, and that Lens drew on more than 45 billion products in its Shopping Graph. The same post said Lens could show a product's price across retailers, current deals, reviews and where to buy it.

You will meet image-led discovery in several places:

  • Google Lens: point the camera, or upload a screenshot, to identify an object and see where to buy it.
  • Google Images: Google's product structured data guide says product information can appear in Google Images and Google Lens, including annotated images of products for sale.
  • Marketplaces and social apps: several shopping and social platforms offer their own camera or image search inside their apps.

The common thread is that a picture has become a search in its own right. For an online store, preparing for it is part of normal product-page SEO; our eCommerce SEO guide covers the wider product-page work.

You cannot control how Google ranks results, but you can make your products much easier to recognise and match. These practices follow Google's own guidance.

1. Use sharp, well-lit product photos

Google's image SEO best practices note that high-quality photos appeal to users more than blurry, unclear images. Show the product from several angles, on a clean background, and also in a realistic setting. Uncluttered photos are easier to match than dark, busy or heavily filtered ones.

2. Write useful alt text and file names

The same Google page calls alt text the most important attribute for giving more information about an image, and gives a simple file name rule: a descriptive name such as my-new-black-kitten.jpg is better than IMG00023.JPG. For a product, that means names such as tan-leather-tote-handbag.webp, and alt text that describes the item in plain words without stuffing keywords. Place images near the text that describes them.

3. Add product structured data

Mark up product pages with name, image, price, currency, availability and brand. Google's product structured data guide separates product snippets, for pages where people cannot buy directly, from merchant listings, for pages where they can. Our guide to schema markup for Indian businesses includes product examples.

4. Keep a healthy product feed

Google recommends providing structured data on your pages along with a feed in Google Merchant Center, where you can also opt into free listings. Keep prices, stock and images the same on your site and in the feed, so the data Google sees in each place agrees.

5. Use fast, modern image formats

Google's image guidance lists the formats it supports, including JPEG, PNG, WebP and AVIF. Compressed, responsive images keep product pages fast without losing visible quality, which helps both shoppers and crawlers.

For more on image search results in general, see our older guide to image SEO and rich visual results.

Both matter, and they work best together. The table below shows how they differ.

AspectKeyword searchVisual search
InputTyped words and phrasesA photo, screenshot or camera view
Best forKnown needs ("running shoes size 9")Inspiration and hard-to-describe items
Key signalsPage text, titles and linksImage quality, product data and feeds
Shopper mindsetResearching a specific need"I want that. Where can I buy it?"

Treat visual search as one more way shoppers reach you, not a replacement for keyword and content work. Our guide to visual search optimisation on Pinterest and Google Images covers the image side across platforms.

Which Products Gain Most

Products bought on look and feel gain most: fashion, footwear, eyewear, jewellery, furniture and home decor. Shoppers in these categories often see something first and look for it second. Products bought on specification, such as spare parts or software, gain less, because people already know the words to type.

For Indian online stores and D2C brands, the practical step is to treat every product image as something a shopper might search with: clean photography, accurate data and fast pages. When that traffic arrives, the store still has to convert it; see why traffic doesn't always convert into leads for the common gaps.

Common Mistakes to Avoid

  • One photo per product: a single angle gives the system less to match against.
  • Generic file names: names like image1.jpg tell crawlers nothing about the product.
  • Missing or stale prices: product data that disagrees between your page and your feed undermines trust in both.
  • Images loaded only by script: pictures that appear only after heavy scripts run load slowly and are harder for crawlers to find.
  • Keyword-stuffed alt text: Google warns against it; describe the product plainly instead.

Conclusion

Google's Similar items in 2017 was an early sign of a shift that has since grown into Google Lens and shoppable image results. The work for a store has stayed the same: sharp photos from several angles, descriptive file names and alt text, accurate product structured data, a feed that matches your pages, and fast images. Do that consistently, and your products are ready for shoppers who search with a picture. For how AI-led search is changing discovery more widely, see our piece on how businesses can prepare for AI-driven search.

Frequently Asked Questions

What is a visual search tool for shopping?

It is a tool, such as Google Lens, that lets you search with a photo, screenshot or camera view instead of typed words. It recognises the products in the image and returns similar items you can buy, often with prices, reviews and the stores that sell them, so you can find something you cannot easily describe.

What was Google's Similar items feature?

Google announced Similar items on 10 April 2017. It used machine vision to find products in lifestyle photos in Google Images and show matching products with price and availability. It started with handbags, sunglasses and shoes on mobile browsers and the Android Google app, and relied on retailers' schema.org product markup.

How can I optimise product images for visual search?

Use sharp, well-lit photos from several angles on clean backgrounds, give files descriptive names, write plain alt text, add product structured data with price and availability, keep your Merchant Center feed matching your pages, and serve compressed modern formats such as WebP. These follow Google's own image and product guidance.

Does visual search replace keyword SEO?

No. Visual search is an extra way shoppers find products, and it works best alongside keyword and content SEO. Typed search still handles specific needs, such as a size or a model number, while visual search helps with inspiration and items that are hard to describe. Most stores need both to be in good shape.

Which product categories benefit most from visual search?

Categories bought on look and feel benefit most: fashion, footwear, eyewear, jewellery, furniture and home decor. Shoppers often see these items in a photo or a shop before they know what to call them. Products bought on specification, such as spare parts, benefit less, because buyers already know what to type.

Share this article: WhatsApp LinkedIn

Want help with this?

Talk to our team about SEO services.

Read next

Comments

Be the first to share your thoughts on this article.

Leave a Comment

Your email address will not be published. Comments are moderated before appearing.