Implement AI Podcast #90 - Why Agentic AI Is a Business Transformation, Not an IT Project

Speaker A: 500 orders a month was manageable. 5000 is madness. Embrace intelligent order fulfillment with Shipstation, the only platform combining order management, warehouse workflows, inventory returns and analytics in one place. What used to take five separate tools ShipStation does in one automate fulfillment and rate shopping to ship up to 15 times more with up to 90% savings. Go to shipstation.com and use code audio to try ShipStation free for 60 days real talent is defined by what people can do, not where they learn to do it. So by stopping at the education section of a resume, you might throw away the perfect hire skills first. Hiring helps you see talent others miss, like more than 70 million stars skilled through alternative routes let their story unfold and gain a competitive advantage because hiring managers who start with skills are 60% more likely to find a successful hire. Hire skills first. Learn why at tearthepaperceiling.org, brought to you by opportunityatwork and the ad Council Home

Speaker B: is for relaxing Whether you're hosting friends for dinner or enjoying a quiet afternoon by yourself with a tasty treat, but it's hard to relax when a couple of flying insects always seem to show up uninvited. Then those one or two fruit flies suddenly become hundreds, and that can make it even harder to relax. Next thing you know, you're spending time swatting them away. Thankfully, the Zevo Flying Insect Trap takes care of flying insects for you. Just plug it in and it gets to work 24. 7. It's specifically designed blue and UV light attracts flying insects like hel flies, fruit flies and gnats, trapping them on a sticky backing without any odors or mess. The Zevo Trap also features a sleek curved design that keeps captured bugs out of sight so it blends seamlessly into your home. When the cartridge is full, you just grab the no Mess tab, throw it in the trash, pop in a new cartridge and then you're done. It's quick, easy and you never have to touch a single bug. Plug in a Zevo Flying Insect Trap so you can work smarter, not harder, and enjoy your time at home with 24. 7 bug protection.

Speaker C: My challenge to every leader in this UK where you're sat on or responsible for a business process. Outsource. Do not pay for bums on seats anymore. That day has gone. You can take, and I use this example, you can take 10,000 FTEs worth of work in any low cost location outside of this country and you can bring that work back onshore. And instead of having 10,000 FTEs or people create jobs for 2,000 or 3,000 people working alongside AI. AI isn't coming for your jobs. Humans that can work brilliantly with AI is what's going to be the premium. That's what I want.

Speaker D: Welcome to the Implement AI podcast, where we explore how AI is transforming businesses in the real world. At Implement AI, we help organizations move beyond AI experimentation to to put AI to work. Through our AOS platform and teams of AI workers, we've deployed hundreds of AI agents across multiple sectors, helping businesses drive revenue, increase capacity, improve customer experiences, and get more done without growing their headcount. Each week on the podcast, we share practical insights from the front line of AI implementation, cutting through the hype to discuss what is actually working, where businesses are seeing real value, and what the future of work looks like as human teams and AI workers operate side by side. So let's get into today's conversation. I'm, um, Piers. I'm Alec. We're going to keep it short this week because we're getting Zahra Boalumian, who is the CEO of Salesforce UK and Ireland, and we could talk for hours, basically, because Salesforce is quite similar to what we do, just slightly bigger. So if you're new to the podcast with Implement our podcast, we've got a lot of information on our website, our resource center. Uh, if you like the podcast and watch it, you listen to it, especially on Apple, please leave a review. If you're on YouTube, please, like, subscribe. Ring the bell, because as you probably noticed, we're having some fantastic guests on the show. So without further ado, let's bring in Zara.

Speaker A: Hi.

Speaker D: Ah, Zara, good to have you on.

Speaker C: Thanks for having me. Nice to see you.

Speaker D: So, as we said, we've honored actually to have Zara Balalumi on m. What podcast? And normally we let you do your intro, but I'll do a little bit of one. We've done some research on your background, which is very relevant to everything we want to talk about, really. You're the president and CEO of Salesforce UK and Ireland, and I think it's fair to say you're one of the most influential voices in enterprise technology and organizational transformation especially. Everyone's probably noticed you're also a woman as well. So it's nice to see women at the top of, um, It. It's not often the case, is it? So you've been doing this for three decades, actually. So tell us a bit about your background from Accenture to Salesforce, and then we can get into the questions.

Speaker C: You make me sound so good. Thank You, I appreciate the intro. It is lovely to be here with you both. Ashok and Lily. You're right. Yeah. You've aged me very well. I've been at this for nearly 30 years now and I would say the first kind of couple of decades I spent in consulting and a lot of technology transformation and implementation, hands on tools. I lived through it all. I lived through kind of mainframe programming, I lived through the client server revolution, I lived through kind of the service oriented architecture revolution. I did digital TV implementations and I worked for 10 years in oil and gas. I've kind of seen all of the cycles. And then latterly within Accenture, I led the technology business and that was really eye opening because I had all the engineering teams, the infrastructure teams, the go to markets and all of the ecosystem relationships. So that really was a massive launch pad for me. And then from there I came into Salesforce nearly six years ago now. So I lead Salesforce in the UK and Ireland, but I also now run the private equity and the VC business globally, which is an incredible business. And again, another eye opener. You stumble across these opportunities and then you suddenly realize what you don't know and kind of seeing this AI revolution through the eyes and the lens of the private equity and businesses and the funds and the VC businesses is something that is a sight to behold, I promise you. Like literally following where the money is and seeing the trends and the patterns that they're forging is quite a privilege.

Speaker D: Let's start zoom out really. So in terms of your view on AI generally the technology, transformational power, then we can maybe talk a bit about Salesforce because Salesforce has moved marks me very, very quickly on this actually. So it's giving you kind of the overview really of the kind of things we want to get into detail on.

Speaker C: I want to talk about the context, if I may, of the UK because it's a bit closer to home and I know we want to zoom out and we'll talk about global trends as well, of course, and kind of where Salesforce is thinking and where we have been thinking. But I think about the economy and the world that we're in right now and we've got so many things going for us to really take advantage of this incredible technology that is AI. And we've got a lot of expansion of the model companies in the uk. We've got our own model companies growing, we've got loads of investment coming in. You know, we've had something like 65 billion of capital since 2025 in January coming into this country. We've had more unicorns than France and Germany put together. And then the talent we're churning out, you know, our Oxford, Oxford University is now outranking mit. So we've got so many things going for us. And so that's why we as uh, Salesforce are committed to investing here in the UK. You know, we've committed a $6 billion investment over the next few years into the UK. We continue with our venture fund to invest in companies like Elevenlabs and Convector. So we've got a thriving opportunity to really trans I think the future of our country, our productivity and I think with agentic AI, just the expansion of capacity around us I think is going to be incredible. And we see that with, I don't know if you come across agent Bobby, Ashok or peers, but we now have an agent that resolves 82% of non emergency queries through a number of police forces. Can you imagine that level of responsiveness and speed? So I think the key to unlocking our stagnant productivity and I think our productivity is stagnant within the uk and I'm sure economists that listen to this will argue with me, but I think we've got so much more potential that we're not tapping into. And I think for us to tap into it, we've got to really focus on a couple of things. We've got to continue to attract that capital. We need that capital in the UK and then we need to create an environment where businesses feel like they can thrive and stay here. I still haven't forgiven ARM for listing elsewhere. I'm terrified these new kind of native AI companies are going to go and list elsewhere when they expand because they will expand. So bringing that capital and creating kind of the environment for people to thrive, I think that's going to be key. I think I talked about Bobby just now. Expanding the capacity of our public services and our citizen experiences is going to make us all feel very engaged and familiar and content and happy with the level of responsiveness that we get and then continuing our capability like think about our talent and invest in the skills. That's why we are in Salesforce. You know, we're investing in our future Trailblazers program. We're going out to schools and young people to teach them about AI. And we've hit the 13,000 mark actually since launching a couple of months ago. We're on a mission. We want everybody to embrace it because when we embrace it, that's when we can unlock um, the power of the tech.

Speaker E: In my View Also let's talk about

Speaker D: Salesforce because Salesforce move very very quickly. So developing a product agent force maybe in terms of as far as you can talk about fin acquisition but the outcomes, uh, based pricing, headless CRM. I mean it moved very very quickly in a lot of enterprise kind of SaaS companies.

Speaker C: We had a really big pivot about three years ago so we decided to rebuild our whole platform, all of Salesforce for agents. I think I'll start there. And we believe in the agentic enterprise. And again just to ground everybody to you know what do we mean by an agentic enterprise? We mean an organization where humans and agents and processes work in a very governed, tightly governed ecosystem and a framework. And that's important in the context of enterprise AI when you think about the security and the control. So we completely re architected our platform to have we think about it now in five layers. Most people think about it as a CRM that stores your contacts. That's not who we are anymore. We are what we call a headless 360 platform and that is really to enable any organization to utilize us and plug and play whatever model and whatever context data they want to into our platform and they can have the power of Salesforce right there. So we think about it in five layers. We think about our ah, data context, our uh, business logic, the agency layer, the engagement and the insight. So this is quite conceptually that's us. We then think about how every organization will want to utilize their own agents. It doesn't always have to be our agent force. We do have our agents right? And our agents, the strength of our agents lie within our uh, heritage functionality. So we've got sales agents that are native to our platform and service agents that are native and commerce agents and marketing agents. And we've got a whole host of those that are kind of inherent and make our platform kind of hyper automated. And we've also got some 200 out of the box pre built industry agents because who wants to spend their time building stuff that's already kind of an industry standard process or an optimized industry standard process. So any agent can MCP or CLI into our platforms, the open protocols into our platform and we are uh, as you mentioned some of our acquisitions we are slowly and steadily acquiring more and more. So we had our uh, just in our process of our fin acquisition, that's our service agent and that is really helping many of our mid market and our smaller enterprises accelerate their service capability right out the box. And we've got qualified also A sales agent. So a humanoid version. We've actually got one on our own website. So if you go onto salesforce.com, you'll see an agent Piper pop up. And that's. Well, that's powered by our qualified acquisition. So our sales agent out of the box. And it's humanoid. It will talk with you and its mission is to understand your needs and your goals and then essentially pipe you, qualify you, and then refer you to a human agent. And by the way, that in doing that we. Something like three quarters of our inbound leads were just going untouched by the way. They just completely. We didn't have enough human capacity to capture and convert it. And then people were only kind of capturing and converting what they knew. Whereas Piper, our sales agent, enabled us to qualify these leads more meticulously and that's generated some $60 million for us in pipe.

Speaker D: It's quite interesting that, because maybe I'll talk about this. So a lot of what we do of our customers is people think it's some, you know, clever agent deployment. It is that is capturing revenue that you've missed is reactivating customers, talking to existing customers, old prospects at scale. Because you can't afford to pay a room full of humans to go and do that.

Speaker C: Exactly. Um, and can I just, like most people think about AI and agentic AI as an efficiency player. Yeah. Okay. It can drive efficiency, but to your point, its ability to capture revenue and meet customers where they are at scale is unparalleled. So it is a growth agenda as well as an efficiency agenda. And those leaders that understand that and embrace that do really well in this context.

Speaker E: That makes complete sense, I think. Zara, how do you find the mindset about the organizations you work with? Because obviously you can provide solutions across support some areas of sales and growth, maybe the things that they aren't even doing right now. And obviously you yourselves in Salesforce are very forward thinking and you were talking about all the different things you're doing. But it'd be good to understand like how you find like many of the organizations that you're working with. Do they have that vision, that understanding about, like, let's transform these different areas or let's do things that we've not done before. Just be great to hear from your perspective.

Speaker C: Well, I'm going to tell you a funny anecdote. Piers was actually in the room when we were around a table of esteemed leaders. And, um, Piers actually pulled someone up on something they said in the room. And by the way, this kind of exemplifies some of the mindset of some organizations that we deal with. This particular individual, I think, said that they were running a couple of experiments and they had delegated it down. I'm not going to say who it is, Piers, but I remember you pulling him up on it. He had delegated it down to a number of people in his team. So they had like three or four experiments. And he felt really good that he had three or four experiments going on. And Piers was like, that's going wrong. He said, the fact you're delegating it, you're not close enough to this. I can't remember the words you use, Piers, but you basically said you're not close enough to it. You're not embracing it as part of your kind of own day to day. You didn't shame him, by the way, but you did call out.

Speaker D: It's not nice to have. We often hear, oh, we haven't got an IT project budget this year. It's like, whoa, whoa, whoa, whoa. This is a fundamental change in how you're going to operate your business. We're not going to bore the ocean,

Speaker E: but we're all fine.

Speaker D: And what Alex alluding to is that. And you'll probably find the same thing, the technology, you know, you've built your technology. We've got our platform. It's the implementation and the mindset and

Speaker E: the rate of learning.

Speaker D: Yeah, learn it. And Zoran, um, you've been in this industry for a long time. Used to have the early adopters, the laggards, all that stuff. It can't be a laggard now. It'll be fatal.

Speaker C: You can't afford to be a laggard. Uh, you've got to understand it, you've got to get close to it. It's not some computer dude in the corner that's going to solve this for you. And it's not like I'm going to point someone on my board or I'm going to appoint someone in my team and everything's going to go through. No, no, no, no. It's absolutely embedded. Why is it embedded? I think if it was okay, you could kind of understand that it fundamentally will reimagine processes and enable your company or business to do things that it was never able to do with humans alone. That's the kind of mindset you need to think about. And I think again, we've got different camps. We've got those camps that are saying, if you're going to deploy AI purely to cut costs and think of it as a Computer project, you're designing it as a 20th century model. If you are thinking about it as reimagining your whole business, if you're thinking about it, like I'll give you an example, Formula one, um, hundred employees trying to personalize the experience to nearly a billion fans and using agentic AI to do that and really capturing that revenue and enhancing the experience of those individuals that are engaging with it. That's a 21st century brilliant opportunity. And I think that the mindset, you're right, some can bolt it on, some understand it's got to be inherent in how they redesign and reimagine their business and the ones that are really thoughtful about it. And I had a very inspiring conversation with the CEO of uh, a very well known bank his last year, last year. So a year ago said Zara, we're going to be an agentic enterprise. There's no doubt I can see the opportunity that this will bring us and bring for our customers. Help me think through the implications of what I need to consider. That's a different mindset. He's like, what are we doing that's inhibiting us? And I went through the whole kind of conversation. Well, you're trying to buy data capability in silos. You need unified data fabric. If you're going to really take advantage of AI, you need to think about your data, you need to think about where your data is, you need to get your data in order. Secondly, you need to think about this as a change program that's going to augment your workforce. Like this is a positive thing, this is going to help their employee experience, going to make their lives easier, going to skill up their cv. They're going to be more valuable, they're going to be more empowered. So you've got to think about it as a change program. And then we talked about kind of the processes. If you're going to take a workflow and you're going to optimize it, go pick any other enterprise software tool. If you're going to reimagine a process and you're going to think about what do you no longer need to do and how do you kind of consolidate that. Great, we can work with you. And then the last thing which is a really important thing kind of mindset wise is security and the governance and the controls and the protection of the data, uh, and the zero retention policy. And it's just, just the kind of governance and observability. Like just some people lose their minds, they're like, oh, I can get a model to do that. I can bypod this and get a model to do it. And it's like, hang on a minute, you're going to be rolling back. Oh, there you are, you're rolling back because you didn't think about the content,

Speaker E: second order consequences definitely you didn't get

Speaker C: it grounded, you didn't get the security controls and the framework in there. You didn't put your toxicity and bias filters in. You didn't think about the agreements that you need to make with all of these model companies around, the zero data retention, you're now having to unwind all of that and you're having to explain yourself. And by the way, the regulator's crawling all over you. So let us help you. So you have two camps, the bolt ons and the ones that really get embracing and think about the broader system around it. And by system I mean those four

Speaker E: things I just said, I mean that requires like quite a journey for them, right? Because obviously like moving from essentially managing basically business as usual and how things are and executing a known plan to almost like reimagining and building learning loops and like, okay, how can we have different things there? We think about security. I mean that is a big jump for many people. And also many people think that I can do my job correctly today. Okay, the company might be managing downwards, but I can do my job because I know what I'm supposed to be doing. But in that new world, I can imagine many leaders are like, I don't know how to do things in that new world. And therefore I'm worried that I'll stand away from everyone else. And I can't just. I see this kind of organizational inertia like in some organizations we speak to because you might get one or two people that have that vision, but unless they can actually bring the other people with them, you're left with like some kind of silent project basically.

Speaker C: Right, I agree with you. And I mean just to uh, kind of balance that point and first of all, I do understand that there is a level of, and I have great empathy with, level of fear, lack of familiarity. And by the way, we don't want the world to become experts in AI. We want people, you know, the power of this technology is to release this human capacity to focus on their own core competencies, their strategy, their businesses, their industries. That is wonderful. And that frees them, you know, the administrative burden. It frees people to focus on their customers and really what matters and that depth of connection. Right, So I do understand that. And I'm not saying God, you're going to fall behind if you don't play with it, you know, if you don't really become an expert. It's about the, like you said, the mindset and it's about remembering, irrespective of technology, don't lose the good discipline. Just because you've got 10 FTE4 deployed engineers experimenting and kind of creating agents doesn't mean you forego, uh, what business value are we trying to address? What processes trying to change or optimize? What capacity is it going to create for me and how is it going to impact my customer? These are fundamental questions that don't change, whatever the technology. And so I think again, we've got so many good examples of fantastic customers. You know, we've got Heathrow that have embraced this. If the police can embrace this and drive value, then, uh, and a regulated entity can embrace this and drive value. We've got secret escapes utilizing this to really transform their service capability. And they're doing that beautifully. You know, as I mentioned Formula one as a really great example, we've got Simply Health that are freeing their teams from the administrative burden in quite an emotional moment when people are suffering from health challenges, to release them, to kind of be more empathetic and spend longer in with the cases that need them. They're actually paying their call center staff more because the level of complexity and the level of involvement in the cases.

Speaker D: Well, let's talk about that then, because that's always the elephant in the room, isn't it? So I will say, yeah, you've got kind of humans and I would say our agents are expanding human capacity, augmenting human resource, always beyond human capability, like scale and speed and that kind of stuff. And often the way we are positioning, implement AI in a way is kind of outsourcing. So outsourcing normally to your background. So it'd be Microsoft and Accenture or Teleperformance. What we're saying is you can outsource a lot of that, like Heathrows didn't do a lot of that kind of repetitive, repeatable words on a scale to AI agents with humans in the loop when you need them. What we're now doing is almost bringing back onshore, it was taken offshore. It's the amount of fewer people doing in the end, let's face it. But they're going to be here, not wherever there is.

Speaker C: Honestly, you've just hit my absolute passion project. So if you think, if you believe that we're in a race for workflow automation, if you believe that to be true, then you've got to think about where have we got repetitive tasks? I mean, where are the biggest cost drags in our organization? It's the back office, where it's not a competitive differentiator. Back, uh, office processing, straight through processing finance. Processing, finance settlement. None of this is a competitive advantage. It's a cost drag. And the days of labor arbitrage now are done like that lemon's been squeezed. And of course people might argue, well we can go to another low cost location that lemon's been squeezed. Now I think agentic AI lends itself beautifully to business process outsourcing. Why? Well, if an organization has outsourced their business process, they're already used to buying an outcome, they've already codified that process. It's probably closed loop and they're used to paying on outcomes. So you've got the most perfect test bed with the least risk to your kind of front line to actually do something to take out a, to take out cost, but to optimize these processes and have a, uh, m. Much more uniform and scalable reach. So then comes the what do we need to do in this country? If you believe my first point, which is stagnant productivity will only be unlocked if we've got the talent and the people that are capable of using it. Well then the logic follows that uh, for people to use it and embrace it, it's gotta be in their day to day work. And so it has to be in their job fabric. It can't just be a pilot or an experiment. And so therefore my challenge to every leader in this UK where you're sat on or responsible for a business process, outsource, do not pay for bums on seats anymore. That day has gone. You can take, and I use this example with what you could take 10,000 FTEs worth of work in any low cost location outside of this country and you can bring that work back onshore and Instead of having 10,000 FTEs or people, create jobs for 2,000 or 3,000

Speaker E: people and more interesting, more upskilled jobs as well.

Speaker C: And they're more upskilled, they're working alongside AI. AI isn't coming for your jobs. Humans that can work brilliantly with AI is what's going to be the premium. That's what I want. I want engineers that can work with AI. I want sales leaders that can work with AI. So create those jobs here. My God, you'd be a hero. You'd be a hero or a heroine. Job creation, brilliant job kind of profile and You've got amazing cost advantage. And the great thing is we're not subject to the same workers councils as the rest of Europe. So we can go faster, we can do this, but it takes somebody brave to do that.

Speaker D: Before we get back into the conversation, a very quick word on implement AI. We provide managed AI workers through our workforce platform, an AI workforce as a service for SMEs, regulated businesses. And we have large enterprise customers too. Our highly configurable agents communicate Through Enterprise Voice, WhatsApp, SMS and email and they can use and operate across all of your systems, even when there's no API. Our interactive agents engage customers, our analyst agents surface insights and our action agents complete work. The outcome is very simple. More revenue, more capacity and um, better customer experiences. So start with a pilot to identify rapid ROI use cases, then scale whatever works. Go to implementai IO to learn more.

Speaker E: You've made a very excellent point, Zara. The whole point there is that like when they already are outsourcing to BPOs, they have got a defined objective, they've got a defined outcome, they've got a process, they've got an internal buyer who is responsible for, for that and they've got that whole loop around there. And I think the challenge is in the other areas they're trying to identify. That's just want a better term is they don't have that. Basically they haven't named those, they just have variable performance amongst different team members. I remember we were on a roundtable with you and some of your senior leadership team and anthropy and then there was like one person was giving example that the difference between their top salesperson and their bottom salesperson was like almost like 10x revenue basically that uh, they were generating and we've kind of got used to this kind of like distribution of outcomes. But if you can kind of like at least put agents in to help augment or the people to actually say like look, the last five meetings you did, this is where you're falling flat basically, right? Or here's the preparation that's going to help you go think do things better. Or if you then had even, let's say for example the qualification done ahead with like an agent like you just mentioned, then you can already fill that knowledge gap for the weaker performing salespeople beforehand so then they can actually focus on going deeper into those questions. But this requires prison that learning loop and basically having everything readable. So you're not like having to go through the old model of like oh, let's do a meeting to find out what happened. Because a meeting to find out what happened is a search you could not do because you're not connected to your systems basically. So I think like this is the whole way to go forward.

Speaker C: I am living proof of that because I use my our own super agent Slack bot every day. So every day I come in, I have a today Slack. So it literally prioritizes, scans all my channels, looks for anything that needs my attention so I don't have to spend time going through it and I can respond in the moment to that number one. Then it lays out my meeting schedule and then it has a prep button next to it so I can actually prep and utilize that so I can expedite it. I missed out. I didn't go to our annual gathering last week in outside of the uk I didn't go but I really wanted to hear the keynote. But I didn't have time to hear the key But I needed to know in that moment what were the salient points. I asked Slackpot because it has access to all of our content. I said can you give me the highlights? What are the key takeaways of a keynote? And it gave it to me in five succinct points because we were grounded in our own enterprise data. The other day I went into a customer meeting and I hadn't appreciated the extent to which this organization is a customer of ours because it's recently transferred into our territory. And so literally a few minutes before the meeting I was able to query Slackbot and it was able to go through all of our channels or any preparation, any prior meetings that have happened last into our CRN system and it was able to serve me within seconds what I needed to know so I could be relevant in that context. So it's helping me. And so you think about that every rep having that just think about the power they would have at their fingertips. No wonder that and I remember the case you met around 10x sales but I think your point is really strong Alok is that it's. You no longer have such wild outliers with a distribution in terms of talent you can actually concentrate it and move that all up up the productivity curve if you like is the best way I can describe it.

Speaker D: Also now that kind of outsourcing of the repetitive I always say automate the mundane and then get humans do the meaningful. But now because of AI it's coming down the future. And so I love talking to you about a salesforce and agent force because you talk about 10,000 customers, 10 FDES we've got the same business but with less zeros.

Speaker C: There is no reason why your organization wouldn't leapfrog. If you think about a small and medium enterprise, it is now with the power of agentiq the large enterprises will get the productivity gain and the efficiency and the growth. I think this is the day of the small and the medium enterprise in that ah, it will be able to leapfrog and punch well above its weight. I meant it when I said nearly a billion fans being serviced by 100 staff in Formula One. That is, I anticipate that business process outsources today are at serious risk. Not because the opportunity isn't there, it's that they're not rotating, the work has

Speaker E: changed, the value has changed. Basically if you're providing the same thing over and over again, that's not the thing because you want a learning and improving organization and they're just basically even give you further insights in the data repeated questions that are coming or different insights and stuff like this. I think it's one question I would have for you, Zara. So some of these benefits that we've got like for some of the ROI is going to be a faster, more agile executive team, more insights, maybe less team meetings internally. These things you can't see on a P and L in an easy way. Right. And in order to like actually see the P and L change, let's say for example like the agent can qualify people on the website and find out more information but you're still relying on the persons then follow up and do those extra steps or do different things and stuff like this. How do you talk about ROI with the businesses that you work in? Because it'd be interesting to understand like how you bill and also like how you help them frame it because it is a transformation. So I'd be great to understand how you think about that.

Speaker C: We can talk very generally and by the way, before I answer that, I'm going to forget this if I don't mention it. Look up the Sequoia Capital framework and if you don't have it, I'll send it to you. And it plots all the business processes and looks at their propensity to go completely what they call autopilot. They say what can you completely automate on autopilot and what's emerging and what will be copilot and that will tell you kind of honestly the direction of where the market could go and which processes could be automated, especially in the context of your business. And what I was about to conclude on was the Challenger to business process outsources today who rely on selling humans is they will be absolutely challenged by an organization like yours because it's going to be a very small enterprise with the ability to automate these processes super fast and it's going to win out. But how do we talk about ROIs? Come back to your point. Think about our organization, salesforce, right? You think about our moat. If you want to grow your revenue, if you want to grow your sales and drive commercial excellence, you come to salesforce. So we have our unparalleled agentix sales capability. If you want to service your customers and transform their experience, you come to us because we have the best servicing capability and we have the best marketing capability and we have the best commerce capability. And all of that is stitched together as I talked about in our headless platform. So when we talk about roi, we don't talk about the things that we know to be true in our own customer zero. Like we put our help agent live service, we put our service agent live in October 24th. That's now generating something like $100 million worth of savings because we're in the 80 plus percent category of it resolving complex queries and we were able to redeploy our engineers. So it took our service cost down. So that is true. So those are the things. And I talked about our sales agent kind of converting those leads and driving 60 million worth more of pipe. Absolutely. But what we talk about is if you think about sales, we talk about leads that you would never have captured and the ability to grow your revenue. That's what our agents will help you do. When we talk about service, we talk about our 24 hour service. We talk about deep personalization that will convert and drive next best action. When we talk about commerce, we talk about being able to buy and transact within an agent rather than customers having to log on and meeting customers where they are. So every metric that you need in terms of sales, uplift, growth, uplift, marketing, optimized campaigns, cost out, automate, all of those things we do in the context of the function, that's what we talk about. So we keep it very relevant to our moat. And that's when I think about other organizations and the customers that we deal with, they have to think about what outcome do they really, really want. And most of them either want an uplift in sales, more efficient marketing, and they want to be able to expand through channels that they never had before. I'll give you another example. Live golf, we sponsor LIV Golf. We also released something Called Agent Caddy. It's called Chip. And we wanted to revolutionize how people bought sports merchandising. So now when you watch a game or a game of golf, if that's your thing, Agent Chip will come flying in. And if Bryson Dechambeau has taken a shot using a fancy golf club and is wearing a shirt that you like, you can actually transact through the agent. How amazing is that? It's like, hey, Chip will just tell you, hey, Bryson just did that shot with this girl I dream about that. I would love to watch, by the way. From a personal perspective, this is really frivolous. There'll be things I'll watch on TV or on Netflix and I'll be like, I really like that outfit. I'd love an agent to be able to pop up and me to be able to buy it. There's no real business case in that at this stage, but I'd love that.

Speaker E: That's what ASOS was to begin with. As seen on screen is to Agentic

Speaker D: commerce, which is a whole different podcast to go and talk about. So just talk us through then. If you get advising, then anyone listening to this, whether you're enterprise, mid market, even smaller businesses, actually what should advice them in terms of embracing and deploying this technology? Because we often talk to so many companies where it's still in kind of project discovery phase, not deployment. And you think, why would you not. This is my frustration. Why would you not use this technology in some way? Because there's always a use case in every single business. So what's your advice to business leaders or business owners?

Speaker C: Well, first thing I tell everyone is just get out there and learn. First of all, come on safari. Doesn't matter which organization you talk to, we've all got a well known playbook that we can embrace and immerse you in it. So take advantage of that. The amount of boards we've taken into our AI center or even here in the tower, or even them to San Francisco and just immerse them in the art, the possible. I think that's kind of point number one. Number two, get really crystal clear about what business objectives you're after. Forget the AI for the minute. What are the business outcomes that you as a business are after? What do you need? First of all, that's the kind of first, um, thing I would say. Secondly, what we encourage our customers to do is identify one high impact, well defined process with measurable outcomes that's got a tight feedback loop that you can actually prove the value. Just prove the value. Because that's a really good kind of control case. If you can prove the value there and tune it, then you can actually scale that process a bit wider. So I think just start with a process, scale it, go deeper into that process before then thinking about expanding it, but always in the context of what you're trying to achieve. So a bank that has had a regulatory warning for not providing consistent customer experience and maybe causing customer harm or the potential for harm, which is terrible against the consumer duty regulation, that's a problem statement. I've got to improve my customer experience. I've got to do it at scale. I've got to do that because my license to operate depends on it. That's a business outcome. So there are processes within that customer experience. Whether it's a bereavement or whether it's account opening or it's a check my balance, it's a closed loop. Those are the things that are really great to start with, but then don't be satisfied with just a pilot proof. Start scaling it across your service capability. That's how we think about it. I could give you a similar example for marketing. Automate the campaign, scale it, uh, go to a broader audience, different market. So I don't actually think this is different advice to technology that isn't AI so any kind of technology, you want to know what your business outcome is, you want to prove it, uh, in a controlled environment and then upscale really quickly and bring people along with you and educate them.

Speaker D: I mean, I think there's one dimension though which is, I think where SaaS in a salesforce, one of the leaders of the SaaS of the revolution, but it's moving a lot faster. And that's the issue I think is that uh, I always do my keynotes is slide of a ship leaving the harbor and it comes to a point where it's so far away that you're never going to make the leap. And then your competition, you're never going to catch them up.

Speaker C: What do you say to people in that? What do you say to your customers?

Speaker D: So again, we're very focused on our wow, we are lock. So kind of show them and you know, discovery.

Speaker E: I think the same sort of thing. Like, I think there's like, there's two types of kind of mindset really. There's one mindset which is like they've got a definite problem like you mentioned. So like they want to like improve the metrics on that thing and that's much clearer to kind of go after. And like we've got Some great customers where they've actually scaled up the volume because they're very clear on what they're trying to do in that kind of area. Then you've got other people that just trying to discover and figure out what's possible. And they could be a bit challenging because they need to move faster and they need to try more things that will kind of guide on that. But then sometimes they could be a little bit focused on one particular area and then there'll be a second order consequence. For example, let's say reactivation. Right. We can reactivate customers that didn't buy anything, but that then requires a salesperson to follow up with them. But then if they don't have people that will then follow up with those people, basically. Right. Then they kind of get stuck at that stage essentially. Right. Or they'll. Someone that reacts to the evening, but they could. They call them during the day and they can't get that meeting. So I think it's like thinking about how are you going to support this function transformation? And I think the biggest thing that's missing at the moment is really like how to like envision what your AI first department looks like. Because right now you've got like this much volume going through it. You have essentially stagnant or stable outcome, let's call it that. Right. And that's what links to what you talked about in the British economy. Let costs go up every year. The outcome stays about the same or doesn't increase. But what you're trying to kind of shift that to an AI first is like a faster learning loop where like each outcome is measured and you can try and understand a benchmark. What was the learning or what was the blocker and what's the next step from it? And then how can the person then improve doing what they can do more or what parts could you actually delegate to the AI to like speed up that, you know, capacity or increase the capacity to then drive more volume through that. But that requires a management that's going to be open to like optimizing different functions around it and going there. And generally we found that with people that owning a transformation function in their business. But in most businesses where they don't have someone like that, but it doesn't quite work. That's what I've seen basically. Right.

Speaker C: I think that's really fair. And actually you can't just drop an agent and leave it be. Like, you do have to spend time, you do have to tune it, you do have to test the outcome. You have to be Hyper vigilant for a period of time before you can, like, you don't just release it into the world. And I would say Thames Valley Police was a brilliant example of that. When they deployed Bobby in the first force, they spent weeks and weeks and weeks just tuning and testing and they wouldn't release it until they were confident that this was going to be safe to use. And they had, I don't know, the best word I could say, pipe, clean, whatever is the verb. But they were so vigilant with it. They were absolutely brilliant. And then they kind of started to release it gently and slowly and then all of a sudden it kind of, uh, scaled exponentially. The one factor I noticed is that where someone's trying to tightly control thing and they're trying to build it all themselves. That's another extreme that doesn't work to your point. Those that can actually think then be open to where it might improve or enhance their business. And actually they don't have to build it all. You can get stuff off. I said, we've got 200 plus pre built agents.

Speaker E: I think there's a job title missing, to be honest. I think there's like a title missing which is like optimization or department optimization. Because literally you have to have the authority to look at all the different components of that outcome. Basically.

Speaker A: Right.

Speaker E: I mean it is like a sales leader, sales manager, but not just in the kind of people function way, also in the technology and overall functional way. And I think that's kind of missing basically in some areas, essentially. Right.

Speaker D: But also I was going to say that the buy or build. So we often people say, oh, uh, technical people on a call, we can build that. Okay. And you know from build the agent force just how complicated, how many blockers there are.

Speaker E: Uh, you could build your own house. Don't want to though, do you?

Speaker D: Right. And then not only that though, even if you build it internally, so you built it, then you've got the workflow, who's going to implement it? You sort of miss that bit in the middle.

Speaker C: Yeah. And we always say now buy to build faster. Our platform appeals to the builders because we say, look, the workflows that are functionally related. So sales, service, marketing, commerce, that is built upon decades of expertise. It's optimized, it's standard, it works. What you want to build is what people might want to experience. The building is around their agents. They might want to build different agents using different technology. Eat your heart out. Do it. You can make use of our platform m beautifully because it's Safe, it's proven and it's got all the inherent security controls and the governance. The thing that the people miss when they want to build everything themselves. They miss the framework of governance. They've got to build around it. The control framework. That's the hardest bit to have. Anyone can build an agent in three seconds on any platform. Anyone can vibe code their way to anything. But having it grounded and secure is the challenge.

Speaker E: Yeah. And completely optimized and upgraded and as things developed. And then if you've got a team member whose job is like developing your core platform, how much is it going to cost in terms of salary? Just like maintain something else. And are they going to be taking the details where it just doesn't cost in Basically, I think it's like having that kind of understanding. I think it's completely agree with what you mentioned. Like I said, some companies actually also then have this view that like, if something's going to be inherent to their business, they feel they want to like do it all themselves, basically. But it'd be great to understand how you talk about that with those kind of companies because obviously what someone thinks they can do and what the reality of that's actually going to happen and that the cost and the time loss and all that kind of stuff is going to be a problem, isn't it?

Speaker C: Well, we say buy to build faster. Take a platform approach, think about the layers, choose whatever. We don't care where your data is, but buy the control framework that can zero copy and get hold of that data so that it can serve it up real time into the workflow. That's what we say. We say, look, we've got your data layer. That context layer is really important. We don't care whether you've got a lake, a swamp, a warehouse. We don't care where it is, we can zero copy into it. You don't want to move your data around, but we can serve it up so that you can make use of it within our workflows. And then we say within that, uh, we've got this, what we call our Einstein trust layer. So that Einstein trust layer means you can utilize any model safely. You can keep your data within your perimeter because our models never keep it. We have zero data retained retention. We also have a zero data retention policy when you utilize any other model. So we don't care. We just want to make sure that your data is your own. So don't bother trying to build this kind of security control framework because it's already inherent in there. That's There. We've spent billions on our platform to make sure it's there. Why would you try and recreate that? Uh, so make use of that natively within our platform. Spend your time building how you want your people to work. Like, where's their place of work, what is their engagement layer? What are the agents that you want to use that, uh, are differentiated? What agents do you want that no one else can have? Don't try to spend your money building the platform. Build the stuff that's going to matter at the top, that's customer facing or employee facing. Or you could use our platform, which is agent force. Absolutely. But I think the market got very excited about enterprise software and they thought that AI would disrupt all of enterprise software and SaaS. You know, that's why we have the SaaS apocalypse kind of narrative. I think what's actually changed is how people will access that SaaS and how people will access those workflows. That's where people want to experiment and that's what we see. We see that everyone brings their own agents and whatnot. So build that. Don't try and recreate a platform. Like I said, someone asked me about SAP the other day, could they vibe code it? And I was like, well, well, you could, but do you want your numbers to add up at the end of every month? Really, I don't bother building that.

Speaker D: So, I mean, we could talk forever about, because we're in similar businesses. But let's finish on where's Salesforce going to go then? So even there, six years, you look out six years if you buy. It's hard to be an AI. Two, three years, what's Salesforce going to look like? What's the plan? How's it going to differ from the business you have today?

Speaker C: Honestly, I couldn't tell you. I think our whole approach is really about a deterministic platform that meets a world that is going to continually move. And I don't even want to label it, and I don't even want to label that what that might be. But in enterprise AI, you still need that agentic capability around something that is secure and deterministic at heart in the core of the enterprise. So our, uh, headless platform, I think, sets the stage for that which allows anything around the edge to be able to utilize our headless capability. It's a fascinating chapter we're in. I think even in the last 90 days, we've all got whiplash. I mean, we're continually evolving. I would never have said a year ago we would be here with Our headless capability. Even though kind of we were building for agents and redesigning for agents, Salesforce continues to learn. I think one of our strengths is that we adapt and we really tune into what customers need. We were the first to move with our headless capability. So we said you don't have to have it all on our platform or you don't have to use our agents exclusively. So I think that's kind of the direction we're heading. And I think nothing changes in the fact that we are absolutely maniacal about our customers and our customer success. And I mean that we obsess about it and everything is pivoted around that. That will never change irrespective of our technology.

Speaker D: Listen, Dara, thank you very much for coming on our show. I know you're very, very busy. It's always fascinating talking to you. We really do appreciate it. I'm sure that, that anyone listening to this or watching this is going to learn something about deploying AI. Thank you.

Speaker C: My pleasure. Thanks for having me. And by the way, a quick plug free step by step guide on becoming an agentic enterprise is on our website. So it actually talks you through the steps to take. Everybody wants that.

Speaker D: We'll put a link in the description.

Speaker C: Yeah. And we've got a CIO playbook as well, so it's really handy.

Speaker D: Zara doing a job selling.

Speaker E: Exactly.

Speaker C: But it's free.

Speaker B: It's free.

Speaker E: Absolutely. It's needed. Absolutely right.

Speaker D: Thank you, thank you.

Speaker C: Thank you. See you later.

Speaker E: Bye bye.

Speaker C: Bye.

Speaker D: Well, as ever, it's always fascinating talking with Zara. Uh, it's a parallels of our business, isn't it? Uh, like the platform, the conversations, the roi. It's the same, isn't it?

Speaker C: It's just the same.

Speaker E: This is different scales and different organization. But it's just interesting. It's just like it literally is the same thing and it's like helping like you know, she was mentioning the CEO playbook and different steps. Ultimately everybody, all the leaders have got to get their mindset engaged into what they need to do to transform and then they can bring their organization with them basically, isn't it?

Speaker D: Well, I hope you learned something there. I think Salesforce is um, amazing business that moved very, very quickly investing billions into Agent 4 and M that is acquiring businesses too. So I hope we added some value. And um, we'll see you again in two weeks.

Speaker E: See you then.

Speaker D: Thank you for joining us for this episode of the Implement AI podcast. If you're interested in learning more about how we can assist you and your business in leveraging AI for growth and efficiency, visit our website at implementai IO. Don't forget to subscribe to the Implement AI podcast on Apple, uh, podcasts on Spotify or wherever you listen to. Stay updated on future episodes. Thank you for listening and we'll see you next time.

Speaker B: Imagine the merging of trusted intelligence into a unified experience. Imagine collaboration amongst teams and across continents. Imagine an empowered ecosystem designed to deliver actionable insights that inspire growth and sustainability. That's the power of the Connect Industrial Intelligence platform to help you see further, innovate faster, accomplish more. That's the Connect effect.

Speaker E: Learn more@thatsteconnecteffect.com 500 orders a month was manageable.

Speaker A: 5000 is madness. Embrace intelligent order fulfillment with Shipstation, the only platform combining order management, warehouse workflows, inventory returns and analytics in one place. What used to take five separate tools ShipStation does in one automate fulfillment and rate shopping to ship up to 15 times more with up to 90% savings. Go to shipstation.com and use code audio to try ShipStation. Free for 60 days. This podcast is brought to you by Thomson Reuters. The best don't just do their work, they change what's possible. Cases won, audits completed, jobs saved. Behind every one of those moments is a professional who needed to get it right and did. Thomson Reuters builds the technology that sharpens insights, speeds up decision making, and powers the outcomes that matter. So when professionals act, the impact is felt by everyone. Be a changemaker visit tr.com changemakers.