Winning the AI Wars with Generative AI and Salesforce

There’s a lot of conversation around generative AI, with Salesforce announcing their Cloud AI launch. After years of development, generative AI builds on core Salesforce platforms. For businesses running on Salesforce platforms (e.g. commerce, CRM, CPQ), this video discusses how to understand the opportunities now opened up. From first-party data ownership to data management - understand the data systems and how they drive eCommerce growth. 


Hey, everybody, Rohit Garewal here, CEO of Object Edge. Super excited to share some thoughts with you today. This is gonna be kinda off the cuff, unstructured. 

I just got back from the Salesforce Connections conference and I'm really pumped up and excited about what I saw. I'm excited about what I saw around AI coming in a way that's gonna be made accessible to businesses all over the globe and AI specifically on platform. 

And when we're talking about AI on Salesforce and what they were talking about at the conference, it's really, they were talking about generative AI, specifically the ones provided by OpenAI, which we all know as ChatGPT, and bringing the power of that language model to the Salesforce platform with the Salesforce data. And I think there's so much there. 

Benefits for Salesforce Customers with Generative AI

There's so much opportunity for customers who have Salesforce to really benefit from this that I just wanted to get some ideas out there. So when I got home, two big questions were on my mind. 

Number one, how are we actually gonna leverage the powers of generative AI now that they're being brought onto platform to actually benefit our businesses? 

The second was, once I figured that out, how exactly am I gonna leverage it, right? And what I mean by that is once I figure out how generative AI and GPT can benefit my business, what is it that I need to do to make it work? Is there anything special I need to do or is it just gonna work “automagically”? 

And I think for the most part, it's gonna work as advertised. That once you've got your data on platform, it's gonna do a lot of the magic. And the GPT language model is gonna be able to provide you a lot of stuff around Commerce GPT and Service GPT and Sales GPT and Einstein in general. 

The Challenge of Differentiation with Generative AI

But the real question I was thinking about is then how are people gonna differentiate from one another? If every single person or every single company that has Salesforce has this power, then ultimately, then nobody has it, right? I mean, if it's all on an equal battlefield, how are we going to differentiate ourselves as organizations to leverage this, to beat the competition and capture market share? And so I don't have the answers all figured out but ultimately, I wanna put forth some ideas and that is really the thesis statement, which is generative AI models are only as good as the selective data and unique data that you're able to pump into it. 

That's the thesis statement. The more unique datasets that you can pump into the Salesforce platform for the GPT model to do their work, the more of an advantage that you're gonna have. And then what I realized is this is just an extension of something I've been saying for a while now, that the value of first-party data in an organization is only gonna increase. 

And with AI, it's increasing faster than I even thought. So let's dive into this. I'm just gonna share some thoughts and then hopefully, there'll be enough compelling stuff in here that it'll stimulate conversations with me and with others within peer groups so we can get this conversation moving forward because no one's figured this out yet. Once you’ve had a look, drop me a line here: 

The world's moving at a really fast pace right now. And all we can do is share this kind of ideas and knowledge and hopefully, it leads to something really, really productive. All right, so I'm gonna be talking about this from the standpoint of simplification. And that is I like to think about this from the standpoint of an experience. 

Understanding B2B Experiences

And when I talk about an experience, what I'm really talking about are experiences between buyers. You've got buyers in a B2B or in a B2B scenario, and obviously, you've got sellers. This could be distributors, this could be end customers, this could be sales reps, this could be distributors. Someone's a seller in a transaction, someone's a buyer. And every time these sellers and these buyers are interacting, I call those experiences. So that's what we're talking about. 

Every interaction point here is an experience, a buyer is having experience, a seller's having experience. 

And then these can be analog or they can be digital. For the first part of this conversation, let's just talk about digital experiences. So if you're talking about a digital experience, there's really two ways of looking at how digital experiences become compelling for you. 

The Data that Powers Your Experience

First off is you're gonna be talking about data and experiences. You've got data that powers your experience and the exhaust of those experiences, that clickstream is there's two types of data that is the exhaust of this clickstream. 

There's data that's gonna power your experience. This is gonna be all the master data in your organization. When we're talking about master data, we're talking things like: 

  • Customer
  • product
  • price
  • inventory

So the master data in your organization is gonna power your experience. And then as people are clicking around in that digital clickstream, whether it's your CRM or your employees are clicking around your CRM, whether your distributors are clicking around your portals, whether your customers are clicking around in your eCommerce sites, the exhaust of all that clickstream is number one, transactional data. 

And this is going downstream to your fulfillment and billing systems to actually execute on the order. And then there's behavioral data and that unfortunately oftentimes just goes away into the ether. 

The Data that Powers Your Decisions

So what are we talking about here? You've got data that powers your experiences and fundamentally, your transactional and behavioral data is data that powers your decisions. Two types of data in your organization. Data that power your experiences, data that power your decisions. When we talk about data that power your experiences, we're talking about the master data and to be fair, transactional data also powers your experience because that's gonna be going downstream to your fulfilling and billing. 

And then when you're talking about data that powers your decision making, once again, transactional data will help you power your decision making but also your behavioral data. So data that powers your experience, is your master data and your transactional data. And the data that powers your decision making is your transactional data and your behavioral data. All right, pretty simple, right? 

So you've got these experiences, you've got these buyers and sellers engaging in digital experiences across your CRM, across your experience, your portals, across your distributor portals, across your customers and your eCommerce, across your service reps and service cloud and maybe your order management. These are all digital screens. And these digital screens, as people are clicking around with them, as your sellers are clicking around with them, as your distributors are clicking around with them, as your buyers are clicking around with them, they're generating first-party transactional data and first-party behavioral data. 

The behavioral data oftentimes is going off into the ether and not being captured or not being leveraged and the transactional data is being passed very transactionally to your fulfillment and billing system so that the customer and the distributor experience can continue but then you've placed an order. 

Leveraging AI on Your Salesforce Platform

All right, now let's talk about the AI piece on this. So now when we're talking about AI, what are we really talking about? Well, as I mentioned, all these experiences are sitting on top of a platform. In this case, we're talking about the Salesforce platform but then we're talking about Salesforce AI. 

So they're sitting on top of your CRM, they're sitting on top of your commerce, they're sitting on top of your service instances, they're sitting on top of your OMS, maybe they're sitting on top of, I don't know, I'm just gonna throw CPQ in there, right? These are all screens that Salesforce provides that fundamentally power these digital experiences and all this data that we're talking about, this master data to power those experiences are really being forced onto platform, right? 

So when you build integration, you're building on and getting all this data into your Salesforce platform. Now, you're not managing the data there most of the time; you have master data management and you're organizing your custom data, your product data, your price data somewhere else but you're integrating it into your Salesforce platform so you can power the experiences. 

Then the data that's being generated, the transactional data as I mentioned is going into your fulfillment and billing system. And again, what's happening to your behavioral data most often is just kind of either going into Google Analytics, but it's not really going anywhere that's being leveraged correctly. 

Now, here we get into where we're talking about AI. If you think about it, underlying all this is Salesforce, I'm just gonna call it Einstein for now but essentially, their entire intelligent data model. So all that data that you're pumping in is sitting in platform and underneath that platform is sitting Einstein and Einstein AI and all the work that Salesforce is doing with OpenAI on the GPT function. 

So now what we're really talking about is as you plug all this data in, as this master data is getting fed in to the platform, all this Einstein GPT stuff is doing its magic and it's feeding a whole bunch of generative AI and recommendations and a bunch of cool stuff into each one of these platforms, right? You're gonna have CRM GPT that's giving you sales-specific generative AI function. 

You've got Commerce GPT that's gonna be feeding you commerce-specific generative AI function. Great, sounds amazing. And then the transactional data for the most part will be leveraged. And I'm sure in terms of, you know, the orders that have been placed but so much of the order transaction work happens off platform. All that fulfillment and billing work that typically happens is typically in most organizations happening in ERP Salesforce doesn't have an ERP. 

What I'm really saying then is all that data doesn't exist within platform. 

Now, here's another thing to keep in mind. For most B2B organizations, 1% to 20% of all transactions are happening in a digital world. So people that are just getting, companies that are just getting started, maybe 1% of all transactions are happening on platform. And if you're really mature, maybe 20% are happening on platform. That means 80% to 99% of the transactions are happening off platform. 

Now, go back to my original thesis statement that the people that are gonna win the AI wars on Salesforce are the people that are able to really identify unique datasets and provide the generative AI models unique datasets.

So going back and starting up, we've got buyers and sellers and they're interacting. In order to put a finer point on it, we're calling these interactions experiences. They have analog experiences and they have digital experiences. In order to power the digital experiences, you've got data that's gonna power the experiences. That's your master data, customer data, product data, price data. 

The exhaust of those digital experiences is gonna be first-party transactional data and first-party behavioral data. The first part of transactional data is primarily being fed downstream to your ERP for fulfillment, billing, et cetera, and service. And then your behavioral data is usually going into Google Analytics. 

Some people that are really mature are putting it into a data cloud and leveraging Tableau on top of it. But most companies are not at this point. 

How Cloud AI and Salesforce Einstein Drive Differentiation

Now, as you're pumping your master data into the Salesforce platform to power the experiences, Einstein is gonna do its magic. Einstein is gonna start providing you generative AI capabilities across your clouds. But Company A, Company B, Company C, they're all doing that. Any company that's powering any experience on Salesforce with their master data, they're gonna be winning against companies that don't have Salesforce because they're not getting all that generative GPT stuff unless they build it themselves. 

But if you have 10 companies that are all leveraging Salesforce in the same space, well, then they're all gonna have access to the GPT and then they're all gonna have it. And then what's the real differentiation? 

Because everybody is going to have, everybody is going to have master data. They're gonna have their own customer data, their product and pricing being fed in. So here's the point. We have to get really good at feeding the analog off-platform transactions into Salesforce for Einstein to do its work. 

That means essentially getting and organizing the data that's outside of the platform, that order data, that customer data, that transaction that's happening off platform today. It doesn't matter if it's on Salesforce or not. It could be on ERP, it could be on Oracle, it could be on any number of different areas that you're powering experiences and you're powering orders through. It could be analog. 

In fact, most of it's gonna be analog, right? 

Your sellers and your buyers working with one another and then filling out paperwork and getting orders of your ERP. The companies that can quickly and efficiently get all that order data and behavioral data that's being generated on these experiences back onto platform, they're the ones that are gonna win. 

And that doesn't mean that you have to take this heavy lift. It just means having the right data governance. It means having the right data integrations and having a plan on how you're gonna be leveraging the Salesforce platform. 

And because you've already paid for those licenses, how are you going to leverage the Einstein that you're already paying for, the GPT that you're already paying for? Well, the way you're gonna leverage it is by feeding it unique datasets, which are unique to you. So that means you have to capture first-party data, you have to store first-party data and you have to be able to feed first-party data back on the platform and that's gonna give you the lift because you're gonna have a factor of four, in some cases, a factor of 20 times more data than your competitors on platform because you're going to have 80% to 99% of your orders coming off platform for a while. So getting those orders on the platform is gonna be really key. Anyway, that's what I want to talk about. Hopefully that all made sense. If you have any questions, definitely let me know and hopefully this will get the brain thinking on how you can leverage data to win the AI wars. Thank you.

About the Author

Blue dotted circle

Rohit Garewal


Rohit is a forward-thinking eCommerce evangelist, especially focused on re-energizing the B2B sector and merging the old disciplines with new technology opportunities. He is passionate about delivering profitable growth through people-driven digital transformation. Watch his talk on digital transformation.

Latest Posts

Cropped photo of a man using tablet device

Looking for help?

We're here for you. Schedule a quick call.