Salesforce World Tour DC - understanding Einstein GPT

Parker Harris goes to Washington

At last week's Salesforce World Tour DC, Salesforce announced new Einstein GPT and Data Cloud capabilities for Flow, Salesforce’s business process automation solution. It also announced new Data Cloud partner applications and expert services partners, and a new professional services offering for Einstein GPT and Data Cloud. On the industries front, it announced new public sector cloud capabilities.

The company also laid out a roadmap for how customers get from where they are today to GPT-enabled automated workflows, and – no big surprise – it all starts with Salesforce Customer Relationship Management (CRM). As Salesforce Chief Technology Officer Parker Harris laid out in the keynote, customers start by capturing sales, service, marketing, and commerce data in CRM, integrate and harmonize it using Data Cloud, and then apply Einstein GPT and Flow to that harmonized data to intelligently automate processes.

For the 20 to 30% of Salesforce customers that Valoir estimates have not yet moved to Lightning, the Lightning migration will clearly be the first step, as only Lightning users can take advantage of capabilities such as Einstein and Flow, let alone GPT. For many customers, the hesitation to move to Lightning has been the amount of customizations they’ve built into their existing Salesforce instance, and the cost of rebuilding that code in the latest Salesforce environment. As Lightning feature parity has grown, that hurdle continues to drop, and the potential of Einstein GPT to take on a big portion of that coding burden will drive it down even further. 

However, because the move to Lightning is a re-architecting effort, it can potentially open customers up to considering other vendors’ alternatives, which explains why Salesforce hasn’t aggressively pushed the Lightning conversation with customers that are still on Classic Salesforce. The capabilities of Einstein and Einstein GPT, as well as the other benefits of moving to Lightning, will make Lightning compelling for more customers.

What is it? 

However, the other hurdle Salesforce - and other generative AI vendors - face is helping customers to understand what Einstein GPT is, and the potential problems it solves. At World Tour, many customers were having trouble getting their heads around what Einstein GPT is and how it differs from ChatGPT. To be clear, Einstein GPT delivers results based on both public data and data in a customer’s Salesforce CRM system, and can use ChatGPT or another generative language model to generate its output.

When it comes to Flow, the story is more straightforward. The DC crowd was reminded that that Salesforce was one of the first vendors to provide low-code process automation, starting with its sales force automation (SFA) 25 years ago, and its business process automation capabilities have evolved from basic workflows to support intelligent, integrated automations with Flow.

Salesforce is making big bets on Data Cloud, and its product messaging has evolved. There was no mention of Genie; instead, Salesforce has laid out the progression from CRM to Data Cloud to generative AI. Rather than 'magic', Salesforce showed how data harmonization functionally worked in Data Cloud to support sales, service, and marketing.

This is important because the choice of a Customer Data Platform (CDP) is a technology decision typically left to IT leadership. Although Salesforce has gained ground in winning over CIOs, its competitors often have stronger relationships with IT, and Infrastructure buyers don’t necessarily believe Salesforce’s incumbency as the CRM vendor makes it a natural choice for a CDP. The company will have to continue to make the case for Data Cloud, and for a CDP strategy as a natural extension of a company’s existing CRM investment.

My take

At Salesforce World Tour DC, the big news was Einstein GPT, but at the same time, the company showed it’s getting serious about Data Cloud. Not all customers are ready for generative AI today, and Salesforce recognizes that it will need to guide customers through the journey from CRM to CDP to generative AI. Laying out generative AI as the natural evolution of a customer-centric company strategy that starts with CRM and ends with AI-driven intelligent automation – with Data Cloud along the way – gives customers a more compelling reason to move to Lightning, and to extend their Salesforce investment beyond just CRM applications.