By Robert DeSisto
Artificial intelligence (AI) has been used to beat chess and Jeopardy champions. Certainly if AI can be used to play a game as unpredictable as chess or require the world knowledge to compete in Jeopardy, it can be used for selling. Selling can’t be that much more unpredictable than chess, or require more knowledge than to compete on Jeopardy. The challenge is that creating an artificial intelligence model to be able to not just emulate, but improve on decisions humans make, to this point has been best case very hard or worst case impossible.
AI has been mostly for the select few who could hire data scientists or PhDs who are experts in deep learning algorithms to not only build the AI model and structure data, but also to tweak the model as conditions change. AI in its traditional raw form is too complicated to be a viable option for most companies, never mind sales organizations. However, if you asked vice presidents of sales if they felt they could benefit from the types of outcomes AI could bring their salespeople, not one of them would say no. That being said, they still likely have a justifiable amount of skepticism that a computer could provide selling recommendations in some cases better than their top salespeople. After all, how many computers have actually had to deliver against a quota?
Moving AI from the boutique examples of chess and Jeopardy to delivering real tangible value for every salesperson would normally be considered too difficult for 99.99% of sales organizations. To be of value for sales, AI must have the following capabilities:
- be able to discover, interpret and reason on both structured and unstructured information such as sales data, email or images;
- surface intelligence in the context of a sales application to provide value for a salesperson;
- make decisions and recommendations with precision and relevance to enhance specific sales scenarios in real-time;
- get smarter by continually analyzing new information as it becomes available and applying experiences from other similar sales situations;
- communicate and facilitate dialog in the customer or salesperson’s communication channel of choice (e.g., text, voice, email, social or collaboration tools).
It is also important that a successful AI strategy be focused on augmenting a sales practice, rather than replacing the practice or the sales operations experts and reps. Although Chatbots have emerged as a potential replacement to some buying and sales practices, salespeople are never going to disappear.
For the reasons cited above, the AI requirements to support sales have previously been out of reach. What I believe is the most significant technical development for sales since the cell phone is now here: Salesforce Einstein is the first contextual AI platform for CRM applications.
What does this mean?
Einstein leverages the most advanced AI technologies — such as machine learning, deep learning, natural language processing (NLP), image recognition, intelligent productivity, and smart data discovery — and incorporates them into a single, purpose-built set of AI CRM platform services. That is lots of technical jargon for sure, but unlike general purpose AI technology approaches where a model would have to be built and data would need to be formatted, Einstein removes both of those obstacles. Einstein delivers AI to all customer facing touch points and employees in the context of the application immediately. Einstein was not built as a general purpose AI tool, it was built for the purpose of helping companies connect to customers and enhance customer relationships. Einstein makes every customer interaction a smart interaction. There are implications across business, but in this post I am focusing on sales teams and operations — not service, marketing, commerce or analytics.
The world of sales will be forever known as BE and AE, Before Einstein and After Einstein. Ultimately, we are entering the era of the death of the AI-less salesperson. The salesperson with Einstein as a teammate will be able to perform tasks smarter and faster than is possible by human interaction alone.
As AI evolves, the opportunities with Einstein will grow exponentially. Imagine a world where a prospect — let’s call her Lisa, sends a salesperson — let’s call her Sara — an email that says deal conditions have changed, and incremental funding has been approved. Lisa also attaches a new annotated product image with changes that would need to be approved for the new funding, and wants to have a meeting in the next couple of days. Being a good teammate, Einstein reads the incoming email and realizes after reading the text that it is very important. Einstein sends a notification to Sara about the new important information. Sara knows Einstein would only notify her if it was important, because Einstein has learned how Sara likes to be contacted and when to ping her in real-time.
So Sara clicks on the notification on her phone so she can respond to Einstein using Salesforce Chatter.
Here is the exchange of Chatter feed between Einstein and Sara:
Sara: Einstein this is great news, can you set up a meeting in the next two days?
Einstein: Sara, I sent Lisa a few calendar options via email, she has agreed to meet Thursday at 3:00 PM for one hour.
Sara: Awesome, how does the new image look?
Einstein: I did an initial analysis and it appears Lisa wants a new product configuration.
Sara: Oh, boy. Do we know if we can deliver it?
Einstein: I ran the new product definition I developed by analyzing the new product image through Salesforce CPQ, and determined we can deliver the new configuration
Sara: Do you have any new competitive information?
Einstein: I also provided generated a list of five other customers who bought the same configuration in the past 12 months against the two competitors we are up against.
Sara: You’re awesome Einstein, I need to create a quote.
Einstein: Actually Sara, I already created three quote alternatives based on the last 12 months of competitive bids.
Sara: You’re awesome, Einstein!
Einstein: Not sure about that Sara, but I hope things are going well today.
Einstein also updated the sales opportunity with the new deal amount and meeting notes for the sales person (I wonder what salesperson would not like that). When Sara got home for day, she thought back to B.E. (Before Einstein), and remembered how difficult it was to make her number. For a split second she thought about getting Einstein a gift card, but then realized, Einstein would not have much personal use for it.
With Einstein the world of selling has changed forever — finally we have AI for the rest of us.
Originally published at www.salesforce.com.