Episode 42: How AI Leads to Marketing & Sales Efficiency

In this episode of The Sound of Automation podcast, we sit down with Michael Janney from Salesforce and Sean Myers from Wipfli to discuss artificial intelligence and how it can help marketing and sales teams become more efficient. Listen in to learn the difference between predictive and generative AI, and how to apply these tools in your manufacturing business to help your team work more efficiently

Transcript:

Sean Myers 00:00

You've got to take control of your data. You've got to invest in your data. You've got to harmonize your data. You've got to clean your data. That has to be a core part of your goal, your strategy moving forward. Not a one time deal either. It's a constant investment into your data because data is the new gold.

Intro/Outro Narrator 00:22

Welcome to The Sound of Automation, brought to you by Wipfli, a top 20 advisory and accounting firm.

Bryan Powrozek 00:42

Hello and welcome to The Sound of Automation. I am Bryan Powrozek with Wipfli and joining me today, I have two guests, Michael Janney from Salesforce and Sean Myers, one of my colleagues here at Wipfli. Gentlemen, how are you both doing today?

Michael Janney 00:54

Doing well. Thanks, Bryan.

Sean Myers 00:55

Doing very well, Bryan. Thank You.

Bryan Powrozek 00:57

Excellent. So, we're here today to talk a little bit, a little bit about automation and artificial intelligence, how some of these digital technologies in Industry 4 .0 are going to impact manufacturers and their plant floors and how things flow through there. But that's not where these tools are limited to and where the benefits are going to come from. We're going to focus today on an office application where artificial intelligence will help improve the efficiency and the operations for businesses. But before we jump into that, just kind of set up a little background, understand a little bit about both yourselves. So, Michael, can you give us a quick introduction?

Michael Janney 01:34

Absolutely. Thanks, Bryan. First off, thanks for having me on this podcast. I love doing these. And hopefully people can get a nugget or two out of out of mine and Sean's experience. So, I've been at Salesforce for over five years now. Prior to that, I was, I was the CIO and one of the partners of a of a clay mining and process manufacturing company, been around technology for 30 plus years in the application of it in supporting businesses and have driven multiple business transformations in multiple businesses. So again, happy to be here and thanks again for having me.

Bryan Powrozek 02:11

Yeah, thanks for joining us. And Sean, I guess a little your background as well.

Sean Myers 02:15

Absolutely. Thanks. Thanks, Bryan. Yeah. As you mentioned earlier, I'm a colleague of yours here at Wipfli. And I've joined Wipfli about a year and a half ago, after spending about 30 years in the software space building or creating my own CPQ software companies, had a few of those. And I joined the organization to really help drive marketing, sales and service automation forward, partnering with Salesforce and really working with our organization in our extensive manufacturing customer base and bringing Salesforce to that customer base. But happy to be here today to talk about all things marketing, sales and service, but more specifically how artificial intelligence is really making hay in that space and really helping a lot of organizations.

Bryan Powrozek 03:02

Yeah, thanks, Sean, that's great. And yeah, I love this topic because, you know, as I look at it, and of course, you know, I look at all of my clients and they're grappling with hiring retention, you know, populations are changing and they're trying to figure out how to continue to do the same level of work, but with fewer people. And that's where, you know, some of these industry 4 .0 tools come into play. But as I mentioned at the outset, the same benefits can be received in different office functions. And so we're going to focus today, as you said, on sales and marketing. But I think just to kind of set the stage for the conversation to come, Michael, do you want to kind of take us through, everybody probably has their own definitions of artificial intelligence and how it's used, but for today's conversation, we're going to focus on kind of two elements, right? Predictive, artificial intelligence, and generative. So, can you kind of just lay some groundwork and define those for us? Absolutely.

Michael Janney 04:02

When we think about AI, we think about obviously data, a lot of people, the first thing that comes to mind is maybe chat GPT, all the stuff, stuff, air quotes, that we've heard about experience in the last 18 months. But really, predictive is the first, to me, it's been out. It's an AI that's been out. I was leveraging it 15 years ago on a plant floor from a predictability and IoT perspective. To your point right now, we're really being able to leverage the data, harness that data not only from a predictive, but from a generative perspective. To answer your question if we lay out those two, you've got predictive, which is just that. It's making predictions based on what it has seen, the data that you have, you think your traditional classic IoT, is this thing going to break? It's just the system looking at the data and seeing the amperes start to go up on a motor and say, hey, you know what, you may want to do some preventative maintenance from an operational perspective that you were talking about. That's also been something that Salesforce has been really good at for 10 years now, is that predictive piece, the propensity to buy, looking at like customers and saying, well, if Sean and Michael are both customers of mine, I know they're direct head -to -head competitors. Sean buys these five products and these three solutions from us. Michael buys one product from us. How can I take that data and predict that Michael may want to buy other things? That's the predictive piece of it. The generative piece more to, I think the topic of this podcast really is, how do you take that data in the AI and leverage it for business operations on the sales and customer service side? How are we thinking about it that way? It's a combination of those two. You've got the predictive I just described. Now, the generative is just that. It's generating, it's creating new things. It could be, we can get into some examples later, but it could be e -mails, it could be just screen pops or customer service that would have taken that person 5, 10, 20 minutes to look through all the data and figure out, and let generative AI just pop it right up in front of them and say, hey, Michael's called in four times in the last five weeks. Here are the topics. These three are resolved. This one isn't, and really helping that customer experience, because that's what it's all about. It's how do you take that data, whether it's sales or service, how are we taking that data, using an organization, taking that data and leveraging it for the benefit of the user slash customer experience, internal and external. I think those are the two big ones. One's predictive, it's making predictions based on what it has seen of existing data, and that generative is really creating new things based on what it's learned from existing data. Still has to have that data, so that's the critical part here.

Bryan Powrozek 07:16

Excellent. Excellent. So, Sean, to kind of turn it over to you then. So, in your role in kind of working with a lot of different businesses and looking at Salesforce and how this can benefit, you know, what have you seen as far as, you know, practical applications of these tools? How are companies using them to improve their efficiency, improve their effectiveness as operations?

Sean Myers 07:40

Yeah, really there's one that comes to mind right out of the gate, and I think it's something that even all companies and customers that we talk to can get their heads wrapped around and it's on that customer service side of the equation. That's where it's easy to understand where the generative AI can really assist me in creating fantastic customer success and it's giving my customers the ability to connect with me for support reasons or connect with me for questions in the ways that they want to connect not the ways that I want to connect with them. So, what that means is making sure that there's a chatbot that is available online. Now, not everybody wants to engage with chatbots and that's fine then you basically make a phone call into me but there's a large portion of society that does want to engage with self -service and chatbots are a perfect example of that where I can simply solve my problems by asking, prompting the chatbot in a customer support dialogue standpoint. Others may want to connect with me over text. They want to text me. Others may want to make a phone call into the equation, but the generative AI element allows customers to engage with their client base in a next new manner and solve problems without the need to be interacting inside the organization. That's a practical application that everybody, every single one of our customers is now exploring online. Every customer that's on Salesforce anyways is exploring that to leverage chatbots into the space on the service society equation. The marketing side and the sales side, there's practical elements on that side. Michael even mentioned from a salesperson's perspective, how can I allow generative AI to advance, make relationship connections as well as to stream on some of the work that I do on a daily basis. Wouldn't it be nice that when I'm in the office an email comes up and it's already prompted to go out to that customer that Michael was talking about that is making certain suggestions on how I can advance that sales process. So, it's someone as a salesperson whispering in my ear, prompting me to engage to advance the relationship with that client base. And then on the marketing side of the equation, that's where there's a lot of fun because one of the biggest challenges for manufacturers over the years is they go to market through distribution and I can't really communicate out to my end customers because my primary customers at distribution channel, a lot of them don't even know who those end customers are. Well now with the maturity of data, I do know who my end customers are. And with generative AI, I can really be strategic in how I'm communicating with that end customer base to drive demand through my distribution channel. So those are some practical elements of where generative AI can really leverage and really advance companies go to market strategies.

Bryan Powrozek 10:45

Yeah, Michael, any examples or anything you'd want to add onto that?

Michael Janney 10:49

Yeah, I think Sean's talking about that. If you think about the blending of the two, I think from a sales and a marketing perspective, or I really should say marketing and a sales perspective, right, flipping them a little bit. To Sean's point, if you go back to an example, I was using Sean being a really good customer, buying five products, three solutions from you as a manufacturer, or even as a distributor, and knowing that, and then knowing that Michael's a direct competitor, well, why not help the salesperson, right? In the very beginning of this, you talked about the efficiencies, right, and doing the same, or maybe more with less, right, or the same with less, and you're doing more with what you have. I think that's really where AI is really going to help businesses move the needle, because if you use the example of the Sean and Michael customers, well, let the system go ahead, and from a predictive AI perspective, understand the differences between Sean and Michael. And now from a generative perspective, to Sean's point, do I internally, do I want to just hand someone an email and say, proofread this, that's going to take you 30 seconds, as opposed to five minutes to think up and right, or sit down and take your time to do it, and then hit send, or from a marketing perspective, do I just want to leverage that data automatically and create a campaign, and just shoot an email out to Michael automatically, as the procurement person within the organization, you might customer and say, hey, Michael, thanks for being a great customer of ours, you've been buying XYZ for, you know, 30 plus years, did you also know that we offer blah, blah, blah, right? And helping that sales rep, because now the salesperson isn't doing anything, marketing, you know, there's not a marketing person that's doing this, you obviously need them, because they need to help set all this up, right, from a business process perspective, but really just automate that whole process. And then, I don't know, then Michael maybe calls in and talks to customer service. And when you're leveraging something like Salesforce, you know, from a platform perspective, everyone's seeing the same sheet of music, so to speak. So that customer service rep knows exactly what happened, right? The system predictably said, hey, Michael will probably also buy these three things or four things and these three services. The system, you know, you took it over from an AI perspective, a generative perspective, you either sent the email to the sales rep to send out to proofread, or you went to that marketing campaign, and now we're getting a call in inbound. So that's from a business owner perspective and a business leader perspective, how can I drive more revenue with what I currently have today? There's a great example of that. And as a sales rep, I wouldn't be offended, hey, mark, mark it all day long for me, if I'm getting a call to come in, because I didn't know that Michael didn't know that over the last 10 years, we've acquired these other companies, we have these other products, these other solutions, and we can help them go to market in that regard.

Sean Myers 13:43

And, Bryan, if I could just add to what Michael talked about from a business strategy standpoint, we interact with a lot of business owners, a lot of chief financial officers, and there's a lot of stress on we need to grow faster. We need to, in some cases, at the current pace, I'm going to double the company in five years. That's too long. I need to double the company in three years. But what does that mean? Well, at the current process, I would need to throw a lot of resources at it, a lot of manual marketing resources and sales resources to achieve that sort of growth projections. But CEOs and CFOs are smart enough now to know, wait, technology can help me get there with the existing resource points and growth as it relates to that resource point. How can I double the company in three years versus five years? Let me lean in here on official intelligence to really help me drive my go -to -market strategies faster.

Michael Janney 14:43

And with the same people. Sorry, Bryan. Another point to that, Sean, is yeah, I'm going to double my revenue, double whatever your KPI is that you're looking at. But that doesn't mean you want to double your employee headcount. So, think about how do you leverage, when you think about software anyway, software as a service, it doesn't matter who the vendor is, right? It's an OPEX. It's an operational expense. Well, so is a human, if you think about it. Right, so which one do you want to trade off? Like, you're going to invest. There's no doubt about it. If you're going to grow your organization, you're going to invest. So, do you invest in more human capital, or do you invest in technology from an OPEX perspective? Because you're going to spend an OPEX. And I think you probably know what our answer is to part of that equation. But it's not for the sake of removing people. I think that's another misconception. I think we're going to talk about that anyway, from a trust perspective, Bryan. But another area. Like, you don't want to invest.

Bryan Powrozek 15:45

I was talking to someone that works in the automation industry and this concept of supervised autonomy, right? That you, it's not a lights out plant floor where the robots are just running 24 seven with nobody there. It's taking the human and putting them over, maybe they're managing one operation today. Now you put them over 10 operations and they're just doing the pieces that are so hard to automate that it's like, well, just let the person do what they do best. And so I think about applying that same thing here is, yeah, why should I spend five minutes drafting an email when I could spend one minute editing an email and now it's out the door. Now I can get five times as much out to prospects, to clients, whatever the case may be. So, well, Michael, you kind of teed us up. So yeah, let's step into our next area here. So, all these AI tools are great, but there are also some considerations that we need to think about before we start going down that path. So, I guess let's get into that or you brought up trust. So, let's start there. When you think of trust in the AI tool, what does that mean to you?

Michael Janney 17:01

To me, it means two things. It means trust in that my data, my company IP is not going to be violated, stolen, abandoned, or however you want to put it, taken, or disseminated out to the general public. So, I'll address that one first. From a salesforce perspective, trust is, that is the problem, right? If you think about from a generative AI perspective, as an organization, as an owner, even as a leader, I would hope even as an employee, you don't want that proprietary information going out. So one of the things that Salesforce did, and I think did really well, if I throw my CIO hat back on and look at how the last 18 months have unfolded and how Salesforce took a little bit of time up front to figure it out from a trust perspective so that we have agreements in place with the large LLMs and the large companies that are out there that are doing all the generative AI, so that information gets stripped out. So, we have an Einstein one trust layer, we call it internally. So, any identifiable information that is in your Salesforce platform is not going out to the general public. It's not going out to the LLMs identifiable, but it's leveraging the large language models that are out there and all the public data that's out there to help your CRM, your Salesforce platform from a generative perspective, make better decisions. And it's pulling it back in. It's not being retained by the LLMs. So, they're not learning from your Salesforce data that does go out. So that's one aspect of trust, right? And we took our time figuring that out. We also had to think about the biases and the toxicity and the hallucinations that we've all heard and seen. There's the recent one about Air Canada. I don't know if you saw that where there was a guy who, you know, John, you're talking about bots, right? So, this passenger went online, was chatting with the bot, was finding out about a bereavement fair, how that worked, how they get a refund. Well, the chatbot basically said, go ahead and purchase your ticket, then submit a refund and you'll get money back on your flight. And so the guy did that, followed exactly everything that the chatbot told him to do. And Air Canada came out and said, well, no, that isn't our policy. But he's like, hey, but I have it. So obviously, you know, the court ruling came in his favor because the chatbots, so there's a hallucination. There was never a policy in place that Air Canada said, oh yeah, from a bereavement perspective, we'll refund your ticket. So, it made it up. It didn't know what to do. So those are the kind of trust things from a data trust perspective. But I also think more importantly, or just as important, I would say, is the trust of the people, right? Like the people that are watching, the workers that are watching AI come in, the sales reps, the BDRs, right? The customer service reps, business development reps, all these people seeing AI come in and think, well, it's going to take my job, right? So, there's a factor there of trust. And I would say that it isn't. It's going to create more jobs. Initially, it may displace positions. Ultimately, it's going to create more jobs long term than, and more impactful jobs. And I'm going to give you a quick example. You know, I'm going to go back to years ago, and this is probably age me and, you know, show my age here. But, you know, my dad was, he worked in at the, he worked in a newspaper in Baltimore. I'll leave it at that. There's only a couple of them. But he worked in a newspaper in Baltimore. And I remember when I was little, they got they just got computers. And I remember my dad, what they used to do is take an exacto knife, and cut out and do the page layouts and the paginations of these pages before it went into print. So, computers started coming on the scene. And you know, maybe it's genetic, I have no idea from a technology perspective. But my dad was like, hey, that's really cool. I want to learn that. Because he could repaginate a page in two minutes that used to take him 25 minutes. So, he didn't fear that the technology coming in, right? But just so that's another trust layer, you know, trust factor for me, it's not just your data trust. It's a human trust that, you know, hey, yeah, this is going to make my life easier as a salesperson. Yeah, this is why I'd want to put data into Salesforce as a CRM platform so that generative and predictive AI can take over and then spoon feed me qualified leads. Well, yeah, why wouldn't I want to do that? But they have to wrap, I think we just have to have our culture wrap our heads around that and not fear that.

Bryan Powrozek 21:37

Yeah, no. And I would almost add another, I guess it's maybe just a different shade of that same trust, but, you know, trusting that that email it's going to generate is the right email. And that's where, you know, the, the individual reviewing it and going through and adding their piece to it is important. But, but I've seen that, you know, in, whether you're talking about AI or I've seen it with clients on, um, you know, like dashboards, KPI dashboards, that all of a sudden the dashboard comes out and people say, well, that's doesn't, I don't believe that that's not what I'm used to, but in reality, they're now seeing the data correctly for the first time. And so same thing here, you know, we've gotten our heads that, oh, this customer needs X, Y, or Z. And now the, the AI comes in and says, well, you should really talk to him about A, B, and C. They're like, wow, no, that's not right. I, I don't trust the tool. So, uh, so that Sean, then I think leads us to another point that, that you were going to speak on, um, and I know you spend a lot of time with, with our clients as they're going through these implementations, but this is all based upon the data, right? So, let's talk about the, the cleanliness of data, the accuracy of data, and how that, that impacts some of this.

Sean Myers 22:46

I will do that, but if I may, I also want to address, you know, just point out something from what Michael had mentioned. He said that Salesforce 18 months ago was looking into this, and they took their time. Just think of what has happened in the last 18 months and the fact that he says Salesforce took its time. Well, nothing was ever evolved and developed within that light speed of just pure speed, go -to -market than what has created since generative AI came on their scene 18 months ago. In an organization like Salesforce, you know, one of the largest software companies in the world was able to pivot and go all in on the fact that the AI element is really going to, you know, be that next generation aspect of things and pivoted the entire organization from the number one CRM company in the world to the number one AI platform in the world and have generative AI inside of all of its different products. The only point there is 18 months ago, that's a blink of an eye and that's the pace that this is coming at us, all of us, and we have a perspective like Michael's dad did of, wow, okay, that's really cool. I'm going to, you know, it's going to make my job a lot easier and lean in and go or we're going to be sitting on the sidelines and watching this kind of fly by us and lose a lot of market share, which I know we'll talk about, but the data element of what you said is all of it is predicated on good data inside your organization. So, I do an AI talk within the plastics industry and I end with what you can do today. Well, you need to pay attention to what the vendors are doing from Salesforce and others as it relates to artificial intelligence and what that means to your organization. That's number one, but what you have control of and what you must do is take control of your data because if you've got a messy data base that has duplicates, massive amount of duplicates, this is not going to serve into the AI elements of what the software vendors can provide. So, you've got to take control of your data, you've got to invest in your data, you've got to harmonize your data, you've got to clean your data. That has to be a core part of your strategy moving forward. Not a one -time deal either. It's a constant investment into your data because data is the new gold and you've got to own that, you've got to harness that and pay attention to that. So yeah, that's a focus there. Yeah.

Bryan Powrozek 25:30

And I really like the point you let off with. And so I want to go back and just hit that again because as you mentioned that the pace of change, and this is one technology, one tool set that's changing at that rapid pace. And everything in the business world is evolving at similar rates, right? So, I guess curious to hear from both you and Michael, you know, your thoughts on business owners and business leaders, they can't sit back, right? Because if they decide, well, we're going to, we're going to let other people figure that out and then we're going to jump in five years from now, you've got that whole learning curve plus, you know, kind of like Moore's law, right? The pace of change is continuing to go faster and faster. So, you're, there's no reason to wait on getting involved in this because you're just going to find yourself further and further behind the pack. So, curious your thoughts, you know, for business owners and leaders out there.

Michael Janney 26:33

Yeah, that's a great question. When you think about the data itself, and keep in mind, from a generative AI perspective, generative AI has only become new to the general public in 18 months. It's been around. So, it's almost one of those things where you think about, you see someone that's successful, again, air quotes. It didn't happen overnight. Generative AI didn't happen overnight. It's been refined over the years to actually get to where it is today. So, from a get started perspective, first of all, anyone out there that has multiple siloed data platforms running their business, which is every manufacturer and distributor I've ever met, to think that you're going to clean those platforms to clean that data, it's not going to happen. Sorry to break the news to you. That's highly improbable to be able to actually successfully do that, because once you do clean the data, then you have to worry about data governance. So, you have to have all that input. So, you can go clean your data. In 18 months, it's probably going to be dirty, if you want to call it dirty again. So really, it's putting a system in place now and today, to Sean's point, like you can't sit around and wait for this. This is not a wait and see type thing, because it's already been in use and in production and being figured out for the last 10 years, probably, from a generative perspective, which is the general public learned about it 18 months ago. And when you think about the data, and it's one of the reasons that Salesforce in particular, we're honestly very excited about what we can do with Data Cloud and our data solution, right? It would stem from a CDP perspective, from a marketing and a customer data perspective and harmonizing the five different Michaels, Michael Janney's that I have in my five different platforms into one and knowing, would the real Michael Janney, please stand up for those of you who get that reference. So that's been around, right? And we launched that product about four years ago from a marketing perspective. And now we're harnessed around being able to leverage that the data that's there and coming from these different systems today. So, the reason I'm mentioning all that is because leave your, I had this conversation last week with a chief strategy officer for a global organization. And the conversation was we stemmed around, hey, he said, hey, I have these literally, I have these 12 Oracle EBS ERPs. We cleaned it up, you know, two years ago, and we cleaned it up seven years ago, and it's a mess again. What's the strategy around that? And really the strategy around that is take something like data cloud, leverage the data out of those ERPs, but only pick the good data that's out of those and expose it into something like data cloud, right? Or Snowflake or whatever, and being able to leverage that data, but leveraging data cloud to be able to harmonize that data from those different ERPs that have different customer IDs and harmonizing into data cloud and then being able to leverage that, that's really where the rubber meets the road in my opinion. It's not trying to clean up all the bad data because you're not going to be able to do that successfully long -term. So, it's coming up with a different solution. It's coming up with, hey, yeah, I need this other data layer and it's called data cloud. And I'm going to connect data cloud into Salesforce to be able to drive my AI predictive and generative analytics. I need something because I have to have good data and there's no way I'm going to get there. And that's a hurdle that I hear from CIOs when it comes to, where do I start? How do I begin? I have these 12, eight, 10, two, five, 14 ERPs. I don't have the time to be able to figure this out. And that's where you leverage something like Salesforce's data cloud to be able to do that because data is the new gold. And Sean, you said that, right? And being around technology as long as I have, it's the new gold. We as a species, honestly, did not have the tools to be able to really refine that gold like we have today, whether it's from Salesforce or others. The technology there is now to be able to leverage that data and make actionable insights in the predictive and the generative AI from a lot of organizations out there.

Sean Myers 31:08

And there lies the value to the pace at which you can really advance your go -to -market strategies. Data's the new goal. Now I've got the technology from folks like Salesforce who can take hold of that data, generate, get the power of generative AI and have additional touches. You know, in selling, there's always the role of seven, there's the role of nine, there's the role of 10. At the end of the day, successful salespeople know that it takes a certain number of touches to a prospect in order to win them in business. That's never going to go away. That trust relationship sort of scenario, relationships are how you win business, relationships are how you win business tomorrow. The question is, how can I accelerate that process? And from a company's perspective, how can I accelerate that process without, you know, exploding my marketing and sales staff to facilitate that with traditional go -to -market strategies? So, executives are not sitting on the sidelines anymore. They're leaning in. It's the topic of, you know, visage groups and EO groups, you know, it's how do I leverage AI to really accelerate my business?

Bryan Powrozek 32:23

Oh, fantastic. Well, Michael and Sean, I really appreciate the time and the insights today. I think this was a very good discussion and a lot for people to chew on. So, thanks for both taking some time and sharing your thoughts today.

Intro/Outro Narrator 32:37