Main content

Salesforce Connections - if AI is the new UI, what does that mean for marketing and commerce?

Stuart Lauchlan Profile picture for user slauchlan June 7, 2023
It's the opening day of the Connections marketing fest and generative AI is inevitably top of the talking points agenda.


Every company has to become an AI company. Salesforce is absolutely an AI company!

That emphatic declaration from Stephen Hammond, GM of Marketing Cloud at Salesforce, shouldn’t really come as any surprise - and we’ll doubtless be hearing that again from CEO Marc Benioff himself next week at a dedicated AI Day event in New York. But before then, the company’s Connections marketing conference kicks off today and inevitably AI, and particularly generative AI, is front and center.

This is a vital topic of interest to marketers, according to data from Salesforce’s Generative AI Snapshot Research Series. A poll of over 1,000 B2B marketers across the US, UK and Australia found that over half of respondents (51%) say they are currently using generative AI, while a further 22% intends to soon.

Other stats of note:

  • 53% reckon generative AI is a “game-changer” for marketers.
  • 71% believe it will eliminate busy work.
  • 71% say using generative AI will allow them to focus on more strategic work.

But there's a but:

  • Over a third (39%) of marketers don’t know how to safely use generative AI.
  • 43% don’t know how to get the most value out of generative AI
  • 34% say they don’t know how to effectively use generative AI at work

So, that’s why Hammond says:

Our job right now is to figure out how do we create tools and technologies that will assist every brand, every company, every user in the world, to also be able to easily use and adapt to using these technologies.

Tech Connections

In terms of tech, Connections sees the official launch of Marketing GPT and Commerce GPT. As per the product descriptors:

With Marketing GPT and Data Cloud, marketers will be able to:

  •  Work smarter with Segment Creation, which gives marketers the ability to create audience segments quickly and improve targeting using natural language prompts and AI-driven recommendations based on trusted data from Data Cloud.
  •  Reduce writing workload with Email Content Creation, which can auto-generate personalized emails for improved testing and engagement.
  • Improve Marketing ROI with Segment Intelligence for Data Cloud, which automatically connects first-party data, revenue data, and third-party paid media data for a more comprehensive view of audience engagement.
  • Deliver the right message at the right time with Rapid Identity Resolution, Segmentation, and Engagement, which automatically resolves customer identities and refreshes segments in Data Cloud to help ensure up-to-date accuracy.
  •  Leverage Typeface’s generative AI content platform to create contextual visual assets for multi- channel campaigns within Marketing GPT based on specific brand voice, style guides, and messaging.


With Commerce GPT and Data Cloud, brands will be able to:

  • Automate growth and conversion strategies while maximizing merchant productivity with Goals- Based Commerce. This cutting-edge tool empowers businesses to set targets and goals, and then provides actionable insights and proactive recommendations on how to meet them.  Powered by Data Cloud, Einstein AI, and Flow, Goals-Based Commerce offers recommendations toward delivering desired outcomes, ranging from improving margins to increasing average order value (AOV).
  • Supercharge merchant productivity with Dynamic Product Descriptions that automatically fill in missing catalog data for merchants and revolutionize the customer experience with auto- generated product descriptions tailored to every buyer.
  • Redefine shopping and loyalty with Commerce Concierge to power personalized, engaging conversations and help shoppers discover products effortlessly through natural language interactions across channels spanning from digital storefronts to messaging apps.


Both Hammond and his Commerce Cloud counterpart, Michael Affronti, talk about the transformative disruption that generative AI is bringing to both marketing and commerce. Hammond argues:

Digital disruption has forced companies to think about how they create meaningful experiences for their customers at scale. Today that shift is really starting to happen at a dramatic scale. We're seeing the seismic shift really driven by the innovation of technologies related to AI. This is something that is forcing companies to think about, how do they use this AI technology to be able to help them create efficiencies, personalization, and really to build to connect better with our customers?

The revolutionary potential in marketing is clear, he says:

Traditionally, when marketers are creating a campaign, they go through a number of steps. The biggest constraints for marketing organizations are generally time and money. When we think about the steps that it takes to generate a campaign, we go through the process of first discovering what has worked historically - this includes scouring through billions of rows of data to make sure that we can understand how historical campaigns worked - and then turn that into a campaign. This is pretty tedious, and it also includes the creation of audience segments again, doing queries and deep dive analysis, which can be very manual.

Add to this, the creation of content. It can be very time-consuming, and also costly, to create content to be able to scale for personalization for these kinds of campaigns, especially when we consider that we want to create very meaningful experiences for every individual that we engage with.

And then finally, when we think about the journeys that we're trying to project for our customers, that has also been a very manual process.

Marketing GPT will change this, he states:

This allows companies to be able to create more creativity by creating efficiencies, and also scaling personalization. This system will bring together the combination of AI from our historical systems, our machine learning to be able to do personalization at scale, to be able to do audience segmentation, and combine that with natural language to help marketers build the interface with systems, using AI in a very natural way to be able to assist the way that they create and manage their content and their experiences.

A case in point might be automating email content creation, he suggests:

One of the hardest challenges when you're creating personalized content is making enough personalized content so that you can engage all the broad audience who you're trying to engage with. Coming up with subject lines and body content at scale has been a challenge. Now, using GPT, we can bring that in an assisted experience, where marketers can engage through AI to be able to create automated subject lines and email body content in a very exciting new way. This will help to create scale and efficiencies conversations.

New day

For his part, Affronti, GM of Commerce Cloud, pitches the idea of “a new day for commerce”, stating:

When we talk to our customers, we're discussing how commerce has really changed from something that used to be just for retail companies, to now being something that is for retail, and companies in every industry and every line of business. I've been working recently with one of our customers in the pharmaceutical space who, thanks to all of the regulation changes that happened during COVID, has been going direct-to-consumer for the first time. We're working with them to develop shopping on different messaging channels because what they realized is that having a standard storefront is important, but most of their customers are interfacing with those [other] channels.

This ‘new day for commerce’ is going to be built on the idea of one-to-one personalization, he argues:

Personalization in the classic form of commerce came through thinking about products and how they relate to each other and what to recommend. But now, with the power of GPT technology in our Data Cloud, we're able to develop personalization [that is] truly at a one-on-one level. And the last thing in this new day for commerce that we're seeing and hearing from our customers is that anywhere a customer wants to interface with your brand needs to be a transactable, elegant moment, where they can interface with your brand, pay for something if that's what they want to do, and have that be a continuous experience, both in person and digitally.

That’s what Commerce GPT is intended to deliver. Affronti positions this as the next evolution of Commerce Cloud:

We've fundamentally rebuilt Commerce Cloud with AI, data and our CRM technology. That's going to show up in a couple of really specific ways for our customers. The first is that we're helping our customers fuel growth of their businesses with the power of our trusted AI platform. The example that really brings that to life is, we've heard from customers, especially customers outside of retail who are going into commerce for the first time, that they don't really know the right merchandizing styles to use or the right promotions to set up to accomplish their goals.

Our goals-based commerce, as a part of Commerce GPT, allows our customers to simply type in, 'I would like to liquidate an outdated product line',  and we will understand their intent and recommend things like storefront design, merchandizing set-ups, and even promos, and then help them target past buyers, and even set up promotions inside of Marketing Cloud. So our goal is to really help our customers accelerate their ability to achieve their business goals.

He also points to a new 'Commerce Concierge Experience' as another example of AI-driven innovation:

This is an application experience built on top of Commerce GPT that delivers a sort of lifelike white glove concierge experience, using bot technology and generative technology to help our customers drive a one-to-one personalization experience when their customers reach out over almost any messaging channel. We think this truly represents the next platform for customer interaction when it comes to interacting with brands and purchasing from them.

Affronti sums up by quoting his boss:

Marc has this great line - ‘AI is the new UI’. If you're not thinking about how that's going to deeply transform your company and your business and your employees, and then the customers of your business, you truly could be missing out on this tidal wave of change coming.

My take

While some critics might argue that the generative AI hype cycle has dominated the tech agenda to an unhealthy degree this year, the potential impact of the technology across multiple business functions cannot be denied, so long as expectations are managed correctly. Barb Mosher Zinck made this point last week in her piece Why generative AI is not a replacement for marketers...whatever the cost-conscious CEO might think!

With that in mind, it was reassuring to hear Hammond make a clear statement of intent here about the limitations as well as the potential of generative AI:

It's such an important question because these technologies are still maturing. But they are advancing in such a way that we can use them today in a really substantial way. They're not a silver bullet. A human is required in the middle of this. We've designed all of our systems so that we can have the the AI or the GPT bring back its version, and then present that to the user to make a decision if they approve that or not. So, there's going to be a human in the loop where GPT is enabled.

For the same reasons of maturity/immaturity, privacy and security at the highest standards are needed, particularly where customer data is concerned. Again Hammond is emphatic:

None of the data that we have in our systems will be exchanged with outside systems; only models and descriptions of types of data will be exchanged. We've also managed to create relationships with these third party systems that we might be leveraging, in addition to our own LLMs [Large Language Models], to ensure that the only time that that information is ever even known to those external systems is at the time of the query, and then it disappears. It's a secure interaction, it's a known interaction, and it's something that we've managed to build in a way that will maintain the privacy and security and trust of our customers.

More to come on these topics at that AI Day event next week. In the meantime, Jon Reed is on the ground at Connections. Check back for more insights.

A grey colored placeholder image