Bazaarvoice + Curalate | Connecting curiosity to commerce Learn More

< Blog home

visual marketing

Introducing Intelligent Product Tagging: Artificial Intelligence Meets Commerce

Image by Brittni Albright.

Image by Brittni Albright.

Even before HBO’s Silicon Valley, the “startup girl/guy” was a tired trope. Startups today are diverse enough that there are few stereotypes that can be fairly applied to all of us. But if there is one constant across all startup founders and team members, it’s that we’re all makers. At the end of the day, we all love building things.

In the nearly five years since we started building Curalate, our product teams have never stopped shipping high-performance tools for our clients, innovative commerce experiences for their customers, and a steady stream of improvements to our underlying platform and infrastructure.

Not everything that we’ve been working on, however, has immediately seen the light of day. We recognized early on that there were some truly hard problems that we would need to solve in order to fully realize our vision — powering visual discovery at scale. Hard problems can’t be solved overnight. Instead, they require brilliant people, ample time, and a healthy pinch of good luck.

Today, we’re excited to start taking the wraps off of just such a major innovation, five years in the making. It’s called Intelligent Product Tagging — technology that can analyze an image and use machine learning to identify the products depicted within that image. For example: If you have a photo of a woman wearing a black jacket, our technology can identify that jacket, then visually match it with the corresponding product in a brand’s catalog — then make the image shoppable.

For our clients, Intelligent Product Tagging means they’ll have the ability to upload imagery that Curalate will attempt to automatically “tag” with the appropriate items in their product catalog. That cuts their workload dramatically, allowing them to approve the automatic tags with one simple click.

With Intelligent Product Tagging, we can match much more than product imagery. Let’s say you’re a men’s fashion brand and you see a man wearing one of your blazers in an Instagram post — our technology can still visually match it with the corresponding product. The same is true for all kinds of lifestyle content created by customers, bloggers, influencers or partners.

We expect to start rolling out Intelligent Product Tagging to our apparel brands in early 2017.

While we love to save our clients time, the real value of Intelligent Product Tagging is that it will allow brands to rapidly increase the size of their intelligent content library without needing to spend more time in the process. Great visual content is the cornerstone of modern commerce experiences — you can never have too much.

We’re incredibly proud to release this innovation, but we don’t consider Intelligent Product Tagging as the final result of five years of research and development. In fact, we’re more excited than ever about the potential future applications for these underlying technologies. Some of those include:

  • Powering customer-facing visual search.
  • Showing related products when a certain item is out of stock.
  • Helping to find cheaper versions of high-end products.
  • Transforming recommendation engines to suggest complementary products based on previously curated looks.
  • Helping brands increase catalog coverage with editorial imagery.

This technology is the result of a major investment by our research engineering and product development teams. It combines a slick user interface with cutting-edge deep learning techniques, all powered by the proprietary data we’ve been collecting on behalf of our clients for years.

Image recognition for consumer goods is a long-running challenge and we know we’re not the only tech company working on it. But our partnerships with 850 of the world’s most-recognized brands sets us apart. The proprietary data that we’re collecting on their behalf gives us a unique ability to visually match images against their catalogs. As our image library and partnerships grow, so will our product matching ability.

Of course, no artificial intelligence is perfect (at least in 2016). If we’re unable to automatically identify the brands’ products in the image, our clients can still have a chance to manually tag products and, if necessary, search through their product catalog. Even in these cases, we believe these manual processes will represent a substantial productivity improvement for our clients.

At Curalate, we are committed to playing the long game. We never cease to be awed and amazed by the innovation we see in the digital marketing space in general, and from our clients in particular. With Intelligent Product Tagging, we’re proud to contribute back to the space, and hopefully make possible some incredible commerce experiences in the future.

Curalate powers consumer discovery for the world’s smartest brands — turning pics and videos into portable, actionable content assets. It enables consumers to purchase your products wherever they discover your brand — social, ads, blogs or anywhere else. Get a demo now and learn how we can increase your average order value, return on ad spend and unique site visitors.

See how Curalate can help your business grow.

More stories to nerd out on:

Stay up to date with Curalate