Salesforce’s Social Studio can now see
The company is adding Einstein Vision to the social management tool in its Marketing Cloud, allowing searches for logos, scenes, foods or objects.
Salesforce is today giving eyesight to its Social Studio.
Social Studio is the social management tool in the company’s Marketing Cloud and, before today, it did not have image recognition. Brands either manually searched images or conducted text searches of metadata or related posts.
Now, it has Einstein Vision image recognition, powered by the company’s Einstein layer of artificial intelligence. Vice President of Product Marketing for Social Products Rob Begg told me that Einstein Vision previously existed in some parts of the company’s Service Cloud.
Brands can search Twitter for images containing 2 million kinds of logos, 60 types of scenes (such as beaches or airports), 200 types of foods (e.g., cheeseburger) and a thousand objects. Begg said support for additional social platforms will be added, and searches can be conducted for combos, like logos in airports.
A brand that has sponsored the backdrop at a music festival, he suggested, might want to search for all the photos on Twitter where its logo appeared. Or a restaurant chain might want to see what foods are popular on Twitter before adding to its menu.
While social media tools once offered text-only searches, increasingly sophisticated image recognition is now popping up all over.
Boston-based social intelligence platform Crimson Hexagon, for instance, recently announced that it has gone beyond the logo recognition provided by many social tools and can now identify faces, scenes, objects and activities. User-generated content platform Chute offers logo, object and scene detection, influencer platform Linqia has photo analysis, and influencer tool Dovetale does image matching.
Begg said that one distinction for Social Studio’s new eyesight is that the results are immediately actionable. A brand, for instance, can call up images of cheeseburgers and then send them to a marketing team or respond to the users.
Opinions expressed in this article are those of the guest author and not necessarily MarTech Today. Staff authors are listed here.