Salesforce Generative AI Expands Integration in Sales and Service Clouds

At the Salesforce World Tour event in New York City, Salesforce today detailed the integration of Einstein GPT into both Sales and Service Cloud. In this quick post I’ll cover the news And then I’ll have a few comments about the sticky wicket of data privacy and Salesforce generative AI. I conclude that Salesforce must swiftly deliver on its promises, or risk losing ground in the competitive enterprise AI market.

Sales and Service Cloud Announcements

Salesforce announced new AI features for Sales Cloud, including Einstein GPT integration, real-time data, Tableau for Sales, Activity 360, and native Sales Planning. These enhancements aim to optimize selling opportunities, enable faster outreach, improve lead qualification, and provide AI-driven sales coaching in multiple languages.

For Service Cloud, Salesforce introduced Einstein GPT and Data Cloud capabilities to streamline common service requests and power service interactions. Einstein GPT for Service personalized customer experiences, empowers agents, and unlocks organizational knowledge. It does this by generating service responses, automating knowledge articles, and auto-generating case summaries. The Service Catalog provides a curated library of flows to automate the fulfillment process for cases like parts replacements or product returns.

Here is a summary of the products and services announced today:

• Einstein GPT for Sales: Faster outreach, better leads, AI-driven sales coaching
• Tableau for Sales: Data exploration, business insights, revenue growth strategies
• Activity 360: Unified, real-time sales activity data from CRM and third-party sources
• Sales Planning: Efficient growth planning, allocation of capacity, territories, quota, compensation
• Pay Now and Reorder Portal: Simplified repurchasing and embedded payment capabilities
• Einstein GPT for Service: Personalized customer experiences, agent efficiency, automated service responses, knowledge article creation, case summaries
• Service Catalog: Curated library of flows, automated fulfillment for common service cases like parts replacements or product returns

Einstein GPT for Sales is currently in a closed pilot, while Tableau for Sales and Activity 360 are generally available. Sales Planning, Pay Now, and Reorder Portal are expected to be available in June 2023. Einstein GPT for Service and Service Catalog will be in pilot and generally available, respectively, in the summer of 2023.

Enterprise AI is Salesforce’s Game to Lose

This week, concerns about AI arose from Bloomberg’s report on Samsung shutting down ChatGPT-4 usage due to a code leak. This highlights customer expectations for security in large language model (LLM) chatbot products.

Salesforce, already trusted for data safety, has an opportunity to excel in enterprise AI. Many customers already use pre-GPT Einstein services. However, Salesforce must act quickly amid the rapid growth of AI tools and services. This is especially true for engineering teams coding in Salesforce languages. Some teams might not wait for Einstein GPT for Developers. They might turn to non-Salesforce AI tools, increasing data security risks for customers and placing Salesforce at a competitive disadvantage.

And speed is of the essence for Salesforce right now due to the Cambrian Explosion in AI tools and services. This is especially true for engineering teams coding in Salesforce languages. Coders are among the first to be impacted by generative AI because of their own personal experiences. Some of these gains are so dramatic that engineering teams will not wait for Einstein GPT for Developers, which is still in a closed pilot program.

This means that some teams are going to try to manage the privacy concerns themselves by working with non-Salesforce AI tools. This activity amplifies data security risks for the Salesforce customer, and it puts Salesforce at a competitive disadvantage because they need to meet new expectations for which they are not prepared.

While Salesforce should be given time to integrate customer data with generative AI, striking the right balance between customer demand and data safety remains a challenge that needs to be addressed soon.