B (5:59)

Ah, it's so exciting. I mean, even a few years ago as I was leading a customer service team and we're like using four tools in order to get a picture of the customer, it felt crazy, but there wasn't really another option. And all of a sudden it feels like we can have it all in one place. And it's really, really cool the way that Salesforce has been approaching it with the agents. And really this agentic AI you've been leaning hard into agentic AI and there's so much to come that I'm so, so, so excited about and we're going to get to all the good use cases and all this good stuff. But I'm curious to know. You get to see a lot of different implementations and we don't need to name names, but where have you seen organizations maybe do it wrong? Like, what are some of the common pitfalls that you're seeing people fall into as they're implementing AI that we need to be aware of and kind of move away from?

A (6:58)

Sure. This is like one of those YouTube algorithms where you're going to take it dark first. Let's get. I'm going to bring you to the light, but okay, let's take you first to darkness. And I think that's a good thing. But let's start with what can go wrong. And there's really too many of them to start. But I'm going to take with one who several months ago and they set up this really great thing. Conversational AI is in form of a bot and it looked really compelling, had a great ui people jumped into using it. But suddenly something went wrong. And what went wrong was this was a siloed application. It wasn't connected to anything else. So if you had an issue or you ran out of Runway with the bot, what happened? You picked up the phone or went into a chat session, or you sent an email, whatever you did. And they're like, who are you? I don't even know who you are. This is like the great corporate amnesia, which everyone just lows. And the second thing that we did was it. It didn't really understand the issues, really, because the. The data was not accurate. So it had a bunch of data, it worked on that data. And this had happened to be tax information. And so people were putting in their taxes and asking all sorts of stuff, this conversational bot, and it was reading stuff from all different sources and giving you answers, because that's what it does when it doesn't know what to do. You know, a year ago, if you asked ChatGPT what are bigger cow eggs or chicken eggs? It would say cow eggs. But come to think of it, that's a bovine and it doesn't lay eggs. And they. Well, thank you very much for that insight. Yay. But that's all improving very quickly. And what you saw was happening is it was pulling in information, actually from places it wasn't allowed to. It was going out to the web, it was looking at social media, it was looking at Pinterest, it was looking at Facebook, it was looking at Google searches. So the problem is. And we're going to get on with it. That's why I talk about clean data. So it was siloed. It was a great algorithm, but it didn't connect with the right information. And it also violated privacy and a whole bunch of rules that would get you shut down in Europe for GDPR or in California or any place where you really have to know the provenance of the data.

B (9:08)

I keep hearing this statement, garbage in, garbage out, when it comes to data and AI. And I really like what you're saying about, like, clean water. We need to have clean resources in this form, data to make sure that whatever system we're putting in place is working correctly. It's like the foundational element of how we can use AI and. And then make great experiences from it. I kind of want to go there right now, even though I was planning on talking about it a little bit later, but you brought it up a couple of times. And I think we just need to talk about how can companies one get the right data and know that they have the right data because there's so much data now, it's hard to know exactly what do we do with all this information and funnel it into the right places that we can create the right outputs from our AI.

A (10:09)

Yeah, and you're asking a question that could be phrased a little bit differently. And how do I do something right away with the data I have so I don't have to do that big? Just like people spent years building a big snowflake repository of information or databricks or whatever and like, well, we'll get around to fixing our customer process when we get that project done. We don't want you to do that because the tools are there now with things like Agent Force and other tools, but to do something now. So the first thing to say is, why don't we take information that you can trust? You probably have a knowledge base around simple things. So we have a client in the UK in the healthcare industry and they said we really want to jumpstart our generative AI program. Let's just take all the emails that we have answered over the last year and take all, analyze them, ask us some very basic things. It turns out there are only three or four things that people ask repeatedly on this site. So I said, that's a great thing. We know this data is clean. We know that the search doesn't go outside of this canonical database that we created or knowledge base we created. So let's point Agent Force at that. And what happened immediately out of the box was that 80% of the emails disappeared now, not 100%, because if Agent Force detected that, it wasn't really sure that the answer it was giving was accurate enough to send directly to the customer. It put it to the agent, the 8% went away, the 20% stayed. And of course it's getting better. It's an iterative process and now they're improving. But having that human in the loop so that the technology can detect. I'm not really that confident with this. I'm going to send it back and we'll talk later about our Atlas reasoning engine, which is also doing an amazing thing on helping automate that process as well, but in parallel. Then you can go a little further and say we have six projects going, eight projects going, but every one of them points back to different places in different Systems. It's in Jira, the data's in. At Confluence, the data's in SharePoint, the data's in a Salesforce CRM. Some it's an SAP or whatever. It Might be. Let's just as we do with marketing, where we built these campaigns and said, okay, we want data to come from here, here and here. That's what we're doing over here. We're saying this information, which, by the way, I can't see because I'm not a supervisor, but you can see because you are a supervisor, or I can read this, but the customer can't read that because it's a price sheet and I can use that information. So the beautiful thing that we did, building something like Data cloud, was build, if you want to think about it as a data ingestion engine, and then put all of those governors, all of those filters on there to say, which data has it been checked? Is it up to date? Is it allowed to be read by this person for this task for right now, and then we go forward with the process. So what I'm pointing out is we are telling our customers we're doing a great job with data. Don't wait until you've got all your data right. You can do probably 20 or 30 things right now with the data you have. And then in parallel, think about that strategic program you're trying to build around generative AI and agent of AI.

B (13:32)

It's thinking through of all the information that we have in all these different places, what is ready to be used versus what do we need to go through a project of cleaning and sorting. That's essentially kind of what organizations need to be thinking about. For example, we have all these emails that have already been sent that the customers already received. We can use that information as a good starting point for how we're going to respond to future customers, because we're just going to be saying essentially the same things in most cases. And then we have information from different places. What I find organizations struggle with, some of them that I work with as a consultant myself, is even thinking that different data sources could be used. Like, I think sometimes we're still kind of looking at things in the way that we've always done it before, where then there's opportunities in different data sources and information that maybe we hadn't thought to tap into before. Is that something that you run into and maybe you have some examples of pieces of information that aren't maybe obvious this to use, but have been beneficial?

A (14:44)

Well, sure. One of the things we're finding is it's imagination. This is an age of storytelling. And the reason we get up on stage, we tell so many stories. You know, you'll see if you go to Dreamforce or to A world tour. You're going to hear story after story after story. And the reason we're doing, we're trying to ignite the imagination because it's the art of the possible. And it's, it's just like with, with language. Until you're exposed to an object and then have a word for it, it's hard for you to even perceive what that is because you have no word for it in your language. So we're giving people the language, if you will, to say, hey, and let's look at customer service, or let's look at marketing, or by the way, did you know, think about you're doing service inside of marketing or marketing inside of service. So I'm going to give an example of one of our customers who their job basically is to sell alarm systems. Very simple. They also have cameras, they have alarm systems. And so it takes a while just to get a technician slotted to go out to that job. You have to look who's got the right tools, who's got the right parts, what's their schedule look like, are they on vacation, how many truck rolls do they have? So we do all that with our AI, our general predictive AI. But what we're doing right now is when they come in on the telephone, as they do right now, to ask that question, we're already running in the background a marketing analysis of that customer. So we have the data about what do they own. We know perhaps what size business they have. We see their install base. When they call and say, hey, I want to go from hardwired cameras to digital cameras, you look and say, hey, by the way, did you know with those cameras, we also have the new alarms throughout your house that are digitized. And by the way, we can also record them. By the way, we can also do the analysis for you. Now, the reason we have the permission or the customer has permission to do that is because instead of spending three, three and a half, four minutes on that call just to get the technician to go to the place, we've automated all that. So now you have the goodwill. And then you are not just throwing out one of a hundred offers, you've pinpointed it, you have made that offer really personalized. So what's the conversation here? It's about how you can do marketing right inside of a support call and that we're seeing in all sorts, B2B and B2C. But it all comes back to, you have to be able to imagine it. So we're now saying to people putting in predictive AI and generative AI. You know, you could also look at sentiment while you're doing this. So now you can start to analyze the words people are using, are those words or the frequency which they write, the cadence, their tone. And you can see, are they a happy customer, are they a content customer, are they an anxious customer? And if they're A, B, C or D, you can now treat them this way or that way. And you also take a learning what task were they performing, what process was going on during that that made them upset? And maybe there's some flaw in that task or in that process that we need to look at. So we bring that list of things in and again we have this flywheel of like Toyota says, better, better, never, best. And people would never have imagined they could do that.

B (17:57)

You're getting my wheels turning now just as we're talking about this. And I think that that's one of the beauties and also one of the difficulties of this new AI generation that we are just on the cusp of is really being able to imagine what is possible. And that's something that we like to do on this show, is to help share those examples of what people are doing so that we can start thinking. But I want to talk a little bit about how can we. Well, actually, how do you in particular help to guide your customers to start thinking about one, what is possible, but then also what boundaries do we need to put in place? Because there's always this balance of, okay, well, we could do that, but just because we can doesn't mean we should. An AI agent your customers actually enjoy talking to. Salesforce has you covered. Meet AgentForce Service Agent, the AI agent that can resolve cases in conversational language anytime on any channel. To learn more, visit salesforce.com agentforce and how do you approach that?

A (19:04)

Yeah, there are a few things in there. The first one pops into my head is don't be creepy.

B (19:10)

Number one rule, don't be creepy.

A (19:12)

Number one rule, do not be creepy. And that's a real thing because you can know something about, about a person or a business, but should you know that and should you let them know you know that? It's just like, you'll see people put anything and everything on TikTok or Insta or whatever, but goodness gracious if you actually put that in one of their emails. Oh, look at that picture of you passed out on the beach. What? Well, of course I can know that. It's publicly available information. I could pull that into your profile. You're creeping up and There's a line. So this whole thing about involving the customer, and we have to talk about that a lot because you want all this stuff at a moment when employees are afraid. You're talking all about jobs being lost or task substitution or labor replacement. That sounds like my job is going away. And the second thing, customers, consumers are thinking, oh, hmm, you're just after my data. And there's a big world between you're after my data and goodness, take my data. I give that example of Insta. You'll post anything there. But if someone uses it inappropriately, suddenly that's bad. And we're finding that businesses who earn the trust, and we're going to talk more about trust, I'm sure you're going to ask questions about trust. If I trust you, and this is large studies done by people like Boston Consulting Group, if I trust you, I will pour out 400% more information, accurate data about myself than if I don't trust you. Just like in a relationship, if I trust you, I'm all known and all forgiving, forgiven. But if I don't trust you, my lips are sealed. And I think that's one of the things. So those two things about getting trust and make sure you don't go over that board, are you always working in tandem with the customer to improve it? I think those are the two big things I'm seeing.

B (21:11)

I love that you brought up trust. It's one of my favorite topics because in customer experience, it's like the gold that we can't always see or measure because if we have trust, like you said, the customer is much more willing to share information that we can then use to help improve their experience. They're more likely to come back, to be retained to tell their friends, when we have a trusting relationship, everything runs smoother and faster and more efficiently and just better. And I think AI is one of these areas. There's definitely a generational component to it. Some of the older generations are immediately going to be less trusting. Where the younger generations are like, here have it, everyone knows it, so whatever. But there is an important role that a company plays, especially when we think about customer data, about how do we, how do we build that trust through these interactions? And I think that a generative AI in customer service environments especially there is a lot to gain and a lot to lose, because if it doesn't go well, if it's creepy, we're really impacting that relationship and that trust that we have with that customer. But if it's serving them and using their Information in a way that actually makes their experience better. We can actually build trust. So what do you think about that? And where do you see companies doing it really well versus making mistakes?

A (22:41)

Yeah. Well, what is the mantra at Salesforce? What's number one? Trust. And what we're really trying to do, all of us are trying to do the same thing, is, if I trust you, I will spend a lot more time I'll spend getting that into the DNA of everyone. The first thing is trust. You know, they sometimes look at above the iceberg and below the iceberg. When they're talking about generative AI, the first thing you. You worry about is, does it work? But I look at relationships and I say, what's below the surface? So above the surface is it's a transactional relationship. I want your money, you want my service. But it's like Peter Drucker said, the customer never buys what you sell. And you sit back and say, what? And you think about it, I don't buy what you sell. You're selling sneakers. I'm buying an experience. The reason I go with on versus this one, versus this one. And I could buy an Adidas I could. I could suck on here, buy a New Balance. Like, I could buy. They're all great, or automobiles or anything else I could buy. But there's something about the relationship, there's something intrinsic that I say, I would rather have this one, rather that one. It's no longer a commodity to me. It's personal. And if you can get into that, get into that, go beyond the transaction of selling this for this, and instead be able to, like Steve Jobs, the iPhone, it drew you in. It true you in. And it did it because they thought you're someone who deals fairly with me and I can trust you. And the perception is that you're an ethical business. Why do we have a office of ethics and humane use? Because we know, especially as you get towards Gen Z in the Millennials, in the Alphas, who knows what they're going to think? But we do know that they want you to be ethical. They do know you want to be fair. They sense you're not fair. They're out if they think you're ethical, if they think you care about your employees, care about sustainability, even though that's going to drift away, it's a desirable place to work. Guess what? There's a sense of adventure. Like, I've been at work for I don't know how many years, and I wake up every morning here, and I wow, I am so lucky I am surrounded by people with great thinking, great brains, trying to do great things. I want to come to work and that's amazing. And I also feel like we're committed to the community, right? If something's going down, we're out there. And places like Wegmans or places like Lego or places like you, you know, you name them, that's how they Muji in Japan, I think I can think of hundreds of them. People know that. And so in the perception of that is that service is not a department only customer service is us. It is what we live for and that kind of stuff. And you get all those things, it becomes a talent magnet. You know, why do we have 90 applicants for every one role? I mean, that's crazy. And we want to have people as influencers. We want to have them sharing information, sharing their point of view, giving us more insight into our product and how we can improve. The best companies that I just mentioned, all those people, that's what they've got going for them. And we think about Weg Maniacs, right? Wag maniacs. Have you ever heard that term?

B (25:57)

I haven't actually, but I get it completely because to the east Coast I was like, wow, people love this grocery store.

A (26:07)

So we had a store they opened in Brooklyn and they're only like Brooklyn. And three times more people than live in that place showed up outside in the pouring rain at 6:30am to have the opportunity to be at Wegmans. So that building that culture and there are people, you look at them go, wow, how did they do that? Right? But Ikea, you know how they did it? They had their way in Patagonia. They had their way. And I mentioned Muji. People said no, no brands, no logos, just great stuff. And of course Salesforce. So all these people, they've got that thing. And yes, they're doing their AI thing and yes, they're doing their innovation thing, but they're thinking more broadly about what sews that customer into the fabric of our being.

B (27:00)

It's an inside out job. It has to start on the inside. You can't build customer trust and have the benefits of customer trust if your employees don't deeply trust the organization that they're working for and then are committed to making that organization successful because it's something that fulfills them. It's not just a job. And that's what I'm hearing you say. And as you describe these companies, it's really something that it starts on the inside.

A (27:31)

Yeah, very much so. And people have flip flopped around and the value of the employee, the need for the employee. And especially now, as we're starting to say, hey, you know, what is the goal of this stuff, this AI stuff, this generative AI, this agentic AI? Well, and part of it, it's to lift that cognitive load off of me. I know that when young people come to get hired at Salesforce or anybody else, they come, they say, well, wow, these are the systems you use. Oh, my God, it's terrible. Not so much at Salesforce because we have pretty good systems, but here they go. It's a spreadsheet, it's field, it's a table, it's a form. That's how I'm spending all of my day. Why are you doing this to me? And we look at 40 to 60% of anyone's time at work is filled with this mundane, repetitive stuff. So far from saying, oh, we're trying to get rid of employees, many jobs we can't even fill. Field technicians, we can't even find them. Nurse practitioners, can't even find them. There are so many jobs which you can't even fill. But if we could take away, especially since the pandemic for health workers, they're burned out. So we're saying, let's relieve that burden from you. Or call center agents who the heck ever grew up. And in sixth grade, when they ask you what you want to be, you raise your hand. I don't want to work in a call center. Like that didn't happen. But today we think that's going to change because the job, we'll get to that in a bit. But all these jobs were trying to lift off the boring, put in the exciting. And for the customer, it's the other side. It's like, why do you have to do all this stuff? It's repetitive, it's boring, it's useless, it's a dead end. Why don't we change all that so that you feel that this company really thinks about me? They invest in me. So both things are happening. We're lifting up the employees. So your top 10% of all performers, we know how they are, but imagine if we could take 80% more of them and lift them up so that they can work just like that, 10%.

B (29:38)

I think that this is one of the best use cases for AI that is not getting enough airtime, in my opinion, is really how we can use AI to improve the lives of our employees, because that then gets transferred to the customer. And I think, I'm sure most people can agree with me here in that One of the things you dislike about your job the most is when you are stuck in the weeds between tools trying to find information and copy and paste things. I remember once I almost quit a job because one of my jobs as a senior manager in a company was to copy and paste 200 lines of expenses from one spreadsheet into another spreadsheet because it was critical information that could not be seen by anyone else. So I had to copy and paste it line by line. It was one of the most infuriating things I've ever done in my life to that point.

A (30:33)

They wonder why there aren't field service technicians. My furnace, I live in New England and it's been very cold. It was about 10 degrees outside. My furnace went, of course, that's when it went. Didn't go in the summer, of course.

B (30:44)

Of course.

A (30:44)

And you know, the furnace doesn't have any IoT, doesn't transmit its signal. It doesn't say what's going on. We'll get to all that stuff in a bit. But the technician finally gets there and the gentleman works on It's a great person, doesn't have the part, comes back and I finally get it fixed. And he says, we're talking about it because I like to do ride alongs with technicians. He said, you know what I've got to do now? I've got to take this whole sheet, which he hand wrote. And now I've got to go back to the depot, turn it into another person. And she is now going to enter all that information. This is the billing, this is the invoice, this is the inventory, this is the time card. Like it's 45 minutes for me, it's 30 minutes for her. And guess how many times we have errors and we're like, there's an app for that. We can turn all that into predictive and agentic AI.

B (31:34)

Yeah, I think that technology we've kind of gotten ourselves into a hole. Like the more tech we've built, the better life has become, but also the worse it has become. And I kind of feel like AI is here to generative AI is here to save the day and relieve us of the, this mess that we've created for ourselves. So it's such an important thing for organizations to think about as we implement AI is not just how do we grow the business with this, but also how do we empower our people to do their jobs in a way that they can not only be better at, but also more efficient and just enjoy their work more.

A (32:15)

I agree with you and One of the things I customers are struggling with is what are the things that I have the, the algorithm do and what do I have my employee do? And that's the big one. And you know, basically I'm like, look, you got to write these two rules right on your wall, your digital wall, your physical wall tattooed on your arm if you have to. But you know, the humans, you know, handle these high value engagements. That's what you want in there for these are moments that are strategic moments that are emotionally nuanced, emotionally complex. My house just burned down, my spouse just died and I want to change the benefits. I don't want to talk to a bot about these things. All those things that you absolutely are going to need a human, then they're going to be aided by the AI. But then the things that we talk about, I can measure this, I can literally do a task measurement and say there's tax task substitution. This is repetitive, this is predictable, this is transactional, this is data driven, this is workflow intensive. All these things are binary and the net, on the one side I have the AI optimizing the human and let's make that list and I can give examples of both if you're interested. And then on this side it's these are the things that are going to be the AI is going to do. And when there's a need, then they're going to flow to a human. But we have to get that right and find out what are the indicators of each.

B (33:45)

How can organizations set those boundaries? What is the process of getting there? Because I think it's not always easy. And I'm curious to know if you have any examples of methods or good results that you can talk about.

A (34:03)

Absolutely. Yeah, absolutely. And sometimes it's when you stop freaking over, complicating all this, why don't you talk your employees, why don't you talk to your partners, why don't you talk to your customers? Why don't you talk to your different segments of customer? You know, like if you have emerging market versus small business versus large business versus a consumer business, guess what, you've got to speak to all four. And you'll find that there are all these indicators that would say, you know, this is just about iq and IQ goes to AI. But then you say, no, this is really eq. This is really the difference between the laws of divorce and what do you think? Should I get divorced? It's the difference between laws of bankruptcy and what do you think, should I go bankrupt? You know, it's one is well, why don't you take the couch and we'll talk about this. And the other is, here are all the rules and AI is fantastic. Better than any human will ever be at many, many things. So is it IQ or is it eq? Because we are so good at body language. You think dogs are good? We are 10 times better. We, we have these social cues and we can get things right. So, you know, basically, I'll give you the example. Take a B2B company. You can throw in any kind you want. Let's say, let's say chip makers are in the news right now, right? So let's say a chip maker. And I'm making chip maker chips for the auto industry and that's cool. So I'm a stubby per company. I sell to Toyota and all good. So what happens? I have my AI powered chatbot, my FAQs, they can tell me the voltage requirement and do order tracking. Let's make, let's make the list ourselves. We know predictive maintenance, the warranty, the rma. Dude put all of this in AI. Energentic AI. But where is the human? Suddenly I've got an emergency. Turns out this chip has a defect and I want a human stat because I am freaking out. Is there a replacement part? I don't want to talk to the bot about that. When will I get it? Do I have to send an engineer or. I want a custom chip designed. What do you think? How much is going to cost? What are the risks? What's the timing? Is it a good. That's what I'm doing. I'm paying you a fortune of money. It's high touch. And that's a B2. That's classic B2B. So in B2B you're trying to give a highly personalized relationship and then you operate, operationalize anything you can. But what about B2C? Let's talk about that. And there it's the exact opposite. Sephora doesn't really know you. They have to make everything as operationally efficient as possible or they can't make a profit. But what do they do? Then they say, but how can I make this personal? So I know, hey, this is my host, guys, this skin's getting older. Who would know? He uses a moisturizer. But not only that, when it comes to the month, when there's, when they've got a sale in my email, I get a completely curated list as though they live with me. They know I buy three of these a year. This is the product they've got A new one coming out to buy this. I should buy three right now because they're 50% out. Price is going to go up. But we also have this new one. You didn't think about it, but you need something for the night. You think about the day. I feel like they know me. They've curated it, they personalized it, but they've really operationalized the heck out of it. So you know, in banking for example, we have a great banking customer checking accounts, loans, wealth management, normal thing. Who's their customer? It's not a small number of people like that chip maker. It's every. It's Main street. It's millions, it's tens of millions of people. So they have customer support and they have a formula where okay balances, inquiries, transaction history, routine account changes, fraud detection, small fraud loans. I want to put out pre fill all the forms. Small disputes, all of that. They are automating with agent force right now. And they're also putting in an agentic layer so they can launch actions. But. And what did we remove that was searching the knowledge base. It can do it faster. Finds the answer to the question. It composed the email, sent it out. It kept an eye on the customer sentiment. Are they happy or sad. It answered the questions about the bill of delivery and the form and all that kind of stuff. And it summarized the conversations. I just gave you six things which they completely remove from the day to day of every one of their people on the phone. But when do they step in? Large scale fraud, your accounts has been compromised. I don't want to get a freaking message about that. Call me. Right? High stakes financial advice. I'm taking out a million dollar loan to open a bakery or a negotiation or I have a crisis. So they what were we talking about? What goes here and what goes here? Sit with your customers, sit with your employees, sit with your different employees. Figure this out. It is not rocket science.

B (38:45)

If we were to simplify this. If I take everything that you just shared and just simplify it into what is the difference of when something should go to AI versus when something should be with a human. And what I'm hearing is that if there is an emotional component to this interaction, humans should be the ones to interact with the customer. If there's something that I might be afraid of or there's something that is not black and white and I need to have a conversation about it. Even if AI could have the conversation, it doesn't mean it should have the conversation. The fraud conversation or the fraud notice I think is such A great example, because if that happens, I'm going to start freaking out and I need someone to talk to. An AI bot isn't going to cut it.

A (39:34)

No, it's not. And if it's a $5 charge, you wouldn't freak out because it's like, oh, you know what, they charge you that because you're late by a week. You know what, let the bot figure it out. And your bot will talk to my bot, which is happening by the way. And instead I've got to call up and go, okay, we'll raise that charge. Why did you waste my time? You knew darn well I'm a long standing customer and you're going to erase it. So how about if my agent speaks to your agent and you negotiate this, you know, because you say, oh yeah, you're a good customer and you don't really do that every single month. It's, it's a rarity. Oh yeah, I am a good customer, thank you very much. Canceled. And all I get is a message saying your five dollar charge was removed. So even in that case, it really, it's contextual. Does it rise to the occasion of this? And if not, and I just slipped in there, by the way, agents speaking to agents. We're going to talk about that a little bit about where things are coming. Yeah, but there's two things we said. One is the emotional component, but the second one was the complexity. And I gave the example of the divorce and bankruptcy because there's symbolic analysis there.

B (40:36)

Right?

A (40:36)

That was Robert Reich's term, symbolic analysis. It's a, hmm. It really depends what you're trying to achieve. You know, buy the house, rent the property, put the stuff, the money in the 401k, buy a Treasury bill, take a, take a gamble on this new nuclear energy fusion company. Let's talk about that. Let's talk about your long term, you know, you're young, you're old. This, those are the kind where you say, let's pop to a human and have a conversation.

B (41:05)

What about the experimentation component of this? Because I think as everyone is exploring new AI, I guess what's the mindset? We should be approaching it with a.

A (41:15)

Few things come to mind. One is can we have a little fun here? And let's not think that everything has to be built here. And so we have companies that are light on it and they totally get that. Like I don't have the IT staff anyway. What do you got? Well, you're trying, you have to understand the buyer's mindset so people who are over on one end of the continuum. They're super innovators and they're up for taking some risk because they see the big reward of jumping out in front. They're a minority. At Gartner, we call those Type A's. You know, I'm willing to absorb more risk than normal. I don't care about de risking because I want the advantage to get ahead of you. And then you see pragmatists who start saying, I see what they're doing. I'm willing to take a little bit of risk. I don't really have a budget right now, but I can redirect some budget that I was going to put over here. I'm going to try this out and I'll even pay you some, maybe a little bit of money for it. The Type A is not going to even pay you money for it, but I'm taking risk. You're going to take some risk. And we're doing that with some of our big customers. You want to go big? Let's. Let's go big together. We'll figure out the economics later. And then over here, it's like, you put some money in, we put some money in, we make a little. You're going to get some value. We have to define the value. We're not there where it's Main street. That's a Jeff Moore term of the pragmatist, goes to the tornado and out to Main Street. And over here is the. Because the Main street needs to de risk. And they're not loving this stuff right now. The pragmatist is the pragmatic person is usually someone in it whose line of business, like the services say, hey, everyone else is doing this. This is fomo. Everyone is trying this. They're doing these cool things with email or with chat or we have one of our great clients in banking in Brazil. And they're not doing almost all their interactions through our conversational AI inside of WhatsApp. So once you WhatsApp, you speak in your normal voice. Right now, you still have to type, but we're getting to voice very quickly. You tell it what you want in natural language. Like, get rid of that stupid interactive voice response thing. Press one, press two. They were already there. They're inside of WhatsApp. Because that's the line of Japan. This is the rest of the world. We have imessage. You know that WhatsApp, all my messages from foreign people. All right, WhatsApp and all.

B (43:39)

Exactly, exactly. It's like, okay, all my European friends are in this, all my Americans are here.

A (43:45)

So they're, they're suddenly, they said, oh, we'll put that for the lower end of the market. Lower end of the market. We can't serve them with high touch. And guess what happened? All those low touch customers, the lower end of the market are just jamming on that WhatsApp, loving that conversational bot. And when it has to escalate because it's a little bit beyond what the bot can do, all the Context flows through WhatsApp to the agent and she says, hey, hi Maria, what can I do for you? I see you're trying to do X. I love it. It's fantastic because it's a continuation of the dialogue. But guess what happened now? Their high end customers saw and they go, why can't we have that?

B (44:21)

I want that.

A (44:22)

We're going to give you high touch. Like we don't care about your freaking high touch. This is something we go back, this is a binary thing. I just want to cash this check, I want to pass this money, I want to find out the current loan conditions. I don't need a human. And so they're actually driving the bank to think more and more about this stuff. So a is have a little fun, start with little use cases and also start doing your homework. Why don't we think about this from the customer journey and this is a very important part and sorry, I'm going a little bit long on this.

B (44:54)

No, you're, you're like hitting on something that is so important to me. So go for it. You can have all the time you want.

A (45:00)

How do I reach this person for this different thing, right? And think, oh, we have 45 channels. But should you really? Because they don't do that at Amazon and they don't do that here and they don't do that here. Why are you, oh, okay, we, we need all these channels. That's fine. But how do customers want to reach you for this task? And it turns out this one, they want it on mobile messaging. Whichever one they choose this one, they want to do it through your app or in your website, which is very information rich and it's a large form factor. That's two they might want to do conversational through your ivr but not through your ivr. They want to just pick up the phone and speak and say, hey, I'm trying to get a fishing license and I'm over 16 and under 60. What do I do? And it gets sent right out to me. Right? That's all there is to it. That's how I'm gonna. I'm rolling with you. It's. It's through your app or your website, it's through your voice response thing that you're gonna now evolve into voice driven, or it's going to be through one of these things. Of course, the other one is if you have Internet of things. Now the car is talking, now the bridge is talking. Now you have machines as customers. And that's a cool and very neat thing where the machine itself, just like your phone is doing all its own updates and your apps are doing their own updates. Now your printer is ordering itself its own printer fluid, and your dialysis machine is getting the reactants and the scheduling when the technician's going to come. That's the fourth case that many of us have, when you're moving into this world of the machine itself as a customer. But that's what I'm saying is figure it out. For each one of those, where does your customer want to start? So now I got the channel and now A, B, C, D, which ones? They should be no touch. So these actions right here, as we defined earlier, they should be no touch. The customers, an automated way. If we have our information architecture in order, they can look it up and get the answer.

B (47:00)

This kind of brings us back to what we were initially talking about. And just thinking of the possibility and kind of getting outside of the box is first off, I just want to reiterate what you're saying is we need to ask our customer what they want. Because so often we're like, well, it's easier for us to send this type of message through this channel, and it's easier for us to send that type of message through this other channel. But, like, what does the customer want? And we really need to ground in what the customer's experience is throughout the journey in order for it to be a great experience that builds trust and has them wanting to come back. But I think you're pushing it a level further, which is, do they even need to talk to us about this? Do we even have to have a conversation? Or is this something that we can just automate and take off their plate completely? And I think that that that's this next frontier of what AI is opening up for us is, oh, we don't have to wait until they tell us their fridge is broken because we've built a system into the fridge where the fridge will tell us if it's broken. And so now life is easier and they trust us more because we're thinking about the Customer's best interest all the way through.

A (48:11)

That's right. And we're doing that already. Think about your health. You're wearing all these devices that tell you about your heart rate or they're telling you about whatever it might be. Your body itself is starting to stream information. So when you get to your doctor for your checkup, they already know. And it's going to get only better and more and more and more advanced. All the devices are doing these things as we build new ones. There's that whole world that's coming online. And the most incredible thing coming forward is it's all going to be voice driven, just like you say, Alexa, whatever. We're starting to see that the new world of applications, the application disappears. And I think one of the most exciting things for so many people is this idea that the interface is your voice and the AI and the workflow and the accurate data in the CRM process. They're the one that fill in what we used to think about that field or that spreadsheet or that form. They're doing it on your behalf and they're negotiating on your behalf. They're doing that exchange of value that you yourself stipulated and defined and you can change it as you change, or it can actually change it for you as your needs. Just like when you put money into a 529 for a child. As the child gets older, the risk factors change and they, they automatically move the investment. That's how it's going to be with your relationship with the businesses you work with.

B (49:38)

It's exciting. Well, I think that's a great place for us to wrap up the conversation. And Michael, we have two questions that we ask all of our guests. And the first is I'd love to hear about an experience that you recently had with a brand that left you impressed. What was it?

A (49:54)

I was putting my expenses into our expense system, which I will go unnamed. So I go into my own system, lodge our expense system and I see a charge I don't recognize happens. I do my best. I'm looking through all my stuff. I cannot figure it out. So I call our credit card company. Credit card company looks at it and they can't find it. There's the date, there's the sum, can't find it. You know, I want you to say that it was a fraudulently or incorrectly and we'll just write it off like that doesn't sound very ethical. Maybe I did. So I think, okay, maybe it's from Amazon. I, I'll call, I'll go onto the Amazon site and I look for my orders on Amazon at that date. Can't find it. Something I really feel just not right about saying it's a fraudulent charge. So I think maybe I can call them. So I've never done this in my life, but I found that they actually have a chat feature. So I went into the chat feature from my order system would come in your orders. I go to the chat and the one says, you know what? How about if I call you? And I said, sure. A second later, my phone rings. So first of all, the chat launched in about 10 seconds. Next, my phone rings and I see it's from Amazon. I'm like, what the hey? And it's this woman, Dot, and she says, explain this to me. It's like, well, here's the thing. And I explained the whole story. She said, oh, yeah, yeah, yeah, yeah. You know what? That's a third party. It's not Amazon. And oftentimes the third parties take up the 30 days to submit the invoice. So let me go back 30 days. And she goes, oh, I found it. $17.20. It's this. Is this. Let me send an invoice to your. To your text message right now if that's good for you. Like, good for me. She's there. Anything else is. I wish I could give you like a digital hug. You are crazy good. This is amazing. But that's how life should be. So my own company couldn't do it. My expense system couldn't do it, My credit card couldn't. Company couldn't do it. And self service on Amazon couldn't do it, but they had a fallback for that. One in a thousand. They had a fallback. And guess what? She knew what I was looking at. She knew where I was on their site. She had all the different forms, right? She had my, my phone, she had my text, she had my records. And that's what you want. You want to be all known. You want a single point of knowledge where you can have things resolved, stats and to your full satisfaction. That's the exciting future that Salesforce is delivering.

B (52:26)

I would love to sit with Dot and see what her dashboard looks like.

A (52:32)

It was just training. She sees that, she's like, oh, and those. That's where the humans are. So good. She's ah, I could engineer that with AI, but you know how much it would cost me. And this is like one in a thousand. Let me impress the guy that we have a human touch. And the invoice to my, to my phone And I just uploaded that into our expense system, and I was done.

B (52:54)

Amazing. Do you think she knew who you are?

A (52:58)

No. Who knows? They might know.

B (53:00)

I know that this guy's going to be on a customer experience podcast, and he's going to say my name.

A (53:06)

Let's have DOT Here next time.

B (53:08)

Yeah, exactly. DOT if you're listening, please reach out. We'd love to talk to you.

A (53:12)

We love you.

B (53:14)

So my last question for you, Michael, is what is one piece of advice that every customer experience leader should hear?

A (53:20)

Yeah. Oh, my gosh. I'm going to tell you this. This is something you're not going to hear. If you're working for a company where your executive leader is not down with this initiative. Find a new job. Get out of dodge. You're probably a really talented person. They're holding you back. You're trying to soar with an eagle and you're walking with turkeys. Get out. Right. Because this is a profit first, customer second. And you want to be at a place like Salesforce, which is customer first. You get the customer by treating the employees right, you build great technology, and you're off to the races. You know, I mentioned, like, you know, Julie Sweet from. From. From Accenture or Colleen Wegman. I mentioned Wegman. They're all about treat our employees right, give them the right tools, give them some autonomy. And I did this from the top, pushing down excellence. It's when Michael Bloomberg ran New York City, he's like, no, they're not. They're not citizens. They are customers. Because realtors work here. They come in from out of town. Financial people come in here, tourists come in here. They're all our customers, and we have to. He drove it from the top until it was done. And it's never done because it's your better. Better, but never best. So that's my. My thing is, like, always be inspiring. Think about it like, it's great. This woman, she started the whole customer service for Lansand, and she was amazing because her boss knew nothing about customer service. And he came to her and he said, well, what do you want me to do? I really need just to have a customer service support thing. And we're in lands then, for goodness sake, the end of the world. What's the first thing you, as a customer experience person need to do? And she said, open a daycare center. Like, no, I'm talking about you need an IVR acd. Do you need? And she's like, no, I need a daycare center. If you can give me that. I can get the best people in town. Like, that's the mindset you want to have. Like, what is going to drive me to be the best? What's going to make me attract the best people? And then what's going to make me attract the best customer? Because every part of you, every part of you is building the brand. You either are building the brand, you're killing the brand. And I don't care if it's the person doing billing or the person like me on the phone with you right now and this great thing. By the way, thank you. Such an awesome opportunity. But if every person in the business that's your customer success, when everyone is aligned, it's like in a, a skull team, when you're in crew, if one person is rowing out of sync, you're not going faster, straighter, you're just going nowhere. So we have to kind of have everyone aligned and then you're just bound for success.

B (56:05)

Yeah. And leaders need to understand this.

A (56:09)

Well, they have to drive it. They have to drive it. They got it, they've got to own it. And that's the best companies. You always see that from the top, like, make it. So, you know, I'm, you're the experts. You in sales, marketing, service, billing, field support. I trust all you, you're great people. But there's, there's no. I don't want to play. It's like you're either on the team and you're a great member of the team or you're off the team and there's nothing else to do. Go somewhere else. We need greatness. And leaders say we need greatness. And there is no, there is no alternative. It can't hide. The business can't hide. There's no excuses. You've got to measure it, report it up. Is it making the customer more loyal? Are they attracting other customers? Are we growing? Are we lowering our costs? It's very simple stuff. At the end of the day, it's not that complicated.

B (57:06)

Cue the standing ovation. That was exactly what I wanted to hear. Thank you so much, Michael. It's such, such, such important advice, both for leaders to own that and also for employees. If you don't feel like you are supported in your role and really a part of a great customer experience, you also have a choice to go somewhere else.

A (57:31)

You have a choice. And what I love about millennials and Gen Z is that they, they will do it. They will just go. They go where they feel the passion because it's not just a job for them. The best employees we hire. It's like you're. You're so. It's so amazing because they're here for more than just the doll. They don't care primarily about the paycheck. They're here for the whole package. Mm.

B (57:54)

Completely. Well, thank you so much for coming on the show today. It's been such a pleasure to have you. I cannot wait for this episode to be out in the world, and I hope you have a beautiful day.

A (58:05)

Thank you very much.