what's next

Salesforce Connections 2023: Introducing the Next Generation of AI

Salesforce made a splash by kicking off Connections 2023 with the introduction of generative AI products—Commerce GPT and Marketing GPT—extending and evolving Einstein’s predictive AI capabilities across every cloud. The core theme of Connections was CRM + Data + AI, positioning Salesforce uniquely in the generative space.

Erika McGrath
Global Vice President, Salesforce Data Cloud and MarTech, at Astound

62 percent of customers expect companies to anticipate their needs (Salesforce State of the Connected Customer Report 2022). At the same time, every brand (not just retail, but health, manufacturing, financial services, and so on) is looking to meet this anticipation in an efficient, scalable way. 

Commerce GPT powers deeply personal shopping experiences, offers, and buying journeys, and represents the next generation of Commerce Cloud. It enables brands to activate “commerce everywhere” in a more personalized, less manual way, at every touchpoint. Salesforce illustrated AI-powered sales and service apps, embedded shoppable social marketing (Instagram, TikTok), hybrid storefront, GPT-powered personalized product descriptions (with the brand’s tone of voice, launching in July), concierge, promotions, and easy “pay now” one-click checkout experiences. 

Marketing GPT enables brands to generate email content, audience segments, and journeys via AI. Today, content creation, audience building, and data management can be very manual. Marketing GPT unlocks efficiencies and automates manual processes, so that marketers can focus on creativity. For example, marketers can ask the GPT functionality to query the system (summer 2023 pilot), build the perfect target audiences, enhance segments, create email subject lines/body content, spin up personalized content at scale, or enrich messages to customers using CRM data. 

Is AI the new UI? Salesforce highlighted how AI can do much more than help brands create a website, AI can be the website. To illustrate, with the right CRM data, AI can generate product images of the actual customer wearing the product or matching a product to a location (e.g., a jacket on a person in a place they’re traveling to). AI can power chat experiences that are two-way conversations, enabling customers to shop together with the AI assistant, not just talk to it. Furthermore, “AI may be the new SEO”: while brands have always optimized their content to surface in Google Search, they must now consider how that content surfaces in Bard or ChatGPT.

Data Cloud as the foundation. The key for brands in leveraging Commerce GPT and Marketing GPT is that the AI models require clean and unified data. We recommend that brands focus first on creating an actionable data foundation through Salesforce Data Cloud to get the most out of the GPT. 

At Connections, Salesforce showcased the power of recent Data Cloud and Analytics enhancements to enable this. Marketers are facing a huge data gap. They must bring together distinct data sources for a single customer view, which often requires complex, expensive processes across siloed systems and billions of rows of data. This single customer view needs to be accessible to marketing and business teams for insights for decision-making.

Data Cloud ingests all these data sources (batch and real-time) via connectors in a few clicks, no matter where the data sits (e.g., Marketing Cloud, CRM, Google Cloud, Amazon, Azure, or POS) with zero copy (meaning marketers don’t need to move the data around), including Google BigQuery (spring 2024, which Salesforce unveiled at Connections). Data Cloud can then clean and prep that data in Data Spaces, also enabling data to be separated across brands or regions. Data is harmonized into a unified profile and then can be used in Commerce GPT and Marketing GPT to activate campaigns and personalize to segments everywhere: site, email, chat, VIP concierge, Google, Facebook, Amazon, and more. 

Through Tableau integration, Data Cloud can also deliver marketing analytics, sales insights, and performance visualizations like email or media outcomes, LTV, or propensity to purchase. Data Cloud is available across every cloud. 

Too much AI? With so much focus on AI, questions arise, like: Can we trust AI? Where does the data go? Who is creating the model? Are humans involved? Salesforce’s AI models can include the Salesforce model, a brand’s own model, or any other external model, and the models won’t co-mingle customer data with other data sources. Furthermore, the AI is here to help humans, not replace them, and there’s a human in the loop to guide the AI. To illustrate, for GPT-powered product descriptions, humans can edit as needed.

That being said, humans must always validate that AI models are making recommendations and suggestions that are true to brand values and goals, and that the data is correct, structured, and harmonized properly. 

Connections 2023 made it clear that AI, with human oversights, can be a new frontier for astounding customer experiences. Ready to get started?  Astound’s Data Cloud Accelerator can help brands get to market fast and take advantage of GPT. Connect with us today to learn more. 


We’d love to connect and discuss how we can power your digital transformation.
Get in touch
© 2024 Astound Digital. All Rights Reserved. Privacy Policy Terms of Use