Victor Antonio is a sales legend, he’s an A.I. in sales expert and he has a killer Youtube channel too.
On this episode of The Salesman Podcast, Victor explains what A.I. is and how it’s going to help you close more deals in the next few years.
Featured on this episode:
Resources:
Transcript
Will Barron:
This episode of the show, we’re diving into artificial intelligence, what it means for you as the salesperson, what it means in the business world, in a whole, and how it’s going to help you close more deals, easier. Hello, sales nation, I’m Will Barron, host of The Salesman podcast. And on today’s show, we have the legend that is Victor Antonio. He is a true expert in the world of AI and sales. His YouTube channel is phenomenal, just Google or YouTube Antonio’s a tonne of great content on there. In this episode of diving into AI and a whole lot more. To all that said, let’s jump right in. Victor, welcome back to The Salesman podcast
Victor Antonio:
Will, thank you for having back. Always great to be back, man, love your show.
What is Artificial Intelligence? · [00:52]
Will Barron:
I’m glad to have you on mate. Diving into artificial intelligence today. And so that’s the first question. The first place we’ll start. What the heck is artificial intelligence?
Victor Antonio:
It’s intelligence that’s artificial. How’s that? I mean, that’s the smart answer. I mean, everybody has their own version of what it is, but basically you’re trying to replicate the human cognitive brain and trying to substitute that with the machine. That’s my simple definition of it. How do we substitute the human mind with the computer?
The End Goal of AI and Machine Learning · [01:10]
Will Barron:
Okay. This is where things get complex. In my mind, when I’m reading about this, reading your book, which is on the table and watching videos and trying to research it, because it’s fascinating to me, we’ve all without sales and business in the conversation, artificial intelligence how it’s going to dramatically change our environment, how we interact with each other. And we’ll dive into this as we go through the show, when we see we’re replicating the human brain is the end goal to have a box that is literally somewhat conscious, whatever that means or are we trying to just do… Are we trying to just systematise some of the stuff that we can do and make it better?
Victor Antonio:
Yeah. I mean, to really understand that… To really try to put this in a box, no pun intended. Let’s start out with how it started out. Years ago, when I first started with artificial intelligence, it was called expert systems. These were rule based systems. And basically you write rules, if this go to then, but don’t do. That whole thing. And so what you try to do is basically take somebody’s domain expertise, plug it in, bunch of rules, very algorithmically, and the machine would actually respond to that. Now what happens over time is obviously that was a very laborious process. You can imagine just to try to mimic somebody’s domain expertise can be millions of lines of code. And you still don’t have that creative part. That intuitive pieces, even in there.
Victor Antonio:
There were limits to what you can do with a rule based system. And that’s why the initial spark of artificial intelligence was like, it kind of died, in one book I read, it said it went through a nuclear winter. Because people basically said, “This is never really going to work. It has a limit.” And there were several limits that it had. And the three that people point to are processing power, being one. Algorithmic power, programming power being two and three was the amount of data. This is pre-internet, I’m talking about. You got processing power, algorithmic power, and the amount of data in the world. Now, fast forward, and now we have these new technologies. Moore’s law has been kicking in big time, we’ve been doubling our processing power.
Victor Antonio:
And all of a sudden we have these powerful machines. I mean, our phones today are probably the equivalent of what a Cray supercomputer was back then in the 80s and 90s. And so the processing power is there now, on top of that, everything is now connected wirelessly, which means we have the internet of things, which means data’s just massively flowing into this stuff. And now we have this processing power with this data, and now we’ve figured out how to write algorithms that can learn. This is that machine learning phrase we often hear. We have artificial intelligence and under the rubric of artificial intelligence, you have, let’s say, machine learning, neural networks. Natural language processing, visual process, visual recognition, all this stuff falls under artificial intelligence. And again, what we’re trying to do is replicate the human mind. And so what we’re seeing now is I think it’s real this time in the sense that it’s going to be more effective than it was 20, 30 years ago.
Victor Antonio:
But then the question is, how far can it go? Again what are the limits, the boundaries of artificial intelligence? Would it reach that point of what people call singularity when it almost has its own consciousness, when it could actually think for itself? And people put it out at 2050, some people put it out at 2070. The real answer is no one knows. All we can see right now is that we have machines now that can process data at an amazing rate and give us insight and just information that we could never garner for ourselves.
How AI will Permanently Change the Way We Sell · [04:41]
Will Barron:
Okay. What does this mean for sales people and people just working in the business corporate environment in general, from a context of, I know if I’ve got a huge list of data, I’m probably never going to bother going through it. It’s going to be quite rudimentary. And for a practical example, I guess here we had, I think about 7,000 replies to a questionnaire that I did towards the daily salesman, which is this newsletter that we put out. 100,000 people on this list now. 7,000 replies, really interesting data. I’ve got no idea what to do with it. It’s probably really useful for some company. I promised the audience that I would never sell it.
Will Barron:
There’s the kind of the financial element of it is taken off the table for the moment anyway, on that perspective. But for someone like me, perhaps let’s scale this up to a bigger company like Salesforce, we’ve got their mug on the table there, they’ve got hundreds and hundreds of thousands, if not millions of potential data points, what can machine learning? What can AI? What can with the processing power, the algorithm power and the data that’s there. What can be pulled from that will help us salespeople, selfishly, close more deals and best serve our customer?
Victor Antonio:
Let’s start with the basic, we used to use spreadsheets. And spreadsheets were somewhat manageable. You had 20 rows, 20 columns. You can deal with that. When you want to understand machine learning and what’s happening today, now imagine a spreadsheet that has… I don’t know, a million rows and maybe 700,000-
Will Barron:
Let me just jump in here for a second to context. Don’t lose your train of thought here. But this spreadsheet I’ve got with 7,000 responses with, I think like between 10 and 15 questions in, crashes the workstation that I’ve got over here that has 128 gigabytes of Ram has two 12 core processes in, it’s what use for video editing. And it absolutely dominates, crashes, smashes this computer system, just for context, for the audience of this lowly kind of spreadsheet with 7,000 responses in isn’t processable on anything other than perhaps a cluster of, or a server system as well. When we’re getting into the numbers that you’re talking about, then this is genuine big data, right?
Victor Antonio:
This is genuine big data. The numbers are just so far out there that, again, from a human conscious standpoint, there’s no way to process it. And so how does it apply to sales people. I think the best analogy I’ve come up with Will to really understand what’s going on. And it’s very apropol that you have a Salesforce mug on your desk right there. Because I think if people look at… And I’m drawing an analogy here, if we look at the iPhone. Steve jobs came out with the iPhone. And then once he came out with the iPhone, it’s a very powerful system. Different gooey, different interface, and we can interact with it differently. Then he opened up the operating system, so people can write apps on top of it. What’s happening now is that Salesforce…
Victor Antonio:
I mean, Microsoft is trying to compete, but Salesforce right now dominates with 20% of the market when it comes to the CRM market space. If you can kind of look at Steve jobs and then Marc Benioff, who’s the president, CEO of Salesforce, think of him as Steve jobs for the CRM system. He’s created the CRM. That’s a very powerful CRM. Now what’s happening is that they’ve opened it up where you can actually bolt in apps onto the CRM, depending on what you want to do with that data. You can do almost say anything, just like you do on an iPhone. This is what’s powerful. You’re actually turning machine learning algorithms into drag and drop apps where you can just drag your data and it can spit out a piece of insight that you would never have imagined, deriving from any type of spreadsheet. And does that make sense, Will? Because I think that’s a good visual analogy.
What is Machine Learning? · [08:25]
Will Barron:
That makes sort of sense. And I just want to touch on this because we mentioned it a few times. What is machine learning? And I guess we don’t need to go into the algorithms behind it, but is there a visual way to describe that? And then what would be a few practical examples of it? Just to paint a picture in the minds of myself and the audience?
Victor Antonio:
No. Break the two words apart, machine and then learning. What happens is when you have a lot of data, so let’s say you have this 7,000 person questionnaire. And so what I’ll do is I’ll take… I’m going to split that data. I’m going to cut it right down the middle. I’m going to put one over here. We’re going to call that the test data, just kind Chuck that over to the side, what I’m going to then do is take the other 3,500. And what I’m going to do is I’m going to write an algorithm. A programme. And what I’m going to do is I’m going to feed data. Now I should know what comes out, the expected output I should know what that is. I will feed data to get the expected output. If I don’t, then I have to go in there and mess with the algorithm.
Victor Antonio:
What I’m doing now is training the algorithm to give me the correct output. Basically, I am training the machine, the programming itself. Now what then happens once I think I got it trained, the machine has learned, I then take the other set that I set aside to test data and I put it through the system. And so if I get the output I want, I now know that the machine has actually learned. Now what’s fascinating. Take it to the next level is now machine learning has to deal with streaming data. Every time we’re interacting, I’m sending you an email on salesforce.com. “Hey, Will, what about that RFP I sent you last week?” You say, “Hey Victor, we’re still thinking about it. It’s in committee.” I respond back. It can take that interaction. Literally take that interaction. It can look at emotions, sentiment.
Victor Antonio:
It can look at the length of the email. It could look at the headline of the email, the time it took for you to respond with that email and the machine will then make basic assumptions through predictive analytics. It’ll make basic predictions on whether that is gone to close or not close. And over time, because we’re pushing through so much data through the actual machine. It’s learning what works, what doesn’t work, what data points matter, what features, attributes matters, what doesn’t matter. And that’s where the machine begins to learn. And that’s what’s powerful about what’s happening today.
Will Barron:
Which is-
Victor Antonio:
Too much?
Gut Feeling Versus Machine Learning When Trying to Predict Sales Outcomes · [10:42]
Will Barron:
No, not at all. Which is more effective then? If we looking at things like, and that also build up into picture in a second, when we’re looking at things like how quickly someone replies? The language that they use, whether they’ve CC’d anyone else on the email, this builds up a picture in our own minds. And perhaps we’re not consciously processing it, perhaps we’re subconsciously processing it. It builds up a picture that gives us a gut feeling as whether this is going to close, whether we need to do more work. And this is through 200, 300,000 years of human interactions and genes being passed down and kinetic changes that has allowed us to achieve this. How effective is our gut versus the data that machine learning can implement and pull over time?
“What machine learning, artificial intelligence does is it removes that subjectivity and it looks at actual data. It’s looking at so many data points that you couldn’t do it yourself, but it’s giving you its best assessment. The more data you feed it, the more results. For example, I feed it information on the deal closed, deals that didn’t close – closed, not closed out. After I’m feeding it so much information, it starts to learn what deals are more likely to close than not.” – Victor Antonio · [11:47]
Victor Antonio:
Oh, our guts are horrible. Our guts are horrible. I mean, you know what I mean? The thing is we have either, selective perception or cognitive bias. Those are the two things, we try to confirm what we want to confirm. We’ve all been on that sales call where it went really well. And we think, “We nailed that sucker.” And then two weeks later, it just goes dark on you after all. And so what machine learning artificial intelligence does, is that it removes that subjectivity and it looks at actual data. Again, it’s looking at so many data points that you couldn’t do it yourself, but it’s giving you its best assessment. And again, the more data you feed it, the more results. For example, I feed it information, the deal closed, the deal didn’t close, closed, not closed out. After I’m feeding it so much information, it starts to learn what deals are more likely to close than not.
Victor Antonio:
Back to your original question, how does this help salespeople? Well, if the algorithm is now learning through every interaction, then what happens is they’re able to… You hear the phrase, lead prioritisation. They’re able to prioritise your leads. In other words, who to go after. And this really the callous for the book, Will, I did an event for a company, I was in South Korea. It’s a big company. And they took me to their telemarketing centre. Just to paint a visual for you, 200 people in that room. Each making about 100 calls a day, run the numbers. That’s 20,000 calls. But let’s just stick with the individual telemarketer. 100 calls per day. Now what they demoed to me that day was something fascinating. They recorded the audio of me talking to you.
Victor Antonio:
Let’s say, I’m the telemarketer I’m calling Will. I’m trying to sell you. They recorded our audio tracks. What they then did is that they grabbed the audio tracks, split the audio tracks and they grabbed your audio track, Will, the consumer, the potential buyer. And based on the words you used, the tone you had, the keyword phrases that you use, the frequency of words that you use, the machine was able to say, “Yeah, Will deserves to call back because I think he’s almost ready to close.” Versus those who are not. And that was fascinating. Their close rate went up like 67% using machine learning to tell them who to call back. Because again, let’s go back to it. If you call 100 people, you close 10. That means I got 90 in the bucket. Who do I call? Who do I call first?
Victor Antonio:
And so now you have a machine telling you. Even if you believe that you have this, you’re like, you got that Malcolm Gladwell blink inside of you where you can thin slice the world and figure it out. Well, guess what? When you’re doing 100 calls a day, there’s no way. It’s in human. On the other side, on my side, they looked at my data, my voice track rather. And then they analysed what I said in response to your questions or statements. And it would actually tell me what modules in terms of sales trading, I needed to go back and review. It was also used to review the actual salesperson or telemarketer in this case. I mean, that gives you an idea how powerful this could be when just used for lead prioritisation.
How Soon Until an SDR is Replaced with a Machine · [14:28]
Will Barron:
If we project perhaps into the future, and I’m going to put you on the spot and ask for a date or time here, and I know the answer’s probably going to be, “Who the hell knows? We’ll come down in a second. But this over email seems to make total sense, emails, more black and white. Perhaps you don’t put as much emotion into an email, both the salesperson and the recipient of the emails. And it’s probably easier to digest. I imagine from a machine learning standpoint, there’s less nuance to it. I can imagine that if it doesn’t exist already, as in, there’s not some blacked out Silicon Valley startup that it’s been funded, that’s doing this. Because I don’t know across any… I’ve not come across any companies that are publicly doing it just yet. An SDR, booking meetings for an account manager through email and no calls that should be wiped out.
Will Barron:
That shouldn’t be long before we can… We could all… I guess, imagine this happening now. How long are we away from… Or how far are we away from this SDR on a phone call being wiped out as well? If a machine can understand context, nuance, if it knows when’s the best time to ask certain questions and especially in a call centre, for example, when people are just using a script anyway, and they’re told not to deviate from the script, because the script works. How far are we away from… Because voice synthesis technology is already there. You can synthesise someone’s voice, more accurate than the FBI and NSA can differentiate between the real and the fake, which is another weird issue that we’re going to come into down the line of evidence in court and all that side of things.
Will Barron:
But how far away are we from there being a software as a service that is an scalable SDR team that can call people, prioritise the leads, get the information that’s needed to perhaps not close a sale, but to book a meeting with an AE, with an account executive, account manager who will then perhaps add some nuance to the call and build a relationship. How far are we away from the SDR in a box?
Victor Antonio:
What’s funny, you mentioned two things. I’ll answer the meeting one real quick is interesting. Now there’s a company out there that actually has a chat box that would actually… Once I start the conversation with you it’ll go back and forth, to set up a meeting and no intervention is required. It’ll just go back and forth and actually be able to interpret that language. That’s already fascinating in terms of when will we be able to get rid of SDRs? I think the unknown variable, besides obviously everything that has to do with programming is the fact that price point and risk will always be something that’s very human. If it’s a simple transactional sale, then you know what, it’s not a big deal. I’ll let the machine do the transactions. It’s like when we go to Amazon.
Victor Antonio:
If it’s something really inexpensive, we can just probably do it via and we know we can return it. But I think as the price point moves up, I think that’s where the SDR is not in jeopardy. I think that’s where they’ll always be around because you still need to hear that human voice I think sometimes, and you can argue with me that, Victor, 50 years from now, you won’t be able to tell the difference. From vocally, you won’t be able to tell the difference and you may be right. How far out, who the hell knows, call that… But are we getting closer? I think so. I mean, for transactional sales, we’re getting very close. It’s getting scary close. I’ll give you one example. I ordered a Spark drone. I can say the Spark drone.
Victor Antonio:
And I remember, it wasn’t delivered on time. I’m trying to figure out where this thing’s at. I get up online, I start… I went into their chat session. And it wasn’t until about maybe four or five interactions, I realise, “Wait a minute, this is a virtual assistant. This is a Conversica product.” One that I mentioned in my book and it wasn’t until five interactions that I didn’t realise that I actually got into… I was really having this conversation. It’s scary that it may be closer than we think, but again, I think SDRs will be around for a while. At least this generation doesn’t have anything to worry about.
Chatbots For Sales Automation · [18:26]
Will Barron:
I love this. Chat bots is what I was going to come onto to. We’ll dive into this now. I created a chat bot for… I’ve started another podcast called Excited Science, kind of a passion project hobby. And on Facebook, I’ve got a chat bot on there, which gives just super, simply a science fact every day, nice little animation, a link to the source. And you basically nine o’clock every morning, depending on what time zone you’re in, you get bing and you get a nice little science fact just to start the day. Just interesting. No value other than just pure entertainment. I wrote the chat bot software or the implementation is an online service that I use. I think it’s called Chatfuel, chatfuel.com. I wrote in the, yes, no answers to the questions that typically could come up. When someone says a specific keyword, it will say specific sentence back to them.
Will Barron:
And I thought it was pretty simple, took a couple of hours, literally a couple of hours. That was it. Then I had a huge problem. People started thinking it was a real thing and started asking me questions of I’d love to see this guest on the show. I would love to do this and because my faces plastered over the Facebook page. People thought it was actually me. I had to then go back and when it was a question that wasn’t pre-programmed I had to put a note in saying, “This is an automated bot. I’m not reading this. There’s too many responses for me to read.” I had to dumb down this chat bot that only took me two hours or so to put together in the first place. And it wasn’t two hours of refinement.
Will Barron:
It was literally me going, I wonder, what’s the most common things people’s going to say. What’s… And for example, swear words, lots of people started swearing at it. And so I come up with some smart comments of when a swear word pops up, it says this smart comment back, and then all of this, then people would get that smart ass comment and start asking me questions, thinking that I’d written it, every single time I’m sat there in front of a computer, just replying to all these comments. That got me thinking all of this. Is there… Maybe a response to this, maybe an answer to this. Is there any way to build or are there any platforms out there that have virtual assistant a chat bot for sales people?
Will Barron:
Because I feel this should be really handy rather than go back and forth over email rather than fluffing round of, for me medical device sales, it would be a surgeon I’d be on the road driving towards them. And then they’d email me saying, “I’m going to be five minutes late. Can you meet me somewhere else?” Then I’d get there because I’d not check my emails when I’m driving and it’d be messed up. Are there any platforms for sales professionals perhaps within the Salesforce kind of world and cloud that would allow them to have quicker engagement with potential customers or actual customers and account management’s perspective when they are unavailable? If that makes sense.
“The thing is, a lot of people think that artificial intelligence or machine learning is something that you wave a magic wand, stick data in the box, and stuff starts flying out. The reality is that the usefulness of that bot or any type of machine learning is only as useful as how organised your data is.” – Victor Antonio · [21:23]
Victor Antonio:
It does make sense. I don’t know, off the top of my head as far as on the run type of chat box, but you bring up a very interesting key point. And that is when you were programming that chat bot you had to put in your data. Took you two hours, as you said to put in that information. The thing is a lot of people think that artificial intelligence, machine learning is something that you wave a magic wand, stick data in the box as stuff starts flying out. The reality is that the usefulness of that bot or any type of machine learning is only as useful as how… In terms of how organised your data is. Let me you use examples, for example, this telemarketing company, if the answers are very standard. People call in, “Do you have it in blue?” “No.”
Victor Antonio:
“Do you have it in red?” “Yes.” Or whatever it may be. If they’re very standard, then a chat bot can deal with this. But once it starts getting off the reservation into deeper questions, unless it’s already programmed in, the chat bot won’t know, it’s not at a point yet where it can… How do you say, intuitively grasp different conversations and create its own language and spit out a unique answer. Everything is still really programmed, but it’s really trying to combine the best of the best data that it has. And I think this is fascinating to point out because I think people expect a lot of magical stuff with AI, without understanding that if your in-house data, your answers, in other words, how you train your agents today, if you don’t have a sales training process, then it’s not going to work for AI. The more detailed your training process is within your company, the better you’ll be able to train the machine. They’re linked.
The Other Side of Chatbots: Risks and Disadvantages · [22:50]
Will Barron:
Would you say then perhaps from the conversation, if we’ve got the process of power, if we’ve got the algorithmic power, if we’ve got the big data, is the thing holding us about right now, the risk of launching a chat bot out there and it starts doing some really weird racist stuff. And it really affects the corporate brand because it’s pulling in conversations at different places that doesn’t have to be racist, but does something that is a human wouldn’t do because a human has rules and fundamental governances that are built up on societal pressures and things that they’ve learned over the years where a chat bot that looks at data can link to the internet and look things up on Wikipedia doesn’t have, is it the risk of corporate embarrassment perhaps that if these other three boxes are ticked, that is holding us back from just going all out in this area?
Victor Antonio:
Yeah. I don’t think we’re in danger of a chat bot developing an attitude or some biases in the near future-
Will Barron:
It happened on Facebook recently. Facebook had this problem and it was blown totally out of proportion, but they had the AI systems basically pushing out content, anti semi content to people who would clicked on these articles. And then that was a huge fuff around and there’s issues with that. It is literally happening right now within our world as we live so far.
Victor Antonio:
Oh, I thought you meant that the actual… The responses were actually going to be produced by the actual machine that were going to sound racist. I’m like, “That’s a little bit much.” I don’t know much about the Facebook. I remember seeing the article, but I think that was more, whatever you’re interested in is going to push something out. I don’t know, but all I know is today, we’re nowhere near. When you look at just the way we talk, the different… The velocity tone, the different accents we have, it’s still having a hard time grabbing that information and processing it, just putting a sentence together, structuring a sentence together, for us it’s very natural. For a machine, it still takes a lot of effort because, again, if somebody says, “Give me a blue dress shirt.” For example, just say blue dress shirt, what do you want? A blue shirt or a blue dress?
Victor Antonio:
Do you know what I mean? It still has trouble picking things out like that in context. And so it’s not there yet where it can actually develop its own language. Now that said there’s a website called Quill. Quill can actually now produce their own articles. Now you have these machine learning programmes that can actually build articles for you. Now, again, you have to train it a little bit and provide a template for it, but once you do so, this thing can spit out articles for you. And I think large companies, I think like CNN are using it and other media outlets because they’re trying to produce a lot of content every day.
Will Barron:
Yeah. I know in the Olympics, a bunch of companies, including the BBC started using these algorithmic kind of… It would just literally spit out, “This team won, this individual got a gold medal.” There’d be machine learning then to find a picture of that individual. Credit the picture. And that would be the article that then would be streaming on the front of the BBC’s website. I guess that’s… This is where we wanted to go with this. And that’s kind of a good turning point here of that gets rid of someone sat there doing a really rubbish journalistic job of sat there, copying and pasting and putting images in, probably not very rewarding for the individual and probably not the best use of their time when a computer can do it better.
The Future of AI in Sales · [26:20]
Will Barron:
What areas within the kind of next five, 10 years of sales. And let’s talk about more the account management role perhaps than the SDR role. Because we’ve covered that somewhat. What areas of sales will be taken away, replaced, changed or helped for an account manager doing larger deal sizes coming from 10,000 to $100,000, which is my world in medical devices, what’s going to be taken off the table, which is boring, crappy jobs that we hate doing and what’s going to be perhaps added that is exciting and intriguing for us?
Victor Antonio:
And I think that you pretty much covered it already is that, all the mundane task. And I talk about this in the book and that when you look at the job of an account manager, and you were to actually lay out everything that person does on a daily basis. You’ll see that there’s tasks that can be replaced through automation, which means that will give you the salesperson more time to go out there and actually sell something. I think that’s part of the automation piece of AI. I love that. The augmentation piece of AI is where…
Victor Antonio:
Let’s get back to our example. You just did a presentation, you sent a questionnaire, they send one back, you send a proposal, they send something back. It goes back and forth. The ability of the machine to be able to predict whether that deal will close or not, is what’s going to help you decide what to focus on. Because again, when we think of a large sale, a B2B sale. If we look at the amount of time, it takes to close a deal, let’s say two months, three months, whatever it may be. What’s your average sales cycle Will on curiosity?
Will Barron:
Medical device sales be probably three month for a deal. For a camera system, it’d be about 50,000 pound. It’d be about three months to close that.
Victor Antonio:
Three months to close that. Have you ever calculated how much money it costs to close the deal, to see an RFP all the way through? Has your company ever calculated that out of curiosity?
Will Barron:
I have no idea.
Victor Antonio:
Okay. Fair statement. If you were to calculate that out and let’s say for every deal that went from beginning to end, the acquisition of that new client, let’s say that the investment time, time, money and effort was let’s say $50,000, which is not unreasonable by the way, when you think about travel time, the amount of time to put proposals, different marketing involved, the proposal department involved, all these people involved. 50,000 is not out of the ballpark, if not more. And so now multiply that by the number of sales people you have, and you start getting the idea this can cost you a lot of money, but most importantly is the opportunity cost.
Victor Antonio:
The amount of time, the three months you spend on that deal, you could have spent on the other deal. And that’s where AI kicks in, in terms of augmenting your ability to say, “You know what? I need to focus on this deal, not that deal.” How do I know? Well, now you can actually talk to your boss and say, “Look, it’s not a gut feel. That’s what the machine’s telling me. Based on past interactions with typical clients, this is what the machine is telling me.” And now you got some sort of justification for why you’re going after certain deals. It can help you in that way.
Will Barron:
Amazing. And I imagine within Salesforce as a platform, there’s going to be a low hanging fruit just to kind of dump it down and take away some of the nerdiness of this. And then there’ll be perhaps a big flashing thing saying, “Take this person out for a round of golf. Now is the moment.” And so interesting. I’ve just thought about…
Victor Antonio:
Just show a picture.
How Soon Before AI Starts Telling Salespeople What to Do · [29:49]
Will Barron:
Golf, meal, get in front of them, do whatever, slash the ties on the car and then be there with a pump to help them, whatever it is. And so this is interesting. I’ve never really thought about this before. We are talking about now our bubble within our CRM system. Is there any way to do this right now, or is this again in the future to pull in outside external trigger events into this of you have your low hanging fruit and midway through the day, some news comes out. The stock market changes, that company has a shift. Someone gets sacked so we’re pulling in LinkedIn data into some of this and of people’s positions, or there’s a rumour going around or whatever. All these external trigger events and this complicates things massively, but then that could literally mean that at 12 o’clock you get an email from your CRM system saying, “Call this person. You can help them right this second. This is when they’re going to be emotionally charged. This is when they are looking for a solution that you can give them.” How far away are we from that?
Victor Antonio:
I think, by the way, that is such a great question because that’s, when it really gets to that point, you know what I mean? I think the larger company’s going to get there first. Salesforce, when you look at Salesforce or Salesforce has Einstein, IBM has Watson. In terms of their platforms. And one of the things they’re doing is now bringing in real time data for these trigger events that you’re talking about. And so to put it in context, if a company hires a new… I don’t know, VP of finance, or a CTO or CIO, we know that’s a trigger event, where we’re able to get into a company that maybe we couldn’t get in before, because there was an incumbent that was well seated. That could be like, “Hey, you need to hit this customer right away.”
Victor Antonio:
The trick is where do you get these sources from? And a lot of this data is unstructured. And so the trick now becomes is, how do you structure the data? In simple terms is when you have something structured data comes in a table, much like an Excel spreadsheet. Well, when somebody does a small post, “Hey, this guy got fired or this guy’s been hired.” Or it’s an article link. How do you bring that data into the system, structure it, and then actually put it in the mix so to speak, put the blend and put it as part of the waiting system within the algorithm. I think that’s where the magic is. And so when you’re looking at all these different applications that are being created. Again, think iPhone, apps, things CRM, apps, when you see these apps, there’s going to be one that says, trigger events, you know what I mean? Monitor trigger events and you’re going to be able to programme that machine learning app to look for just those things.
Victor Antonio:
And everything behind that is looking under the engine. That’s all programming underneath, but we’re not that far off Will, not at a high level, not big companies that have big data and have all this power and keep in mind and not to be long winded about this, but Amazon set the standard when it created AWS. The Amazon Web Service, because what it did, it democratised, literally democratised machine learning, AI, because it put all the apps in the cloud says, “Hey, if you’re a small company, you don’t have to hire data scientists or mathematicians. You can just use our apps.” Salesforce and Microsoft are doing the same thing. Now small companies in the future will be able to plug into a lot of these apps. And again, I think it’s going to get really interesting when it starts happening when we’re all using machine learning to market our products and services.
Ways AI Can Help Improve the Relationship Between Sales and Marketing · [33:07]
Will Barron:
That’s going to be amazing. And one final thing I want to wrap up here on Victor is, how does this change the dynamic or potentially change dynamic between marketing and sales? And what I mean by that is marketing for the longest time has been spread out, just spam a newspaper, spam an email list. Nothing’s personal. Now that we’ve got this machine learning, they can use people’s names, the context, the fact that there’s a trigger event, an email can automatically be sent out from perhaps an organisation rather than a company. And perhaps in the future Einstein is a avatar that a company has that sends out the email to make it a little less corporate and a little bit more personal. And you can reply to that as opposed to kind of Will the salesperson on the end of it.
Will Barron:
How does all this pull together between marketing and sales and marketing are going more and more one on one and salespeople, this has been what they’ve been good at. This has been the competitive advantage versus marketing if we see them as competing units, which clearly shouldn’t be, but they have been perhaps in the past, how does marketing then intertwine with sales? And does this now just get rid of the division between the two?
Victor Antonio:
I think it’s like a Venn diagram. There’s an overlap, and I think the book that challenger sailed to me was the first like warning shot in the air that you need marketing involved in the sales process to provide insight, within your messaging and your presentation, voice of the customer, blah, blah, blah. Well, AI is pushing that a little deeper now. There’s one company I mentioned in there is called Phraise in the book, where you can actually bolt in this little plugin into your Google mail or Outlook, your Gmail or Outlook. And one study it did was it showed that just by changing the headline, the subject line recommending a subject line, it was able to increase deal closes by or open rates by 30 to 35%, some crazy number like that. That’s where marketing’s going to play. They’re going to use the machine learning stuff to help sales people say, “Hey, this is where you want to go.”
Victor Antonio:
But I think marketing is now going to be able to use these products. And now come back to sales and said, “Look, I got some real stuff for you.” I think both people are going to share the platform. Now, I think that’s exciting, but things we didn’t touch on. I mean, you look at machine learning today, talk about price optimization. What’s the right price to price a product? Let the machine do it. It can actually tell you when somebody may or may not quit. That’s was another mindblower. They have these programmes now where they can actually anticipate whether somebody’s going to leave the company or not based on email interactions. Now some would call that spying, but man, it is what it is. You know what I mean? But there’s all these things that are happening right now that I think we’re going through an exciting time right now with this Will.
How to Improve Your Understanding of AI in Relation to Your Sales Role · [35:48]
Will Barron:
I agree and for anyone who’s listened to this now, Victor who has there’s been caught up in it, they’ve really enjoyed it. They are perhaps doing well in the sales role, but they’re looking for a shift somewhere else. Perhaps they’re looking at sales ops, perhaps they’re looking at marketing. Is there anything that they should be doing, reading, watching, learning? Even if it’s building chat bots, whatever it is so that they can be in a position perhaps to be the sales AI guy within their company in kind of that five to 10 year period, when probably will be somewhat of a role there, is there anything they should be doing now, if they’re excited about this, to learn more about it and to become equipped, to perhaps add value to their organisation in the future?
Victor Antonio:
Well, I don’t mean to be a Salesforce cheerleader, but one of the things I like about Salesforce is first of you have the CRM, you have the platform. You also have all these things in the background you can play with that are AI. You know what I mean? And if you want to stay, I guess, on the leading edge of what’s happening in selling, just go beyond the basic CRM, start understanding how some of these algorithms actually work, how many… These predictive analytics, what you can do with these things. And on top of that, the reason I like Salesforce is if you go to their website, they have so much free content, even on their, on YouTube, their Dreamforce channel. There’s so much great free content on how a lot of this stuff works that I think people would benefit from them. If you’re in selling today and again, I think if you’re at a manager level or hope to be at a manager, a VP level, you need to study this. This is where the future’s going.
Victor’s Advice to His Younger Self on How to Become Better at Selling · [37:22]
Will Barron:
Amazing stuff. Well, I’ll link to some salesforce.com and quotable.com articles as well. And yeah, with that Victor, I’ve got one fan question, mate. Something that I’ve asked everyone who comes on the show. It’s something I’ve asked you in the past as well. I’ll pitch it to you again. Be interesting to see if your answer is similar to last time. If you could go back in time and speak to your younger self or be one piece of advice, you’d give him to help him become better at selling.
Victor Antonio:
You did say that to me, didn’t you? You did say that to me. And I believe my answer was, don’t be so hard on yourself. I think that was one of the answers I had. Don’t be so hard on yourself. Sometimes we’re so hard on ourselves that we kind of defeat ourselves in the process. Look, we’re just learning as we’re going. You know what I mean? We all get better. Don’t be so hard on yourself.
Parting Thoughts · [38:08]
Will Barron:
Amazing stuff, great memory as well. You the only guests who I’ve asked that question to perhaps like a year later that’s remembered it. I appreciate that mate. And with that, tell us a little bit about the book where we can find it and then where we can find out more about you as well, and specifically your YouTube channel. Because there’s a tonne of awesome content on there.
Victor Antonio:
Well, I think if you go to YouTube channel, just type in Victor Antonio, you’ll find me, but the channel’s Sales Influence. The book, you can find the actual electronic copy on Kindle version on Amazon Sales Ex Machina, which is how AI’s changing the world. By the way, Sales Ex Machina, means sales in the machine as to the Dale’s Ex Machina incase people have missed that. If you want a paperback copy right now, it’s only available on my website, salesexmachina.com within the next month or so we’ll have the paperback available on Amazon as well.
Will Barron:
Amazing stuff. Now how about this? I’ll buy a bunch of copies and we’ll do a giveaway in the daily salesman email newsletter that comes out daily as well. We’ll do that. You can go on Victor’s website or get on the email news list and we’ll give away some copies of the book there. And with that, Victor, I want to thank you for your time, mate. I want to thank you. And I’ll shut you off for this of, I guess, projecting things forward making… I’ve read the book. I’ve got it here in front of me. If everyone’s listening on the audio, clearly you can’t see this. Really interesting, really well written and digestible. I’m a huge nerd, but anyone who isn’t a nerd will get through this and enjoy it as well. I appreciate that. And you having a kind of an angle and put this perspective and moving us forward with all this Victor and without want to thank you for joining us on The Salesman podcast.
Victor Antonio:
Thank you for having me Will. Always a pleasure.
