3 New Einstein ‘Platform Services’ Released

Share this article...

Almost a year ago, Benioff revealed his new creation to the world. That creation, was Salesforce Einstein.

Salesforce Einstein kickstarted a mission to change the face of CRM and take it to the level: from reactive to predictive[1].

Although Artificial Intelligence (AI) wasn’t a new concept, what was revolutionary was the possibility of AI becoming mainstream in enterprise software, a “scalable and cost-effective approach for incorporating advanced AI capabilities to your mainstream operations”[2]. These AI capabilities of the Einstein brand, together form ‘Einstein Platform Services’. Being able to take analysis at such a granular level, and roll it out on mass to the tremendous datasets companies are working with today, is something pretty remarkable.

As each technology piece is showcased in it’s own glory, the Salesforce Ecosystem and beyond continues to wait, engaged by the anticipation of ‘what’s next?’. For instance, last month Einstein ABM stole the spotlight – an ‘AI-powered solution for B2B Marketing & Sales that do business development in an account-focused, persona pin-pointed strategic way…’[3]. Shortly after, Salesforce announced that Airbus will become Einstein users – a flagship customer case study for sure[4].

The momentum continues to build as this week, new Einstein capabilities were unveiled which add to the evermore sophisticated “1st comprehensive AI platform for CRM”. Emphasising the fact that Salesforce is a platform, these Einstein AI technologies are available for companies to use in app development. I’m sure you can think of many use cases in your organisation that could do with a little AI injection!

Let’s take a look at the 3 new treats that have been lined up.

Einstein Object Detection

First up is Object Detection – which in short, means recognising multiple objects within one image.
In the Spring, Einstein Image Classification was released, meaning that brands can be recognised in logo or text in image format – a great addition to the Social Listening capabilities of Social Studio.

Object Detection takes it a little further though. Not only can it recognise certain objects, but it can also determine the quantity, sizes and locations of them. A unique use case for this would be in retail display and inventory management – or under its fancy, technical name: ‘share of shelf’.

Einstein Sentiment

Sentiment is opinion that is expressed, and when we frame it in the context of Einstein Sentiment, it’s the analysis of opinions weaved into text – such as social media posts.

Think about the places where you most often broadcast your feelings towards a product or service – yes, there’s multiple possible channels. Take this to scale and you have a difficult time segmenting good, from bad, from absolutely dire. Plus the headache of the double negatives that merely rating by keywords can’t rate with 100% accuracy.

What Einstein Sentiment does, therefore, is rates the opinionated texts on a scale of emotions that loosely group as Positive, Negative and Neutral. Just by doing this, you are able to add a lot more finesse to the process, including case routing and applying next best action.

Einstein Intent

If sentiment applies to opinion, then intent, on the other hand, is the indirect expression of an aim or a plan. Slightly different, but with similarities.

The similarities mean that the use cases overlap. Without running the risk of repeating myself, by knowing what your customer hints at doing, or where there expectations focus, you can effectively route leads or cases, plus of course automating the ‘next best action’ piece too.


‘Einstein Platform Services’ continues to expand, and I believe we’re still on the warm-up act. With these 3 new additional treats (Einstein Object Detection, Einstein Sentiment and Einstein Intent), we can start to have a more realistic picture of organisations using this new breed of ‘mainstream’ AI. As more functionality and use cases that are added, our confidence in the “1st comprehensive AI platform for CRM” rises, because it becomes both relatable and tangible.

2 thoughts on “3 New Einstein ‘Platform Services’ Released

  1. Avatar

    We recently ran our Einstein readiness report and one of the items listed as being a problem is “Amount field on opportunities is populated on at least 80.0% of all opportunities”. Is it possible to exclude this field from being required for lead scoring if we (as a company) don’t use it?

  2. Avatar

    I would expect/hope the Object Detection will be used in tandem with their Fields Service capabilities to bring huge benefits to Field Service Engineers. Will be interesting to see how accurate this functionality is under field service conditions.

Leave a Reply