Let’s Get Visual: How Search Is Changing [Infographic]
It's the perfect digital storm. Over the past few years, technologies such as smartphone cameras, image recognition software, and machine learning tools have all rapidly improved at the same time, leading to the emergence of something truly transformative: visual search.
What Is Visual Search?
Visual search replaces the traditional input for searching for things online—text—with images. Essentially, instead of typing "tulip" into a box, a user submits a photo of a tulip to the platform.
It starts with a visual that can come from anywhere, including:
- A image/stream taken in real time by a camera on a device
- An image that's already in a device's photo library
- An image pulled from somewhere (website, email, etc.)
Next, the platform must identify and interpret, which involves:
- Recognizing what the image is
- Determining what the searcher wants
- Delivering the most relevant results
Why Is Visual Search Suddenly a Thing?
Using visual search feels like magic: simply share a photo and the search engine quickly figures out what it is and what you want. This ease masks the fact that the task is immensely complex.
The technology is just becoming a thing because up until recently, it wasn't possible to execute well. To make visual search as powerful and accurate as text search, these elements had to come together:
- The image: The widespread of adoption of smartphones with good cameras and speedy Internet access has made it possible to consistently capture and share high-quality images.
- The data: Search platforms need to be able to recognize what is in an image, which requires accessing and processing huge libraries over time.
- The intelligence: For visual search to work, the machines must be be smart. Advancements in AI and machine learning have made it possible for platforms to learn more quickly and accurately.
What's So Great About Visual Search?
Visual search matches what we see with the vast amount of information available on the Internet. That makes it possible to do a number of useful things, including:
Answering questions that can't be asked via text: Sometimes we can't type in a search query because we don't know how to describe what we're looking at. Visual search is great for answering questions such as:
- What is this?
- What looks good with this?
- Where can I buy this?
Linking the physical and digital: Often there's a disconnect between what people see in the real world and all the information available online. Visual search can bridge this by:
- Enabling consumers to match products they see with online information/purchasing
- Utilizing barcodes and QR codes to link products and services with digital content
- Making shopping experiences truly interactive with visual cues and prompts
What Does Visual Search Look Like in Action?
Although visual search is relatively new, it's already been integrated into quite a few products. In particular, a number of the major online platforms have focused on the technology:
- Google Lens: This extension of the search platform enables users to point their device's camera at any range of things—from a hat to a monument—and get relevant results.
- Pinterest Lens: Share a photo of something with the platform, such as a necklace or cake decoration, and this tool will deliver Pinterest Pins of similar products.
- Amazon Rekognition: This platform powers a number of visual search tools, including Snapchat's, and connects images with products.
- Find It On eBay: Share an image from social media with eBay and the company's mobile app will find listings on the service for similar products.
How Popular Is Visual Search?
Are consumers actually using visual search? Is the technology part of the mainstream? The data shows that it hasn't made it there yet, but that adoption is quickly accelerating.
Most people still don't know what visual search is:
- Only 27% of US Internet users are aware of the technology
It currently accounts for just a fraction of all searches:
- Less than 1% of all searches are currently visual based
But its use is on the rise:
- The number of monthly visual searches on Pinterest jumped from 250 million in 2017 to more than 600 million in 2018
What Can Marketers Do to Optimize for Visual Search?
While visual search is still gaining in popularity, there are some things marketers can do now to optimize for the technology:
- Invest in images: If you want visual search engines to recognize your offerings, then it's necessary to have a robust library of images that can be accessed.
- Connect and catalog: Ensure that your images link to the correct product/service pages and that your site is structured in an easily crawlable way.
- Think of the physical and digital as intertwined: Consider how you can structure things like in-store displays to cater to those who may use visual search.
Finally, keep in mind that this is only the beginning. Consumers are just becoming familiar with the approach and the underlying technologies are still improving. Expect visual search to become much more popular and much more powerful in the near future.