Visual Search - A Status Update on its Latest B2B Progress

Gartner hype cycle including visual searchimage credit: Gartner

This post is the latest in our series on new technologies impacting industrial manufacturers and distributors. Earlier posts have included Pricing Optimization, IIoT, 3D Printing, Generative Design, 5G, Artificial IntelligenceDigital Marketing with CAD, Blockchain, Digital Twins and this one is on Visual Search. As we've seen with other new technologies, they are often combined to create solutions - in this case visual search uses image and search technology plus artificial intelligence.

Before getting into the adoption of this particular technology, here are a few observations about all these new technologies:

  • We all face a bewildering number of new technologies and change, we can't do everything so need to be selective in deciding. This blog post, How to decide which technology to keep up with and apply may help.
  • What sets the pace of adoption of a new technology? History shows us a repeatable pattern of invention to commoditization to ubiquity. However, it's not invention that drives adoption, it's commoditization - Karl Benz invented the gasoline-powered vehicle, but Henry Ford's mass production commoditized it much later.
  • The rate of commoditization varies. Music, movies, and media have been transformed, manufacturing is changing more slowly. Technologies like 3D Printing, Robotics, and Artificial Intelligence were invented in the 1960's but only recently became available at generally affordable price points and that has accelerated adoption.
  • The internet commoditized IT communications and networking, to enable multiple innovations including email, the world-wide-web, online search, eCommerce, social media, video calls, content marketing and more. For manufacturing marketing specifically we gained capabilities like Online Catalogs, Configurators and CAD Model and Drawing Downloads.

The future of search will be about pictures rather than keywords." - Ben Silbermann, Pinterest CEO

What is Visual Search?

Visual search is a search query initiated with an image rather than typed or spoken words. The image may be captured by a camera or other means such as a CAD drawing or model. Gartner classifies visual search as an ‘emerging technology’ that depends on AI to understand the content and context of images and return a list of related results.

Who uses Visual Search?

An Aug 2019 eMarketer survey showed that only 3% regularly use visual search, and only 10% have ever used it and 7% are familiar with the technology.

Who enables Visual Search?

Gartner's hype cycle for digital marketing (chart above) follows the hype of new technologies up to their 'peak of inflated expectaions' and usually inevitable drop into their 'trough of disillusionment' before they achieve a slower rise up the slope of enlightenment. It is a very useful model of technology adoption to help us think about new technologies. Visual Search is listed on it's way up the hype curve and artificial intelligence (AI) which it uses is listed at the top of the hype curve.

Google, Microsoft, Pinterest, Amazon and other search (and augmented reality) vendors continue to invest in visual search. Multiple platforms now use computer vision and AI to identify images or an object within an image. As Gartner analyst, Mike McGuire, put it, “Marketing success depends on the maturation of these technologies and mature data and image metadata and tagging capability on the marketing team.

Visual Search platforms already available (for example, Google Lens, Google Apps for Android and iOS, Prism, Amazon Partfinder, Pinterest) try to enhance the customer experience in eCommerce, content marketing, product and search marketing. For example:

  • Amazon Part Finder uses visual search to find partsAmazon Partfinder includes an iPhone app that lets you find parts for building and repairing things by photographing a part on a white background with a penny in the frame for scale.
  • ASOS, a U.K. online fashion retailer, offers a visual search tool, Style Match, so customers can zoom in on a man’s suit in a magazine picture, for example, and receive suggestions on similar suits.
  • Pinterest just announced that it's Lens camera searches can now identify more than 2.5 billion objects within the fashion and home industry.
  • VizSeek is targeted at the needs of engineering/CAD companies to quickly find parts or products by using a photo, 2D hand sketch, 2D drawing or a 3D model.

Conclusions

1. Visual search is much more complex than text-based search, because the image itself is the query. Visual search technology is continually improving, and usage is expanding across various platforms.

2. Most use so far is in B2C applications but as history shows what starts in B2C soon moves to B2B - as the VizSeek example shows.

3. Visual search is closely related to other visual technologies which collectively enable visual commerce/interaction with a brand’s products in a visual, immersive manner. Gartner identies visual commerce as a top trend that, "spans 360-degree video, 2D and 3D configuration, visual search, augmented reality (AR) and virtual reality (VR)."

The real challenge of all these new technologies is our society's (politics, businesses, individuals) ability to adapt to the pace of change.in the way we live and work. For example, think of how the gig economy has, and is, affecting employment.

As always, if you have comments or suggestions please make them below, click to ask us a question or call for a discussion. In the meantime we'll continue to keep you up-to-date on eCommerce and new technologies impacting industrial suppliers.

Comments (0)


Add a Comment





Allowed tags: <b><i><br>Add a new comment: