Webinar Recording: How Artificial Intelligence is Impacting LeadGen

Mike Gibb 0:05
Hey everybody. Welcome. Thanks for joining us today. My name is Mike Gibb. I read lead gen bulletin.com. And I want to welcome you to today’s webinar on how artificial intelligence is impacting lead gen. Artificial Intelligence is changing the world around us right before our very eyes. This includes how we create and manage our lead generation operations. There are tremendous opportunities to improve efficiency and output, but also a lot of risks and potential pitfalls. So today, in this webinar, we got a panel of experts here to break down how AI is impacting lead generation to make sure you have the knowledge and the tools and to make sure you aren’t falling behind. Let me take a second and introduce you to our panelists. Brad Biggs is an avid startup executive. With more than 18 years of hands on product development and digital marketing experience. He understands the challenges facing modern startups, small businesses, SMEs and enterprises firsthand. Christopher Williams is a Stanford University graduate and seasoned professional and marketing software development with more than a decade of experience he currently serves as a technical product manager for trusted forum at active prospect wears expertise fuels, innovation and digital marketing solutions. His passion lies in crafting cutting edge technologies that empower businesses to thrive in an ever evolving digital landscape. Everyone who’s watching us live today is absolutely invited to ask questions and submit comments during the course of our webinar. You can use the chat or the q&a Actually, then we show the chats on the chat. You can use a chat or the q&a to submit your questions or comments. If you wish to participate anonymously, then go there, click on the q&a button there is an anonymous button that you can push in anything you send to us is said anonymously. If you’re watching a recording of this and there’s a question or anything you see or hear shoot me an email, my email address is Mike at lead gen bulletin.com. And I’m happy to forward it to the panel to work on getting you an answer. All right, with that, let’s jump into the conversation. So the first thing I want to throw out there for the two of you is let’s just talk sort of high level, you know, the low hanging fruit, what are the most significant ways in which artificial intelligence is transforming lead generation today, Chris, you can blast alphabetically, so I’m gonna make you go first, Oh, I

Christopher Williams 2:21
love that I never get that honor. So appreciate it. You know, there are just so many ways that it definitely can get us and I think today, it’s still growing a lot. But we definitely see some really interesting ones. Some of the things that I’ve talked to a lot of customers about are just number one is using machine learning to kind of find the high intent leads the ones that are actually like a really good fit. If you have a lot of data, there’s a lot of patterns and trends that you can kind of find from that, you know, and actually find out which leads are going to be good. And I know that we’re probably gonna talk a lot about this, because he’s specializes there for sure. So I’ll leave some of the meat there for him. But that’s definitely that’s one. Also just using it to kind of hold conversations and actually follow up with consumers. That’s a big one, because no one wants to kind of do all of the manual work that’s involved there. We’ve been seeing it being used for transcription a lot just in, you know, normal things, when you want to make sure that you have a transcript of the conversation that actually took place. It’s also really useful for coaching purposes as well, if you’re you know, having a, a, a call center agent that’s talking to someone, you want to coach them on how well they’re doing things, maybe some improvements they can make AI can give you real time improvements. So that’s like super, you know, useful there. And then for me, personally, being the product manager of trusted form compliance, compliance is the big thing. You can use AI to tell a lot of things that normal people just can’t tell from a lead event. The FCC and FTC are confusing everyone right now with the 20 million new rules and regulations that are coming out. So if you have AI, you can kind of go through this and see if all those requirements are met that people have interpreted and kind of do it not just on feelings, but moreso on data. That is a huge plus. And we’re actually using some of that now with one of our products called trust to form verify just to help identify what that consent language actually is. So yeah, it’s being used in a lot of super important ways.

Brent Biggs 4:13
Chris, I can’t agree with you more. Some of the topics that you cover that I jump in from a different perspective, which is on like, look like modeling and for us, what I’ve always focused on is how can you cut out people who will never be your customer without having to use really high expensive enrichment services. So I think that what we focused on and what I’ve seen, the best use of is minimizing who you’re calling, by not purchasing the lead or reaching out to them if they’re not going to have the potential to be your customer. The other area that I can’t agree more as on compliance. I think there’s a lot of fear around like the robo, calling Robo taxi and getting worse and worse and worse. And I think the benefit that people are missing is that when you programmed or what’s supposed to be, when you’re programming, a dialog flow, that using pre recorded message or building some LLM models, you’re really able to confine what that experience for the consumer is going to be, where you have a live agent that you go off. I mean, we hear calls in QA where they’re just going off the left field. But when you’re programming that, I want to call it a bot, some people get offended by it, but the bot, you know, you can’t control that conversation. And it’s just going to repeat itself and be consistent and delivering. And so you can have more compliance. And some of the areas that we’ve seen problems is in the deregulated energy market where the agents just go off script. And they do certain things and end up getting company sued by the regulatory agencies. But with I believe, with the bot or with the conversational AI, you can’t control that conversation every single time. So I can’t agree with you more that there’s a lot of positivity coming out of it with compliance, and also eliminating who you’re going to call.

Christopher Williams 6:03
Yeah, that is just so true. And when you’re talking about going off script, of course, you can go off script with calls, the websites that are used to generate leads can also in a wait go off script, because we’ve had a lot of websites that you know, may present an ad that isn’t 100% truthful, that you know, may do some things that many, of course, wouldn’t consider compliant, it’s really hard to have a human go and look through each one. And you know, kind of figure that out. But using AI, you have a resource to actually get that done and make sure you can check it for every lead that you’re going to actually call. Are you

Brent Biggs 6:34
able to tell them that change the page? Like, like, I know, one of the problems we’ve heard is on some of the offers with in Medicare, whatever is that there was those incent people thought they were getting some free card. And then as the they get to the call center floor, it’s like what free Carter you’re talking about? Are you guys able to to use AI to detect when the pages are changing? If they’re changing the ad after? In the middle of the flow? Yeah, so

Christopher Williams 7:01
the way that we work, we just capture the entire lead event to happen. So any changes that happen on the website? So all of that is just data? Right? So um, anything that we can feed to AI, of course, it would be able to detect the change, and he’s kind of anomaly. So we don’t have that yet. Let me be clear there. But definitely, it’s a possibility that we’re looking into.

Mike Gibb 7:24
Are you seeing the current tools that companies might be using starting to incorporate integrate AI into their solution, as sort of how most companies are using it versus say, you know, a standalone tool like, say, chat, GPT, or the other large language models, where someone’s kind of have to sort of proactively go? And actually, you know, sort of access and take advantage of the tool? Or are you finding sort of a mix of both?

Brent Biggs 7:57
I mean, from our, from my side, we bundle all the individual tools together into one, one platform, right? I think it’s, I think he can get confusing because there’s so many people saying AI today, everything’s AI, you know, versus, you know, what type of machine learning? Is it really what is it really doing? And I think there’s just a lot of confusion. So for what we’ve seen is we just try to simplify and talk about the outcomes. And as you know, from our side use the AI to achieve those outcomes. I think there’s a lot of confusion about what some things are machine learning and some things aren’t and and where AI is. But that’s at least from our side, most of the people that we’ve talked to most of the potential customers we talked to is they care more about the outcomes and how they’re going to get there, then really like what is the, you know, basically, how’s the sausage made? Through the process?

Christopher Williams 8:55
Yeah, and I can’t agree with you more that everyone is just calling everything AI right now, I’ve seen several tic TOCs that are like just throw AI on it, like toppings on top of burger and they just keep throwing more and more on there. But yes, you’re absolutely right. Not everything is exactly clear to people what type of AI that they’re using. And I think what we’ve seen is that most people actually are using machine learning to kind of like classify, once again, which leads are good or bad. They’re using that they’re using a lot of data sources, just to kind of take in a lot of data correlate that to their conversion rates, you know, whether or not leads actually converted over to customers or not using that they can kind of you know, come up with a model that somewhat identifies which leads actually are likely to convert. That’s been the most popular use case, especially for those who are using ping post bidding. Granted, that may change very soon how that all works. So you know, use it while you can. But yeah, that’s definitely a huge use case there.

Brent Biggs 9:55
One thing that I’ve noticed which has been great is more the democratization of AI, right how simple it’s becoming for everyone to use it in some form. One of the challenges that we had building, you know, on our challenge we face on really scaling lead scoring was the ability to have the ML ops and have the back end. And it was going to be very resource intensive. In an expensive to have a data scientists run that operation full time. So we just weren’t able to deliver it at a cost effective price until we found an application that has a complete ml ops does makes the time of developing lead scoring models cut down from three weeks to a day. And that’s where I think the power is going to be as more and more accessibility and the easy it is to access these tools, and build them into products. I don’t know if you guys saw I mean, I think everyone’s seen the new track GPT and K mile I just started playing with it. Today, it’s pretty interesting. But the, the thing that I noticed about that is like how fast it’s changing. That, you know, there’s like, you know, firstly, a dialogue flow, we have pre recorded messages, and you have a little intense pass, and it does, and then people were building LLM models combined with that, and now this comes out. But I think that’s a great thing, because I think that’s going to continue to be more accessible for smaller businesses, where you’re not going to have to have a middle person, or some very expensive staff in order to operate, or they’re gonna have to know vertex AI or ml ops or whatnot. So we were very thankful that we found this company to work with and partner with that can cut that big chunk on the back end down for us when we’re building the lead scoring models. So I think democratization, I think it just the way it’s so advancing, it’s going to be a lot easier for companies to just access it and use it. Yeah,

Christopher Williams 11:55
definitely, that’s gonna be a big, like part of it growing and being adopted more widely. I think right now we’re in the state where a lot of you know, of the less technical people know how to go and play with it and kind of see what it can do, but they don’t know how to incorporate it into the actual systems yet. So that’s the next stage that I’m really looking forward to is when incorporating it becomes a little, you know, more simple, you don’t necessarily have to hire somebody to do that, or at least hire as many specialists to, you know, handle that. And then we’ll we’ll actually see it being used a lot more than we do today.

Brent Biggs 12:31
We were looking at this, I was digging around. I don’t know if you’ve seen it or not. But talking about like just coding and production and coding. I saw I don’t know what’s going on. But our production encoding has gone up. And I think it’s related to co-pilot, but just talking about like efficiencies and stuff like that Not needing, you know, the more specialized people. I was actually seeing this thing where where you wrote out your complete project plan, basically was for building a video game, but you wrote out your complete video game, and what you wanted. And I mean, like, it’s like, Pong, right? It’s not like some advanced thing. And you you’ve you put it into this chat, LLM model. And it actually went and had a project manager, a CTO, a CEO that had developers, and it actually coded your complete project, just from the natural language that you provided to in the instructions that you had, and then even had a Yeah, the project manager would come back and ask questions about particular things that it got hung up on. So anyways, I went on a tangent there about about that. But I think with digital labor, I think what my tie in with that is, is that it is going to contribute a lot to making things more accessible, and creating more efficiencies to digital labor. So that’s exciting.

Christopher Williams 13:45
Yeah, it really is. And the copilot thing just struck me because I used to have my hand in code a lot. And I didn’t have co pilot kneeled down at the time. And I can only imagine if I did, maybe I would have been a better engineer. So yeah, hopefully we’ll see a lot of engineering teams kind of advance with things like that, that can actually help them out too. So I love that.

Mike Gibb 14:08
Great stuff. I’ve got a few follow up questions, and we’ve got some stuff coming in from the audience. So I’ll take a break from my questions and ask those. Let’s see here. The first one is will AI voice on an automated outbound call after a consumer completed a FORM FILL be considered a robo call, since it is not a manual that like human, that’s more of a compliance question, but I’ll throw it out there. If anybody wants to share an opinion that is not going to be considered legal advice. Yes,

Christopher Williams 14:36
I love that. I started a lot of my calls off with I am not a lawyer or expert. Anything I say it’s not

Mike Gibb 14:41
I didn’t read my attorney disclosure. I can read that. Well,

Christopher Williams 14:45
I think we got it now. But yeah, so of course, not providing any legal advice there. But I would be very wary of using any kind of regulated technology without express written consent in order to do so. So whether or not you know, is considered a robo call, which is still a term that hasn’t been clearly, you know, elaborated on yet, I would still make sure you take all the proper precautions and check this all with your compliance and or legal team for sure.

Brent Biggs 15:16
Whether it is or isn’t I think that’s the great thing about single consent coming into effect is, you know, I don’t I nerded out and watch the hearing. And one of the things that was what that I walked away with, I was really happy to hear is one of the commissioners, you know, said like, what was one of the one consent, we’re okay with Robo calling or Robo talks thing like that, we understand that, of course of business. So again, I’m not an attorney, but you could refer to the transcript and say that, you know, if you have seen that’s, that’s a great thing, but saying single said, if you have it, even if it is it’s still permissible.

Mike Gibb 15:53
Okay, let’s see here. How are you hearing folks are using AI to address the new FCC CMS rules in the Medicare Advantage base requiring one to one match between the agent and beneficiary?

Christopher Williams 16:10
Yeah, that is a big one. Honestly, these rules and all the changes are still so new that I don’t know a lot of people who are using AI yet, there’s a lot of plans to use it for sure. Kind of like what I was talking about before, in terms of using AI, so use all of the data that it has available to determine whether or not a lead event was collected in a compliant way. So when you’re using a service, like trusted form, AI has access to all of the assets from the webpages and all the things that they clicked and all the options that were there, you know, AI can tell whether or not this looks like what it is, you know, things one to one consent is, I would still personally be a little weary of it. And I am so weary of it, which is why we haven’t released anything yet. But AI can make mistakes, of course, right? What AI may think is a one to one consent isn’t necessarily what your lawyer will think one on one consent or what the FCC thinks one on one because it is. So you still have to be very careful there with anything that you are using.

Brent Biggs 17:11
Way out of my subject matter knowledge, though. I’m glad you’re here, Chris.

Christopher Williams 17:15
Happy to be here for this when I talked about all day.

Mike Gibb 17:21
Someone says so we were just weren’t able to deliver in a cost effective price until we found an application that as a complete machine learning operation does make the time of developing lead scoring models cut down from three weeks to a day. So it’s asking for the name of the company that offers that.

Brent Biggs 17:36
It’s a key to a K. And let me a KK i o.com

Mike Gibb 17:44
There you go. A que kio.com I’ll put it. Okay. Um, so you were you were talking before about the confusion. And maybe we should take a second just to kind of clear that up in terms of you know, what we’re talking about? Sort of like, I’m wondering if I’m understanding correctly, what you were talking about. It’s sort of the confusion between AI, artificial intelligence and machine learning? Or is it confusion, sort of just among the different types of artificial intelligence that are out there?

Brent Biggs 18:14
And my personal opinion, I just think, kind of work because we’re saying everyone just slaps AI, you know, slap AI on something, and it’s not, you know, really, I, I tend to, you know, when talking to people, I’m like, well, let’s examine, like, what exactly is the machine learning that they’re using? You know, is it a regression model, a classification model, and try to really understand the science behind it, versus just, you know, slapping on and it’s, you know, some notes AI that’s rotating something or doing something that’s not really AI. And I do think a bot is AI, it’s, it’s basically doing classification. So, if you’re using dialogue, flow, or intent models, I just, my personal opinion is Dig, dig a little bit deeper into what type of machine learning modeling they’re doing.

Christopher Williams 19:04
Yeah, there’s, um, there’s definitely lots of types of AI, like when you dig into it, you know, as I did before, this, of course, you see all the different ways that it can be classified and the various, you know, terms that are kind of thrown out there. And I think that for the average person, a lot of that gets super confusing. So they just once again, think everything is AI. And we’ve used that term for a long time. Like I even think back to, you know, when I was playing Sega Genesis on Mortal Kombat, if you’re fighting the computer, you call that AI. Was that really AI? Probably Are you not, but some people may argue that it is so yeah, there’s always going to be a little confusion when things are a little bit subjective.

Mike Gibb 19:44
It was always AI because I could never beat it. There we go.

Brent Biggs 19:51
stay up all night. Yeah, well, my daughter does.

Mike Gibb 19:56
Christopher you were talking about sort of kind of seeing it. sort of beginning to kind of use it and then sort of needing to incorporate it. And I’m just curious how you sort of see that process occurring? Is it just something as we get more experienced as we get more comfortable with using it and the technology being out there that it sort of happens sort of just kind of naturally? Or are there steps that people really sort of need to take to get to that point where they kind of understand what they what they can, what they can and shouldn’t be doing with the technology?

Christopher Williams 20:27
Yeah, I think a lot of those steps are aren’t really going to come from the technology providers themselves. So, you know, we have chat GPT now, like, the reason everybody can use is because someone decided to go and make Chet GPT and make it available to everyone, right, so that’s step one. Step two, let me making it easy for companies to make their own version of GBC, which that’s already out there too, honestly. But it still may not be well known right now, sometimes you need those larger companies who are well known throwing something out there, that just makes it really easy for people to use it before you actually get to the next step. And I think that we’re going soon, there’s going to be more companies, first of all, who are you know, technology providers, like active prospect, and data score, just offering you know, solutions that use it, then people become more familiar with it, they want to use it, they inquire more than companies start to realize, hey, I need to sell something that allows other people to use this tech. So they create something that makes it really easy to code, you know, AI programs, or to build l M’s AMS and machine learning models and all of that. So that’s, that’s when you’ll kind of get to the next step where you have things like co pilot, which just makes it easier to actually use AI in whatever it is that you’re working with.

Brent Biggs 21:48
So let’s say this roadmap for Databricks. And, you know, their objective is to eliminate the need for data scientists altogether. And I don’t know if you’ve anyone’s ever hired a data scientists or paid their salary. But these people are, I mean, I picked the wrong career, but it’s I think out of college, they’re getting 350 out the gate, you know, some are 500 plus. So, um, you know, I thought it was fascinating, the Databricks, which is used by data scientists, as a mission to eliminate the need for companies to have them. And actually, I think it’s good, because it just lowers the costs, or operating costs of delivering, you know, the services or delivering the products or packaging the product where it’s affordable. Um, I might jump ahead. But actually, this I think this is kind of a decent gateway into into something of talking about, like the cost of AI, and how do you deliver it cost effectively in Legion, right. And so, when we were starting with lead scoring, you can build the greatest lead scoring model out there, it would just cost you so much money to enrich the data, that it just became not cost effective. Right? So the challenge was as well, how do we get lead scoring just good enough, that it makes a big impact, but the cost is very low. And so I think that’s the same thing, when you’re looking at a lot of the conversational AI that are out there that are using MLMs. You know, they’re not cheap to use 20, I’ve seen 25 cents a minute. I’ve seen $1 $2, or record that’s loaded. I’ve seen it across the gamut of what’s out there. And what’s I think is that if you think about the pyramid of a phone call, you know, the first five seconds of that call, you’re not really sure what that phone call is going to be worth. But it’s not worth really worth that much. There’s opportunity there. So what technology do you deploy, if you’re using if you don’t want to use a human? Do you deploy to handle the five seconds, and as you move, kind of, I guess, down the pyramid, and you get to like 10 seconds, 20 seconds, that call becomes more valuable? They’ve gone through one qualifier, well, is that when you introduce LLM models, and you spend a little bit more into it, but that the biggest challenge right now is how do you build a product that can be cheap on the front end, and get more experienced and better on the back end to actually handle the whole call? And you know, talking about it, because I mean, obviously high intent leads, you got a Facebook lead, someone’s going to expect to talk to a machine, I think they’re fine with that. But when we’re talking about like Legion at scale, it’s just not I don’t think it’s feasible to have a bunch of phone calls where you have 35% of your of your calls are five seconds or less. And you’re paying some increment of 25 cents or 35 cents or whatever it is. So I think the real challenge is is how do you build how do you deploy AI long term where you can take diploid at each value level of the of the call? Does that make sense? I don’t know if I’m just rambling there. No, yeah,

Christopher Williams 24:55
I think that makes perfect sense because that’s pretty much the same thing that happened with software right? costs of creating any kind of software application used to be amazingly high when you had to go buy a physical server or have some storage space for it, and yada, yada. Now you can just go sign up online for a Ruby on Rails course, spin up a Heroku app, and you have a program made in minutes for free. So you know, that’s kind of the same thing that I see that’s going to happen with AI, there’s going to be more resources do it’s super cheap, super, you know, it’s simply and then we’ll get more adoption then.

Mike Gibb 25:31
One of my favorite shows is for all time is the West Wing. And I remember they were talking about drug research and an episode and someone’s like the pill costs, you know, 25 cents to make is like the second pill cost 25 cents, the first pill cost them like $100 billion. So I think that’s kind of maybe sort of the dynamic that we’re talking about, too. And is it bright? Do you think it’s just a matter of time? You know, as is, you know, like, I guess I can, I’m older through everyone cell phones first came out, we were paying by the minute for that, too. And that sort of eventually went away? Do you think that’s the sort of those sort of the same kind of model that we’re going to see, you know, when it comes to sort of AI tools and AI software?

Brent Biggs 26:09
I do, I think it is a matter of time that the call that cost continues to fall down lower and lower. I just think the challenge and lead gen is there’s not you know, you’re dealing with defined marketing budgets that you have. And so I think when you’re targeting, particularly in the chat, you know, the chat bots and everything that’s out there, how to best use them to maximize it, where you offset your human capital, but you’re using digital labor? Because now I just don’t I don’t think it’s my opinion, that it’s possible to deliver on a large Legion. If you’re paying 25 cents a minute. For the majority of phone calls that are I mean, everyone’s looked at $1 disposition hangups, that or whatever it is. So but I do think I do think it’s a time is a prime example is, you know, when we started doing the lead scoring, when we started this five years ago, I had two data scientists, and three data engineers, and that’s been minimized dramatically, just by a keel that we were just talking about. And that’s, that’s actually made it really affordable, affordable for us to deploy lead scoring at scale, you know, and so we can take in hundreds of 1000s of leads, score them, reject them and make a decision on them and not break the bank. Right? Because you’re paying for 35 cents for a lead or 50 cents a lead. You can’t have a lead score. That’s 50 cents. Right? So, yeah, so I think it’s gonna come down the more than democratize it. I mean, yep, that’s my two cents. Yeah.

Christopher Williams 27:44
And maybe that’ll make us see adoption in certain verticals sooner than others. Because, you know, there are some verticals where each lead is a lot more than 25 cents, so maybe they’ll be able to afford it. We’ll see.

Mike Gibb 27:59
I’m curious as, as we talked about price. And as you were talking before, Brian, about data scientists, you know, one of the sort of the debates that’s going on with respect to AI is, is it going to replace workers? Or is it going to sort of help in augment workers? Like, you know, the, what I think the conventional wisdom is, is that people who know how to use AI, are the ones that are still going to are the ones that are going to be the most employable, or sort of the winners, I guess, in the AI race. And I’m curious, you know, with respect to this industry, do you see it more as a replacement for either agents or other types of employees? Or, you know, just kind of workers in general? Or do you see it more as a tool to improve efficiency and productivity for the people who are already here? Or both? I guess, again, it can be both.

Brent Biggs 28:54
I mean, I, I see it as both. You know, one of the things that we want to play with is taking in calls and having, we’re actually going to do this with the client, have the client use low cost labor to initially front the call, and then prep them with something like this. I’m not a scripter. So forgive me, you know, great, is it okay, if I transfer you to our automated system that will set your appointment, it’s going to protect me from receiving personal identifiable information about you? And when the consumer says Yeah, I’m ready, you’ve sort of prep them and then you can send them into that more expensive 25 cents a minute because you’re you’re getting someone to schedule a Home Services appointment or solar installation apart apart, I’m sorry, appointment. So to me that’s complimentary, right? Because you’re able to increase your SPH is with your Fronter and let the let give the consumer an option to use it and then use the bot to set the appointment and in take care of that process. I don’t know I mean, I watched South Park and according In the South Park, the only people that are going to have jobs are going to be blue. You know, people that build stuff with their hands. Lawyers are in trouble seeing that South Park episode, have you seen it?

Christopher Williams 30:12
I haven’t seen that one.

Brent Biggs 30:14
Oh my gosh, it’s amazing. You gotta watch it. It’s funny. It’s hilarious dude pulls up in this like limo truck to like do plumbing at Stan’s dad’s house. It’s amazing. So

Christopher Williams 30:23
I’m gonna watch it after this now.

Mike Gibb 30:26
We’re gonna lose everybody, as I go watch it now. Another question, yeah, send us a link to the episode. See, there you go. Another question, can an AI lead gen product, have an agent monitor the conversation and interrupt at any time as they watch the progression of the conversation? I know, you know, there are AI tools that can listen to a conversation and prompt, you know, on a screen with suggestions based on what it’s hearing in the interaction it’s having? Are there actual ways where it will interrupt a call? Does anybody know?

Christopher Williams 31:01
Definitely possible. I can’t say I’ve seen that anywhere today. But you know, if enough people say they have the need for it, because let’s say you need to interrupted for a compliance reason, for example, or just something else, they definitely I can see a company, a company tackling that and creating a way to, you know, create that stop.

Brent Biggs 31:23
Yeah, I hadn’t, I have no clue outside of my scope of knowledge. I try to stick to my lane here. What I know.

Christopher Williams 31:31
Yeah, the normal rule is that anything is possible if you have enough money to build it.

Mike Gibb 31:38
fairpoint what new skills should marketers be developing to stay relevant, you know, as we enter the AI enhanced marketing landscape?

Brent Biggs 31:51
I mean, I think a great I think a good place to do is like Google YouTube. I mean, that’s actually what I do a lot of times, and just stay on top of what’s out there, I there’s something called I think it’s got sound storm. And I was kind of educated myself a little bit more about the way that dialogflow and vertex AI works together. And I found the south store thing, and it’s fascinating to see where the technology is going, just so that you’re prepared for what’s what’s next, and what’s out there, and what the future holds. But anyways, it sounds storm, I think it is what is called the base basically had a conversation with itself. But what was unique about it was the empathy and the inflection and the tone. So when when one version of the bot would say something that was sad, you know, in the storyline, the other the other version, the other bot would feel empathy, or express empathy, sorry, and their tone would change. And so I think it’s fascinating just to go to YouTube, and Google the topic, and, you know, see what’s, see what’s out there and stay on top of it. So you know, what’s coming? Yeah, maybe innovate. So,

Christopher Williams 33:01
I’m just saying that you actually took my answer, because YouTube is where I go to learn a lot of new things. But since you said it, I’ll also say that podcasts in general are great. So I use or, you know, listen to a lot of those on Spotify. But in general, I think for marketers, like you don’t have to go become an AI expert, you don’t need to learn how to build anything, don’t go start taking Python classes, you know, or anything like that. But the important thing to realize is that most of you know, AI, we’re throwing that term out a lot, but most of it is narrow, in the sense that it’s really made for a specific task. So yes, there’ll be a lot of tasks that like, you know, we can now use AI to automate, but there’s still gonna be a lot of things that marketers need to do to coordinate the overall process at least until we get super in, in intelligent, you know, AI that just puts us all out of our, you know, lives and jobs. But until then, this is gonna definitely still be a lot that you need to actually manage. So just learn how to use AI to make yourself more efficient. You know, rather than expecting you to actually do your entire job, I think, is the key there. The more ways that you can like become more strategic, rather than tactical, if you can make the plan for how your company is going to use the AI and how you know, it’s going to increase something like as long as you’re more focused on the overall strategy, then you’ll be very, very useful in order to make your marketing successful.

Brent Biggs 34:27
Well, so I think not falling for the trap of this is so special and unique. Give me this massive deposit. And we’re gonna go build you this amazing bots. That’s going to replace your call center. I mean, go back to YouTube and just take what they’re saying and type some keywords in there. lol Mala development, you know, lambda, and see, you know how to build a model or how to build an LM model on open AI or whatever it is right? And educate yourself and realize like, okay, it’s not it’s not that it’s not hard. but it’s not that hard. Right? It’s getting easier and easier and more democratized. And so as you’re talking about being ahead of the company, you may be saving your company money by saying, Well, I don’t need to pay the setup fee. Because you guys don’t really need to do all this. So if we’re going to test it, you got to not have a setup fee. I don’t like setup fees, if you can’t tell.

Mike Gibb 35:24
As part of that learning process, you know, one of the things we’ve talked about is sort of companies that are just kind of slapping AI on everything and not net and in some cases is not necessarily AI. Are there ways that an individual can sort of figure out if it’s really sort of AI that’s, you know, that that’s integrated or that you know, that whatever the company is touting actually uses AI, or, you know, any sort of rule of thumb or or questions that they can ask or a way to sort of how do you figure how do you learn enough to find that out?

Brent Biggs 35:57
And I think he just asked him to explain like, which what AI they’re using, what machine learning models they’re using, and then research it afterwards. You know, I think a lot of this stuff is out there a lot of informations already out there where you can cross reference with someone saying that’s my Yeah. Because yeah,

Christopher Williams 36:20
I mean, yeah, that’s definitely a really hard one, right? Because someone if they want to lie to you, I think that they’ll be able to, but some ways that you can catch them in maybe lives if they’re not as well versed in it. Ask them, how is there? Like, how are they actually training their model that they don’t have an answer for that they’re not using AI properly? Yeah,

Brent Biggs 36:44
that’s that’s really good asked about the features they’re putting in their their data sanitisation beforehand, or what’s their ETL process?

Mike Gibb 36:53
Because I just figured it’s easy to throw buzzwords out there and make it sound like you’re using AI. And it can be very impressive, but not necessarily explain anything?

Brent Biggs 37:03
Well, that’s actually the Christopher brought up a good point. I mean, there’s the thing that makes the model is the data that goes into it. Right? So have them explain what they’re putting into the model to train the model, what’s their training data? What is it like to collect? What did they do in the process of collecting training? How did they do data exploration, on the data to decide which attributes go in and are relevant to the model? There’s a lot more than just having a model, it actually starts on the data side.

Christopher Williams 37:33
Absolutely, if that’s the key, and you know, a lot of it too, you can probably do some testing on your own too, because, you know, if it really is artificial, intelligent and should behave intelligently. If you notice that something is wrong at first and your first test, hopefully it can learn from that. And you know, change it, of course, depends on what it is that you’re testing. But there are probably in general, just ways that you can look at test to see if it actually is learning from a model. If there is something more advanced going on, then it just having a couple of, you know, conditions that say If This Then That, and that’ll help you tell whether or not it’s actually using AI or just some simple logic that they’re slapping the AI sticker on.

Mike Gibb 38:18
Okay, all right. So let’s see here. I skipped over this question. Let me come back to it. What are their ethical considerations and potential risks associated with using AI and lead generation? And if so, how do you mitigate those or address those?

Christopher Williams 38:36
Yeah, this is a big one. So the first thing that stands out to me is that whenever you train, your models will have the potential to create a bias. And you really need to be wary of that. And I think that this is one of the problems that technology just overlooks a lot of the time, honestly, in general. So you have to be really observant, and like watchful to actually even catch it. And then to solve it, you probably have to change a lot more in your data models and things like that. To give an example of what I’m talking about, I remember there was what years or years ago when like facial recognition and type technologies kind of really took off. I was working at Google at the time. And I remember I was like testing one. And when I tried to use it, it didn’t work. And then my coworker, you know, use it and it worked perfectly difference was that my skin is darker. There, her skin was much lighter, and the technology relied on light being reflected off of your skin. So of course my skin absorbed it. So it just didn’t work. So it kind of created an inherent bias that no one really thought about at first. So you can have a lot of things like that pop up with AI too, if you feed it a lot of data for a specific type of person, that it only targets that person and those are the only leaves that you’re now getting. You might find yourself in some hot water after that when people realize you know that you’re excluding certain groups.

Brent Biggs 40:01
Yeah, it’s funny, Amazon early on one of the build, I think one of the early models that they’re a good example of pre selection bias. They built a human resource once to process all the applicants there, were only hiring males. Because at the time that they were using the resumes, as you know, classify and positive note and class, the only people applying were males. And so they weren’t really hiring into developers that were females, even though they were applying because they weren’t getting through the the model. But that’s Yeah. So you got to be careful of biases and inadequate for that. I guess, I guess that would be an argument is why you have a data scientist because they understand how to deal with that.

Christopher Williams 40:41
That is true. That’s why it’s hard to replace everyone. Complete job for sure. But that’s only one thing to answer, there are just so many things that you have to watch out for. I mean, we know that once again, AI works off of data, data collection is a hot topic when it comes to whether or not it’s morally collected and things like that. So if you don’t collect it the right way. And you once again, might find yourself in hot water and having to, you know, change some things. And of course, my recommendation to everyone is that when you’re collecting any kind of data that may be sensitive, make sure that you give people notice about that, once again, not legal advice. But that’s just the nice thing to do, and can definitely help you in some cases.

Mike Gibb 41:27
Right? Full disclosure, transparency, right? If a company is using AI, if they’re interacting, interacting with the consumer using AI, I think the standard is to sort of make that disclosure and let them know or in some way, shape or form that it’s not a human that they’re having a conversation or engaging with.

Christopher Williams 41:45
Oh, yeah, you’d want especially after we all heard what happened with Joe Biden in the deep fake, they’re people who would just want to use AI to misrepresent what’s actually happening. I mean, those are the people who are just the scumbags. They’re doing things and doing it, you know, wrong on purpose. So they’re probably not worried about the ethics of it. But uh, be good people don’t don’t do things like that. Yeah,

Brent Biggs 42:09
and I think also the the ability to store and analyze personal identifiable information. So if you’re doing, you know, if you think you have a bot, I’m gonna say it again, conversational AI out there, and you’re doing a solar setting appointment, or you’re doing an intake on something, it’s different than a bunch of agents doing it and typing in and I think that with AI, it’s, it can be stored easier, and then analyze easier. So the information that’s being asked in the bot, I think consideration needs to really be examined, what is the company that we’re working with? What are they doing with that personal identifiable information? Are they blanking it out in the transcript? Because it’s also going to be transcribed? So what are they doing with the transcript? How are they storing it? I think that could open up some ethical concerns.

Mike Gibb 43:02
Another question here from me, the audience, what roles in lead gen, do you think will be augmented by and thrive with AI? And what roles will slowly disappear?

Christopher Williams 43:17
That’s a good one. I mean, I, I honestly want to say that the simple answer is I think all of them will thrive, it’s gonna be really hard to get rid of any full roll, like I was saying before, because they’re multifaceted to every one job, and I don’t think you’re going to be able to just easily get AI to replace them all. But I’m certainly for performance, performance, marketers, just determining once again, which strategies actually work, which ones don’t, which one, create high intent leads, and that, that’s where AI is going to really, really shine, like, it’s definitely going to be a huge way to improve all of your conversion rates and KPIs that you’re seeing, just kind of using the determination that you can get from these large models, actually, instead of just, you know, relying on gut and instinct.

Brent Biggs 44:01
I think there’s gonna be less humans in the call center. months, you know, long term, and the ones that are there will have a more, I don’t want to say that. I don’t want to just say by saying this, that I don’t mean someone’s job is meaningful right now, but more meaningful and impactful, and more rewarding role, besides answering a call hang up, during the first five seconds of a pitch, hang up. You know, I think that I think there’s just gonna be less humans having to do that.

Christopher Williams 44:32
That’s one of the things that I’m really waiting just to see how society also reacts to that in general, because there’s a lot of people who don’t want to talk to a robot. But I do agree there will be you know, less than need or a human actually do that. But I do think there will still be some need, even if it’s just for the people who are resilient or resistant. I mean, to be new technology.

Brent Biggs 44:53
It’s a little bit about conditioning though. Like I mean, I think about like the what Mike the way my kids communicate versus The way I do or what they’re used to or not used to social media, I didn’t have social media growing up. Do you think that it’s just conditioning? Like, it’s just the younger generation will be used to it? And thus they will prefer it? That’s a good well, mind. Yeah.

Christopher Williams 45:19
Yeah, that definitely could happen if they grow up like that. I mean, at this point, the younger generation already doesn’t want to talk to them, period. So by the time they take over and we’re all gone, a lot of things are going to change drastically for sure there. But as long as you know, those of us who know the difference are still here, we’re going to have some some trouble fully converting over.

Mike Gibb 45:41
Another question, not legal advice. But is it So is it your opinion that on the form fill disclaimer? You need to mention the contact via voice or SMS can include AI technology? Probably? That’s my opinion. Yeah. Better safe than sorry.

Christopher Williams 45:57
Yeah, I think a lot of people have that opinion. I think it actually has been clearly stated in a couple of, you know, rules that have been put out. So I would definitely, once again, consult with your legal and compliance team. But they’ll point you to some great resources, I’m sure straight from the FCC, and FTC.

Mike Gibb 46:14
And let’s see here. So we spent all this time sort of kind of talking about sort of where we are. Now let’s talk a little bit sort of turn our gaze forward and talk about where you see things going. What future trends do you foresee within the, between the intersection of AI and lead gen? And how can marketers start preparing for them, especially if they’re already sort of a little behind or way behind, and not up to speed? This is changing so fast and evolving so fast. I mean, we are 50 months 16 months for when chat GBT first started, hit, hit the world. And now look at where we are and how fast things are going. You know, I know you mentioned sort of YouTube and podcasts before, but where do you sort of see things going?

Christopher Williams 46:58
So I see it just becoming or having the ability to personalize so much more of the entire, like lead generation process. And I mean, from when they first land on a landing page, and then showing them you know, the appropriate set of questions afterwards, maybe that completely becomes personalized per consumer. Or maybe we really buy or make improvements and how we actually match them with the actual sellers that they want to purchase a service from. But all of that AI is going to touch in some way. And it’s going to make everything way more efficient, way more accurate, way more simple to get people to where they actually want to go. And in terms of staying on top of it. I think outside of what we’ve already said, the big things are just going to be to keep your eye on the market. Like make sure you know what other people are doing. Make sure that you’re at least looking into it. You have someone at your company who can explore this and kind of give ideas and figure out what’s feasible, what’s not all of that. But if you’re not aware of like, you know, where the rest of the market is moving, then it becomes really easy to get left behind.

Brent Biggs 48:03
Yeah, I mean, I agree, I think it’s all positive. in the right direction. I think it a lot of the fraud, a lot of the BS leads that get pushed through. A lot of calling people that just are not your customer won’t ever be your customer will will make a better experience, at least on our side for outbound. And I think it’s going to continue to get cheaper and be more democratized. And but I, I believe it’s, it’s going to move faster, faster, faster and faster, and you’ve got to get caught up now. Again, I’m gonna go back to IKEA, and I’m just gonna sing their praises. I was extremely frustrated that their time to go live. And I went to trafficking conversion. The only thing I got good out of that. So thank goodness, my CFO made my CFO happy. But you know, I found that one application was like, Oh, this exists, this is here. And now I’m upset with myself that I wasn’t even like, you know, Googling it and staying on top of where things were going. But it’s going to change fast. And I think it’s exciting. I think it’s all positive.

Mike Gibb 49:13
Wonder if I could send them an invoice for I

Brent Biggs 49:16
did a lot of advertising, I want you to put my affiliate link, I need

Mike Gibb 49:19
to give you an affiliate link and I wouldn’t have been that would have been your best bet the affiliate leaves.

Brent Biggs 49:23
I’m gonna hit just like Yo, dude, I just promoted you. So I need some free credits.

Christopher Williams 49:31
Everyone will go to your LinkedIn page and find

Mike Gibb 49:35
it. So we’re almost out of time. We’ve got a few minutes left here. Before we wrap up. What I want to do is ask each of you what you know, for people who are watching this or they’re watching this live or they’re watching a recording. And as soon as this is over, as soon as we end this webinar as soon as they disconnect what it is over the first second third thing that they should do based on everything that we’ve talked about today, based on everything that they’ve heard and see, what are the first steps that they should take? Brian, I’ll ask you to

Brent Biggs 50:04
go first go watch that South Park episodes. My number one, it’s in the chat. So there you go. It’s classic. I’ll just put it in there.

Mike Gibb 50:12
I put I put a link to the Wikipedia. So everyone knows what Oh, nice.

Brent Biggs 50:15
Oh, good, good, good. Um, I think the second one is just maybe a lot, a little bit of time to do research on on YouTube to see what’s out there. And, you know, it’s, I think the real magic is happening with the, with Fang, you know, Facebook, Apple, Microsoft, Google, that’s where all the heavy lifting is being done. So I think, you know, don’t believe the snake oil, too much of people, you could do a lot of things yourself. You know, and especially if, if, if companies are asking for exorbitant amount of money to do something, see if you can learn it on your own and do it on your own because it’s getting there.

Mike Gibb 50:56
Excellent. Thanks, Christopher.

Christopher Williams 50:59
All right, I want to say one of the first things I would do if you’re interested in learning more and keeping up to date with things go to Google set a Google alert for artificial intelligence and performance marketing or something along those lines, you’ll get a lot of you no resources there. That’s how I find a lot of things just that then lead me down more rabbit holes. You’re already listening to this to this webinar. So you probably do things like that already. Keep it up. The more webinars and experts that you hear talking about it, the more you’ll learn, the more you’ll figure out that you know, you can go research and you know, find out cool things too. And last piece of information, talk to companies who are using AI to do various things that you need, you can learn a lot just from talking to them, maybe you’ll find products that you want to actually use, or you’ll just get inspiration and ideas. So definitely talk sorry, decide talk to that company right there, as well as data store two. Excellent.

Mike Gibb 51:52
Well, Brent, Christopher, I want to thank both of you so much. I really appreciate it on behalf of everyone who’s watching and listening. I think walking away with some not only some new tools and some new ideas, but also a deeper understanding of sort of where AI is, and where it’s going in the lead gen space. Thank you both for being here. Really, really appreciate it. To those of you watching live or watching a recording of this. Thank you. I appreciate you giving us an hour of your time. And I hope you found this to be worthwhile. If there are topics you’d like to see covered in a webinar, are there things you’d like to learn things that are keeping you awake at night, by all means shoot me an email, Mike at lead gen bulletin.com. And I’m happy to put a webinar together to talk about it. If you are watching this live, there’s a link in the chat. for next week’s webinar. It’s three o’clock on next Thursday, on what to do with best practices in dealing with consumer complaints, whether you’re getting them directly from a consumer or you know from an organization like the Better Business Bureau, how to respond to those and why, you know, it’s a good thing to actually have consumers complain to you. Because if they’re complaining to you, they’re likely not complaining to the Better Business Bureau or a state attorney general or somebody else that could lead to some bigger problems. So get a great lineup for that you can find the link right there in the chat or I’ll post it on Legion bulletin.com If you want a recording of this Mike at lead gen bullets a.com Shoot me an email, happy to share how to get access to that. So thanks everybody again for watching. Really appreciate your time, and we look forward to seeing everybody soon. Thank you, Mike. Thanks, guys.

Brent Biggs 53:25
Thanks, Chris. Nice to meet you.

How Artificial Intelligence is Impacting LeadGen

During the conversation, speakers discussed the potential risks and ethical considerations associated with using AI in lead generation, such as biases in data collection and models. Speaker 3 emphasized the importance of compliance and minimizing unnecessary calls, while Speaker 2 highlighted the potential of AI to detect changes in lead pages and ensure compliance. Speakers also discussed the challenges and opportunities of integrating AI into businesses, and the potential of AI in complying with new FCC CMS rules. Later, Mike Gibb and Speaker 3 discussed the potential of AI in lead generation roles, while Speaker 3 raised ethical concerns about data storage and analysis.

Action Items

• [ ] Research using machine learning to identify high intent leads based on patterns in data
• [ ] Consider using AI for transcription and coaching call center agents
• [ ] Explore using AI for compliance purposes like assessing consent language
• [ ] Follow up on using AI to address new FCC/CMS rules requiring one-to-one agent-beneficiary match

Outline

AI transforming lead generation, including using machine learning for high intent leads, automating conversations, transcription, and compliance.

• Panelists discuss AI’s impact on lead generation, sharing insights and expertise.
• Mike Gibb and Chris discuss AI transforming lead generation, including using machine learning for high intent leads and conversational follow-up.
• AI helps with compliance, such as identifying consent language and ensuring FCC/FTC regulation compliance, per Chris.

Using AI to improve lead generation and compliance in telemarketing.

• Speaker 3 emphasizes importance of compliance and reducing unnecessary calls.
• Speaker 2 mentions using AI to detect changes in website ads, but clarifies that it’s not yet implemented.
• Speaker 3 discusses incorporating AI into existing tools for lead generation, rather than using standalone AI models.

AI democratization, lead scoring, and digital labor efficiency.

• Speaker 3: Confusion around AI, machine learning, and lead scoring.
• Speaker 3: Democratization of AI simplifies lead scoring development.
• Speaker 3: Democratization of AI tools will make it easier for smaller businesses to access and use.
• Speaker 2: Incorporating AI into systems will become simpler, reducing the need for specialists.

AI use in compliance with FCC regulations for telemarketing.

• Speaker 2: AI voice on automated outbound call after form fill may be considered robo call.
• Speaker 2: One-to-one match between agent and beneficiary required for new FCC CMS rules in Medicare Advantage.
• Speaker 2 expresses caution about using AI for lead scoring, citing potential for mistakes (14 words)
• Speaker 3 advocates for deeper understanding of machine learning models used in AI applications (14 words)

AI technology and its cost-effectiveness.

• Speaker 2: AI technology is classified in various ways, causing confusion for the average person.
• Speaker 2: As technology providers make it easier to use AI, more companies will offer AI solutions, increasing familiarity and adoption.
• Speaker 2 discusses Databricks’ roadmap to eliminate the need for data scientists.

Reducing costs of AI technology for lead generation.

• Speaker 3 discusses challenges in deploying AI for phone calls, particularly in handling short calls (5 seconds or less) while maintaining cost-effectiveness.
• Speaker 2 compares the adoption of AI to software development, suggesting that as resources become more accessible, more people will adopt AI technology.
• Speaker 3 believes cost of AI tools will decrease over time, making them more affordable for businesses.
• Speaker 2 suggests adoption in certain verticals may happen sooner than others due to lead cost differences.

AI replacing workers or augmenting them, with a focus on marketing and customer service.

• Speaker 3 suggests using AI to augment human workers, not replace them.
• Speaker 3 and Mike Gibb discuss the potential for AI to improve efficiency and productivity in the industry.
• Speaker 2 mentions that interrupting a call with an AI agent is possible, but not sure if it’s been done before.
• Speaker 3 suggests marketers should stay informed about AI technology and its applications in marketing.

AI in marketing, importance of understanding AI basics.

• Speaker emphasizes marketers don’t need to become AI experts, but should focus on strategic planning.
• Speaker 3 advises asking companies about their AI and machine learning models to determine if they are genuine.

AI ethics, bias, and data collection.

• Speaker 2: “If AI model is not well-versed, can be caught in lies.”
• Speaker 3: “Data sanitization and ETL process are key to AI legitimacy.”
• Speaker 2 highlights potential biases in AI models, emphasizing the need for careful data collection and consideration of diverse perspectives.
• Speaker 3 shares an example of pre-selection bias in Amazon’s early hiring process, demonstrating the importance of addressing biases in AI decision-making.

AI’s impact on lead generation, including transparency, ethics, and future trends.

• AI transparency and disclosure are crucial for ethical lead generation practices.
• Speaker 2: AI will replace some call center jobs, but not all.
• Speaker 3: Younger generation may prefer interacting with AI.
• Speaker 2 predicts AI will personalize lead generation processes, making everything more efficient and accurate.
• Speaker 2 advises marketers to keep an eye on the market and explore new technologies to stay ahead.

AI in lead generation, including benefits and challenges.

• Speaker 3 recommends watching South Park episodes for first steps in understanding AI marketing.
• Speaker 3 advises researching on YouTube and relying on heavy lifting by Fang (Facebook, Apple, Microsoft, Google) for AI marketing success.
• Christopher suggests learning AI on your own by watching webinars and talking to companies using AI.
• Brent recommends setting Google alerts for artificial intelligence and performance marketing to stay up-to-date.