Episode 1: AI Readiness for Salesforce Podcast

In the first of a series of podcasts on the topic of AI Readiness for Salesforce, Luca Benini talks to Sam Chappell, CTO at Performa IT about his experience in implementing Agentforce in EMEA.

Sam talks about the project with Secret Escapes, touches on challenges and learnings, and suggests how other organisations can get ready for AI in their Org. (Hint: it’s all about the data!)

Video transcript

So thank you, Sam, for being on the podcast. Hello, everybody. So this is our first, of hopefully a very long series of podcasts dedicated to AI readiness for Salesforce. In this series will run a few episodes. We’ll have guests helping us understand how companies culture, people technologies, Salesforce itself, are preparing for AI, with all, you know, the aspects possible hopefully to on different episodes by different expertise that is brought through, with our guests.

So without further ado, thank you very much. Sam, you are, you know, our first guest. Very welcome. On this podcast, you want to introduce yourself? Yeah, absolutely. So I’m Sam Chappell, CTO at former IT. So a Salesforce side partner and, very happy to be able to say as well that, I was the lead architect on the first implementation of Agentforce in Europe.

So that was, a really exciting, cool achievement to be part of. So, yeah, really excited to talk to you today, Luca. Well, 100%. And, you know, quite exciting and, you know, maybe self-explanatory. Why, we invited you to be our our first guest. You know, a revolution is taking place in our ecosystem. And, you know, best, to to tell us a bit more about, you know, what is going on and, you know, are we ready?

And are companies ready? You know, we ready in the ecosystem. X was ready to welcome AI in the form that is coming to life right now. So actually let’s start there. Know Sam, why don’t we, can you tell us a bit more about, you know, how you, you know, took on these, you know, responsibility, you know, how did the, you know, the conversation with Salesforce and, client, you know, take place?

And what were the steps that led to this amazing project? Yeah, absolutely. So, we’ve been working with, Secret Escapes. Who’s the client? For? Must be around about five years now. And we were brought in, you know, specifically to work on the service cloud element of the, Salesforce implementation. So we were trying to upskill, both in-house teams on that and then also bring up the functionality and really make it a best in class.

And through that journey, of around about four years, I’ve been building up the knowledge base and the interactions in there and building up the omnichannel reaching and everything that that you would expect within a service. Implementation within CRM. And that’s, Secret Escapes up to have a really good foundation for Agentforce. And I think when it, when it came out.

So, Secret Escapes were approached by Salesforce actually to become one of the, partners for the pilot program for Agentforce. What that meant is, before we all knew what Agentforce was as a product, they were being, talked to around the potential benefits. And would they like to take on this new AI functionality that was coming out?

And that’s when we got involved with the conversation there as well. So probably about six months before anyone had heard of Agentforce, we were starting to work on the project. And, it started with a desire to help, improve the menu based bots that were placed on bots that were available on the website at the time, and make that more conversational.

There are a luxury travel brand, so you know, that personal feel is really important. This seemed like the perfect use case to, to start to to look at Agentforce there. And that’s where we started to have all the conversations there. And, it was a really exciting, prospect to be working directly with Salesforce product over in the US as well, being able to, you know, actively feedback on the product, give our thoughts on what was working well, what wasn’t working well, and get real insight into how it was being built.

So yeah, I think, you know, a little bit of luck. But I mean, all the hard work that went in before it, I set them up in a place to mean that, you know, they were they were a no brainer to try. Exactly. So let’s talk about that. You know, kind of you just mentioned, you know, they were, a good target because they were set up, you know, with a lot of hard work getting to a place where they were ready or, you know, Salesforce and you guys felt, you know, there was something happening that made them ready.

So can you decline that? So, you know, we talked about service. We talked about a long history of using, Salesforce in the right way with your support, setting it up in the right way. So I would imagine there’s a component of data. There’s a component of, culture of innovation and technology. You know, give us a sense of the ingredients that made them feel like they were ready.

And Salesforce feeling like, yes, these guys should be in the pilot.

Yeah. So I think there are a couple of elements to that. As you’ve alluded to. The first one would what I would always, would, always would say would be the culture. So there was a real drive for innovation.

And that is something that I think is really key when looking at new technology. Is having an open mind, you know, it’s not going to be perfect the first time it goes in. But that you’ve got the ability to see where it’s going, and make sure that what you’re defining for those use cases that you’re going to start building to, is really clear, and everyone’s on board with it.

The second point, as you said, they’ve been using cases for a while. And I’m gonna talk about cases specifically for service. They had a lot of data in the, and that is that is brilliant. And there was some cleanup that needed to be done. And there was some effort as soon as, we heard about the potential for being on on the pilot, there was some work effort that went into making sure that all the data is going to be in, in, in a good shape.

But where they started to adopt Einstein features before then as well, a lot of effort had been put into thinking about now, how clean is that data? Do we have all the data we need? Do we need to go and fetch data from, you know, the internal booking, software, for example, and make sure we’ve got that information.

And there was a project that we’ve done, you know, a couple of years earlier to pull some of that booking information needed, which made a big difference in having some pre-built connections that we could, we could use to to extend what we wanted to do in the future as well. And just the other thing as well, I would say particularly for service, having a decent bank of knowledge articles, and a proper publishing process behind how they’re getting updated and funneled and pushed through made a real big difference for for getting a really quick, implementation of the sort of back and forth conversation around knowledge articles really, really quick.

We we ended up doing that in three weeks. From getting hold of the product to going live, it was three weeks. Wow. Just for the knowledge article piece. Yeah. And I think, you know, we could probably do that quicker now that we know more about Agent Fulsome. We’ve been working on that. But I think even still in the in software development terms, the three week turnaround is is pretty quick.

So a couple of very different angles that I want to pick up on. One is you said the the culture was probably the one thing that made them a good candidate to, to experiment with. So that’s a culture of innovation and experimentation. Right? So being ready to fail or to accept that is not going to work the first time.

And how how do you think they achieve that in a leadership is that, you know, years of, you know, I don’t know, is it because it’s a fun company? Because they sell travel and, you know, amazing destinations? Or was it, I think it’s a mix of of all of those. So, I think there is a real passion behind the client to, to put the customer first.

You know, they want to make sure you have an amazing journey. And I think that then naturally lends itself to thinking, well, okay, what cool new stuff can we do to help people? Which is really key. And I think that’s something that lots of companies can adopt if they if they haven’t or don’t it already. There’s a big, excitement from the board level as well, which I think, you know, the key key for any, you know, big large transformation project is having the board on board, makes a real big difference to how the messaging is is lay it down throughout the rest of the company.

And there was a there was a big excitement from the, customer service agents as well. You know, they want to make sure that people are happy. And one thing that was really key to that was making sure everyone was involved from the start. Yeah. So obviously there was a big push for innovation. From the tech side of the business, from the I.T department.

And really brilliant, sort of messaging from there and coordinating everything they needed internally. But we also made sure that, you know, as soon as we got our hands on the agent and they could see something, we got it in front of the customer services team because they are the people that are answering those questions.

So their members day in, day out. And really, they’re the people that can tell us if the answers that the the agents generating are correct or not. And I think getting everyone involved makes such a difference to making sure that people are really excited about it and not only makes projects more, more successful, it makes sense. It makes it look.

Give me a sense of, the scale of, you know, kind of, you know, how many articles are we talking about? How many, people, how many customers, how many interactions expected? You know what what was the the landscape where where Agentforce was deployed first with the, with the articles. Yeah. So, there were around about just over 200 articles already.

Okay. I think Salesforce’s recommendation is over 100, and I think very good learning. You mean over for good? Yeah, for for a good based within the model, for it to learn from it and also have enough answers to come back with across the sort of the variety of questions you might get from a customer. Yes. I think that is probably a roundabout.

Right. I think sometimes people will think maybe that’s well, we don’t we don’t have a hundred knowledge articles or we, we can’t necessarily get all that in. But I think if you look across your business, you look at your FAQs on your website, you look at all these different things. You know, a project we’ve done earlier was pulling the facts off the website and holding them within Salesforce, and that put another 100 articles in just off the back.

So I think that there are lots of quick wins to get that information on board and in there. And the other thing you can do is just talk to your agents, you know, you know, your existing employees. What what sorts of questions are you getting, you know, and pulling through. And it won’t take you long to get that that level of information.

So that was the first side to it in terms of interactions. We were looking originally, being able to answer, around about 15% of the interactions. So we’re not trying to, you know, blow away the water, and take everything on at once. What we wanted to do is if people are dropping off, we want to sort of try and make sure that we can answer their questions and focus on that, and sort of look at that deflection rate in and around, and, yeah.

Without revealing too much, we’ve absolutely achieved that. And the second sort of side to that was then looking at sort of decreasing the amount of time people spent on there to get an answer back as well. So, decreasing the average handle time. We, we doubled what we thought we did on that as well, which is really, really cool.

Well, and are we talking about, like, hundreds of, conversations a day is a, you know, in the thousands, you know? Yeah. So it’s, I think on average need to double check. But. Yeah, certainly in the hundreds. Wow. Okay. So it’s a big impact there. From, you know, we talked about data and you talk about, an effort that went on before to bring data into or exposed data into the CRM from a different, different set of, you know, database basically from a different database.

Is is that the case? Is that necessary? Was data cloud then? You know, part of the the picture not because of these. Yeah. So there were and I, I, I always think of this because I have this question quite a lot from customers and things as well. So I think, I think there’s two sides to it.

The first one is, and I’ll come back to this probably throughout, throughout this whole conversation is, is there a real use case to, to need that data? So what we focused on is chunking it down into five key use cases that we wanted to do. The first one was answer questions with knowledge. And then we wanted to expand that out into additional bits of information.

So specific information about your booking, you know, what your flight details are, for example, how much luggage you’ve got. And then we wanted to focus on, you know, cancellations, rebookings, things like that, which obviously get massively more complex as you get into that. And that’s kind of where we started to go. Right? We know we want to do these.

They’re not going to be out the back, you know, in the in the first iteration of this agent. But we know we need that data for later. So we’re going to start doing that now. Populating and connecting up what we need to do. Some of that was done with data cloud. Some of it was done with existing, integrations with, with platform events.

So it was kind of if we knew we already had an event bus that we could use, we’re going to use that because that’s the existing architecture will continue to use that. If it’s information that we’re going, you know, we we don’t necessarily need to show this to people internally. You know, they would very rarely go and look at it.

But we need to have access to it. That’s when we were sort of starting to use the data cloud functionality to, to pull that in, and which gives us the option if we need to later down the line of showing that information with it within the CRM itself as well. Very interestingly, talking about, you know, that particular use, of the technology and broadly speaking, the use of both old technology, you know, our good old Salesforce in the CRM, relatively newer technology with Data Cloud, and then the brand new, which, which is represented by, by Asian for us.

How did you feel like you aware and your team at perform it where the right combination of skill sets and expertise or maybe, you know, with the right attitude to learn, to take on this new challenge. Yeah. So we, we we looked at data cloud already. So we’re happy with that. Obviously the core CRM cloud, we’ve got lots of people that that know that inside and out.

And then, you know, we’ve got a couple of, people within the business, myself included, that are very, very interested in AI and have been for a long time, and do a lot of reading up and learning externally as well. We, you know, we’ve been playing around with connecting ChatGPT up and things like that within lightning components within the org and things like that, to see what we could do and how we could do.

So when we got told by Agentforce that might complete sentences that, you know, why Salesforce were doing this and had a lot of answers to the questions we had around data security when we were doing it. And so I think it was absolutely, you know, I think we did some work with Professional services, Salesforce Professional Services, the global AI team on this project as well.
They were they were involved. And we had a conversation at the start, which was, you know, they said, how involved do you want us to be? Do you want us to take on the whole thing in you learn or do you want to do it? And then you know, and feedback to us and we want absolutely we want to do it.

All we want to know is, you know, can we pick your brains around what you know about it? And it was very much through the pilot stage. We’re all learning together. And so it was sort of it felt like we were very much at the cutting edge of what is new. And I think for any technology team that’s working, that is the most exciting prospect you can possibly have.

It’s a dream scenario, right? It’s a very scenario. So we had team members fighting over who was going to work on the project and yeah, myself included, you know, typically, less involved in day to day delivery of projects. With my role. But I was first hand up was like, a project. So let me let me ask you a question with that expertise in mind, with it, you know, hands on expertise in mind.

So what are the things that you guys did wrong? And if you could go back, you would change them in? Broadly speaking, what are the biggest mistakes that you see happening in the world of, you know, getting ready for AI or embracing and deploying, Agentforce? Yeah. So I would say there was there was only one thing for me that sort of shone is as is the the thing that we’ve changed in our process with the, the relevant or subsequent, deployments and implementations we’ve done with other clients since.

And that was factor in how much testing was going to need to happen. And you’ll see that now that Salesforce has released the, testing center for Agentforce, they identified that that was something that that was hugely time consuming. You need to do you need didn’t have that. You didn’t have that luxury. No, no.

In your own tests, it was, an Excel spreadsheet, of people writing, you know, scenarios over and over again and checking manually in the agent, checking again manually going back and forth. The we weren’t expecting that we would need to do automated testing for it. I mean, we thought we probably would, but we didn’t think it would be worth the time investment.
You know, looking back in hindsight probably would have been, but yeah, that was it. It took, you know, any software project you would expect to have a good amount of testing in the, it was at least double what we thought it was going to be. So it was it was a real big commitment. I spoke to Salesforce Professional Services after the project.

We had sort of the debrief and were all sharing notes around everything we’d learned across the board and, and looking at that because, you know, we’re all still trying to sort of create the best practices product still changing every few weeks, it feels like. And so we’re constantly in touch. And, they said, yeah, every other other partner since it was released globally.

And the internal professional services team, when it was about for for every day of, you know, actual hands on making changes, it was about five days of testing. So it’s a 1 to 5 ratio and, and they said that was across the board. Everyone said the same thing. And I think that’s what testing centers come out. So that’s probably the biggest change.

But the biggest thing that I think people are also facing, I think we were we were quite lucky in that we were very proactive with thinking about this at the start. But I have seen it in, in other areas. It’s trying to take on too much as well, trying to go, you know, trying to boil the ocean with the first implementation of Agentforce.

You know, it’s the same as any other technology. Start small and grow it. I think that’s that’s the best way to look at it. Doing it. And it which is a great point that you’re making thinking about that experience of, you know, pioneering, you know, being part of the very first implementation and in the, in EMEA, what do you what impact do you think it has or it had on your business and your team?

And broadly speaking, what is, in your opinion, the opportunity that is in front of a company like to perform it, but broadly speaking, the ecosystem that operates now with Agentforce, you know, front and center. Yeah, it’s I mean, it’s made a massive difference for our business. We’ve had a lot of new, clients that have come through, you know, working with a year on, on Agentforce, implementations and deals there.

I think it’s made a massive difference for our team. In, you know, they were all everyone’s always excited to learn is a big part of our culture that we have here at the team. But I think it’s just it’s like it’s like adding rocket fuel into it that everyone’s just been all over it. I think everyone’s got an AI set in the business now.

Sales team included. Everyone’s just on it, really excited about it. So there’s been, you know, a massive opportunity for us. Absolutely. I think it’s probably been one of the biggest shifts we felt for a long time. But within that real drive towards a product, and it certainly put us on the front foot of trying to start to build out what best practices look like ever changing, of course.
But I think that’s been really fun to work on, like a blueprint of, you know, step by step approach and. Yeah. Yeah, what what can go wrong, what what goes well, how to make up for it, how to be ready. You know. Yeah. Yeah. Absolutely. What what makes someone ready for for for implementation. It made us go and look at our systems and go, right.

What what do we need to do to get ready for it as well? And I think the biggest thing is the opportunity for creativity that it’s created. Everyone’s thinking about what new use cases can we use this for? What can we use this for? What? What could we change? People are thinking about internally within our business. We’re thinking about, our current clients.

We’re thinking about industries that maybe we don’t do as much work with as our core ones, for example, that’s been really powerful. And I think it’s, it’s created, almost a new focus on innovation of how we can, you know, do a bit more blue sky thinking about what what the opportunities could be and how we can make those changes.

And I think that’s been, you know, so brilliant to work in that culture. And, and help foster that in terms of, you know, the opportunity for, for everyone else, I think I’m sure we’ll probably touch on it, at some point. But I think it’s sort of, it’s something you’ve got to dive into because I think AI is going to be here to stay.

The I think I always think of Salesforce as a toolbox. Any technology really as a toolbox. And AI is, is, is another tool within that toolbox that’s going to be used. And I think, you know, once you start to see some of the really big, companies that people use every day. Yes, but to integrate AI into their, you know, into their customer journey, that’s the expectation for customers from that point onwards.

And, you know, everyone’s going to expect you to have some kind of AI in your business, talking to them or wherever it’s, it’s place and it’s it’s going to be the expectation as a, as opposed to the innovation, I think within the next year. Certainly. Oh, wow. Well, I think definitely the narrative that we hear, you know, from Sees for themselves has been definitely going in in that direction to the extent that maybe they’re opening themselves up to some criticism, with regards to shifting too far, too soon in into the world of AI, maybe leaving behind some of the more current, you know, CRM related needs.

Have you seen that? Do you feel that’s the case? Do you feel like Caesars is doing the right thing? How do you how do you feel about that having worked, you know, hands on on the first implementation? Yeah, I mean, I can certainly see why they’re doing it. I have not felt and I don’t know if we’ll feel this for a little while.

Either way, I’ve not felt like there’s been any drop off on the new features that are coming out in core CRM, I think from, you know, all the conversations we’ve had with Salesforce, they are absolutely still fully committed to that. And, really want to push that as that is the core of the business and that that’s what makes all these agents tick.

I think, you know, from a business sense, looking outside in, in, I think it makes absolute sense as a product company to be, you know, really focusing on that. I’ve, I think they are ahead of the game with Agentforce at the moment, and it only makes sense to sort of focus on that momentum and sort of become that big player and double down on it.

I want to pick up on the the point that you made that they they’re still thinking about and creating these new products, these new ways, of innovation centered into their CRM. And I think that probably is what, you know, makes makes it successful or we’ll, you know, lay the foundation for its success, going and going forward.

Yeah. Fantastic. Look, I wanted to, come to a wrap up in a way of our conversation. It’s been super interesting to to understand more about your journey and how you guys were ready, how your customer and why, they said they were ready. And, you know, it’s and, you know, us and, you know, the rest of the ecosystem are also also ready.

I’ll leave you to kind of give it, a sense all together. But I think the key themes that come out, if I have to sum them on for one, readiness, definitely at the leadership level, with an innovative mindset, readiness to embrace, innovation, fail even. And I think the second point is, sent to, around that innovation is a step by step approach as well, an ability to understand and, a piece, that needs to be to be taken when embracing, new technology.

Then we talked about data, and I think that’s very, you know, crucial, their readiness for Secret Escapes and probably for the rest of the organizations that you’re working with centered around, data readiness and having the right data. And then we touched on, creativity, which I think was the highlight of of the conversation, thinking about the opportunity and what it does for your team, what it does for the ecosystem.

Your answer was centered around, creativity, which is, you know, unleashing new thinking across the organization and across both your clients and your internal, colleagues and employees, which is which is amazing. So if you had to, you know, sum up what you know, what’s just behind us with this, you know, first implementation and all the ones that you’re working on right now and what’s ahead of us, you know, how would you, would you see how would you frame that as an opportunity?

Yeah. So I, I think, as you said, you know, leadership point is, is key. And I think getting leadership to think about, you know, what, what those problems are within the business. I know most people within the business, you know, at a leadership level, will know exactly what’s wrong and where they need where they need to update things.

And it’s, you know, putting that, making that visible and getting everyone, I think thinking about what what could you do to help change this? That’s where those innovative ideas come out. Obviously all of that needs to be grounded in the business case. There’s got to be a use case for this. And I think making sure that you can see a clear return on investment or a clear metric that you’re going to be able to use to measure this is so important.

I think that’s what people try and bite off more than they can chew. With with implementations of Agentforce is think right? We’ve got a problem. Let’s fix this first. Or revolutionize this. And then we can focus on how we expand it and bring that out. That’s been one of the most important things. The other thing that I think I, I annoy people by saying is I don’t think we’re in the AI era. I think we’re in the data era at the moment. Every everything that we do with AI is underpinned by the data that we’ve got available to, to work with. We can use AI, to improve the data quality. We can use it to train the data, but it’s the same as an employee, right? If you train an employee badly, then you’re not going to get the outcomes that you want from that employee.

And it’s exactly the same with any AI. If you give it bad data, which is essentially bad education, that it’s not going to give you the outcomes you want. So I think that’s the big area that people need to focus on. And I think that’s the area that’s going to take the longest to get up to speed with.

But if you have a clear use case, you can focus on the data that’s going to help that use case first. And that’s going to be a lot less time than trying to fix the data across your whole org. If you know you just focused on answering questions with knowledge, let’s say to start with, or outreach with the SDR agent, something like that, you can focus on the quality of the data that’s coming into your leads.

You can focus on the quality of how well your knowledge articles are written. And then that’s a much smaller project to work on, which gives you the benefit of unlocking AI than trying to update all the data and everything within your org. And the other thing that I see a lot of people are starting to implement in, and that is going to be key for business, is putting AI within your technology plan.

Creating comprehensive data governance plans and building not only a culture around creativity and innovation with the AI sense, but a culture around, pride within the data that you have within the system and the understanding of how that can benefit everyone by having that up to date pride. I love the use of word of, you know, being proud of your database and your discipline around capturing the right amount of data and the right set of data.

Yeah, absolutely. Because I ultimately that is what’s going to drive everything. And I think that’s, that’s something that a lot most businesses have known for a long time, based on the reporting that tricks and things. But I think if you, you have been a little bit behind in that now really is the time to focus on, getting that up to speed and that will unleash the ultimate creativity, I guess, you know, business impact through new thinking.

Yeah, yeah, we will opportunities and and that is that’s what’s going to drive all the new opportunities we’re going to see in the space. I think we’re very much at the start of the journey of seeing how far we can push things. And we’re only going to discover that more with the new product enhancements that are going to come out, with, you know, now that testing can be done through the test center.

That’s going to help speed speed up. So you’re going to see more use case delivered, more use cases delivered more quickly. And then again, you’re going to get that sort of pipeline of creative ideas flowing through and helping build out, you know, what the new age of business is going to look like. I think over the next ten years, we’re definitely in for a very, very interesting journey.

Absolutely. Thank you very much, Sam. This was an amazing conversation. Thank you for sharing. You know, your experience and your knowledge and your understanding of this new, exciting era that is opening up with Agentic AI and Agentforce. Hopefully we’ll all be ready to embrace it. You know, also, thanks to your recommendations and your advice in this, in this episode.

Thank you for being our first guest. Thank you, everybody, for being with us. Yeah. Goodbye. See you next time. Cheers