Crowdsend Launches to Use Crowdsourcing to Identify Products in Images

Companies like Stipple, Thinglink and Luminate have been trying to take photos from static to dynamic with their tools for adding information and attribution to images. Australian startup Crowdsend is taking the same idea, adding product info to images, but combining it with the wisdom of the crowd with their new crowdsourcing platform that aims to identify products within images. The company launched in private beta last week, and has 100 retailers on board with over 26 million products already tagged in their database.

Unlike Stipple, which lets publishers and photographers add information like product information and links to their images that stays updated anywhere they appear online, Crowdsend doesn’t require the image owner to assign or update information. Rather,  anyone can add product information to an image, for example adding a link to an online store where consumers can buy the item. And if someone comes across a product they want to find, they can ask Crowdsend’s community for help identifying a product, which any user can then tag themselves. It’s like a crowdsourced version of reverse search engine TinEye, which allows users to search a photo and see where else it has appeared online, except focused on product information.

That product tag would only appear on the Crowdsend website and on the original publication where the image was published, rather than dynamically updating anywhere the image appears online, though founder Tim Davis said that’s something they’re considering adding in the future.

Davis said that while Stipple focuses on letting photographers and companies make sure their information stays with a photo anywhere it appears online, Crowdsend is more focused on the end consumer who comes across a photo and wants to identify a product. Users can tag items they know in images using the bookmarklet, and can also use it to ask the Crowdsend community to add tags if they want to identify a product. Users can like or ‘want’ products, leave reviews, and request for a tag to be fixed if it’s tagged with the wrong product. As they tag images users earn points, and the company is aiming to work with retailers to let users redeem their points for real-life rewards.

“What we’re trying to do is essentially create a place where if you visit a site and you see an image and you think to yourself ‘I’d love to know what that is,’ what we want to be able to do is allow you to ask that question, and allow other people to suggest what it could be,” Davis said in an interview.

Crowdsend also works with publishers, who can add a line of Javascript code to their website to tag images on their blog or publication, or allow visitors to tag images, which they can approve or deny. Every tagged photo gets its own page on Crowdsend, and all the products tagged to a certain brand are aggregated to a separate page. Davis said they’re really trying to target brands and site owners, since they can provide analytics for site owners on which images from their site have been tagged, and for brands on everything from which of their products have been tagged to who is viewing any given product.

“What we’re trying to do is create brand pages, site pages, image pages, and product pages, so what it means is when you link all that information together, you can look at a Gucci bag, and which site it comes from…or go to the product page and see all the images the product has been featured in,” Davis said.

Davis said that while right now they focus on products, eventually they could branch out to people, places, and other points of interest in a photo. The tool is free for users and publishers, and right now they make money through affiliate links on tagged products.

While identifying products in images and providing information on where to buy is a value-add for brands and consumers, Crowdsend will need to answer the “what’s in it for me?” question for both sides. Analytics for retailers, not to mention affiliate traffic, will likely be enticing, and for users the promise of rewards in exchange for tagging images will likely work if the rewards are attractive enough. While the company works on ironing out bugs during the private beta in preparation for a public launch, they will need to focus on building an engaged base of users, publishers and retailers so visitors aren’t left using Google image search to identify products.


Erin Bury

Erin Bury

Erin Bury is a Co-founder and CEO at Willful, an online estate planning platform. Also a former Managing Director at Eighty-Eight, a creative communications agency based in Toronto. She was formerly the Managing Editor at BetaKit. Follow her on Twitter at @erinbury.

0 replies on “Crowdsend Launches to Use Crowdsourcing to Identify Products in Images”