
Details That Make Salesforce Commerce Cloud Win: Predictive Vision
- ArganoUV
- Commerce
- Salesforce Commerce Cloud

Stefan Zweig’s classic observation about Brazil — “Brazil is the country of the future… and always will be” — often feels a lot like the state of too much of modern technology. The Jetson future of flying cars is in the future…. and it always will be. (But least we have Tweets!)
Or maybe not. Sometimes, the future does arrive, and one of the ways the future is here is with Salesforce Commerce Cloud’s mind-blowingly powerful functionality known as “Predictive Vision” or “Einstein Vision” that Salesforce built for Salesforce Commerce Cloud using its Einstein platform.
Here’s Predictive Vision in a nutshell, at least one particular case of it that’s the primary use case for Salesforce Commerce Cloud:
Imagine, a potential customer of yours snaps a photo of a product, uploads it to your mobile site (SFRA, of course!) and — BAM — the site knows precisely what product it is, so it can open up that product’s page on the store, so the potential customer can buy it and become a real customer.
Snapping a photo of something and it being immediately matched to an existing product? That’s not quite Jetson’s flying cars, but it is well on it’s way. The future of online commerce isn’t coming — it’s here.
On the back-end, the way this works is using Salesforce’s aptly-named Einstein technology for its machine learning. One of the key challenges of AI isn’t particularly building it, but it’s getting good data and training it. Einstein Vision includes methods for easily uploading and training large amounts of visual content. Neural nets need to be trained, and Einstein makes training them easy. So it’s not just an AI technology, but a way to make it easy for you to teach the AI what you need to.
Predictive/Einstein Vision has other uses and applications as well, should you want to be even more creative. You can use it as a visual filter so users can narrow products into particular subsets that they like. It can be used to detect uses of your brand that are inappropriate — especially important in marketplace contexts where many products are uploaded to be sold without rigorous control. Maybe you want to analyze advertising, or social media images.
In our experience, this feature is one of Salesforce Commerce Cloud’s most under-utilized features. Why have most of us never even seen websites that offer this, given that the technology is both there and easily commercially available? (Aside from Salesforce not marketing their own awesome products as aggressively as they could have.) Few sites go through the effort to implement many of Salesforce Commerce Cloud’s best features. Sometimes due to cost and time constraints; sometimes due to creativity constraint.
Of course, creativity — in the real world — always does have price and time constraints. But, on the other hand, that is, too often, an excuse that marketers use to not stretch their creativity to the bounds. That perhaps could be one explanation for the vast under-utilization of the Einsteinian Predictive Vision. Someone is waiting to make a viral sensation in which their store lets people upload images of competitor’s products and then it recommends their product instead. Now that is a neural net to be trained — in irony. The media machine will be all over you. Use Einstein Vision to its max and get all your potential customers talking about you. What are you waiting for?
PS: UV is one of the world’s leading Salesforce Commerce Cloud specialists. Contact us to see how we can work together.