Loading summary
Transcript31 lines
- [00:00]
A
Welcome to the AI Chat podcast. Today on the podcast, we have the pleasure of being joined by Ashley Gross, who is an expert at AI and who helps companies implement AI and does a lot of really interesting things. If you're not following her already on LinkedIn, I don't know what you're doing. You got to do it. Certified LinkedIn celebrity over there. We're excited to have you on the show. Welcome, Ashley.
- [00:22]
B
Thank you for having me. I'm excited to be here.
- [00:24]
A
Okay. So super stoked to have you on. For the listener, can you tell everyone a little bit about your background and kind of what you do today with AI with companies?
- [00:33]
B
Sure. My background is I've been a marketer for 10 years. I started using AI back in 2020 because I wanted to consolidate my work week from 40 hours to 15 and maintain the exact same output and quality. It's also the year I found out I was going to be a mom. So that coincided really, really well with being intentional with my time. Thank you. And I never looked back. I started using AI and implementing AI pre chatg to enterprise marketing organizations, and the moment that I could actually tie experimentation and AI implementation to revenue was the moment that I never looked back. So within three months of rolling AI out to an enterprise marketing org, we actually overachieved our pipeline target by 25 million, which is not easy to do.
- [01:19]
A
Incredible.
- [01:20]
B
Thank you. And so I just thought this is way better for me because I love being a business owner. This is my third business I've owned. But I also thought that I could help more marketers and more businesses in general by sitting in this role. Because to answer your latter question of what I'm doing today, I work with mid market enterprise companies and me and my team of consultants will go to them and diagnose their business's main problem and then create use cases for AI that go to solve that problem. So we're aligning their work force at the top, but also the bottom. So it's not just top down, it's bottom up. And we're tying everybody's individual impact to the overall goal of generative AI for helping the business. And it's just fundamentally changing and shifting how people view themselves and what they want and their aspirations. And it's really cool. It's a really exciting time and I love what I do. I never want to stop doing it.
- [02:14]
A
Amazing. Well, my background also is in marketing, so I get the passion for it. It's exciting stuff. Talk to me. So right now you're, you're, you Know, you're working with companies and you're helping them implement AI stuff, you're working with consultants. What has been give us maybe like one or two of the most exciting or transformative areas that you see AI having like just the most lift in a company, like being able to do the most. Is that like in blog writing, SEO, is it in like content creat? Like where are the areas that you're most excited about?
- [02:41]
B
No, it's interesting. A lot of our clients are coming from health space, government, and those are not typically the industries that you see experimenting with generative AI or you don't hear about it a lot. And so for me, my answer would be it kind of widely differs, but I think that the biggest potential for AI in those organizations are pulling the data that they already have within their tech stack and gathering insights from it. Whether that's to iterate their actual product and service or whether it's to create a better customer experience for prospects. It's in the little, little things. And I think that that's it. I don't think that it's necessarily a one use case because with healthcare, for example, you know, we are working really hard to create content that can be translated into different regions. So you have a global campaign that a lot of budget's being thrown at and localizing that so that it's culturally representative of every area and everybody feels like this experience is for them. A great example of this would be the newest Apple ad that came out where it was about hearing aids. And everybody resonated with that, even if it wasn't, you know, in a specific language, like there wasn't really any, any words. You could just feel it. And so I would say that there's not one specific use case. It's more so just making brands come alive in ways that they already knew with tools that they already had. They just needed maybe that generative AI glue piece or the strategy behind it to bring it to life.
- [04:12]
A
Okay, I love that. I love the customization aspect. You know, I've seen a bunch of companies recently and I'm sure you're looking at this kind of stuff. I'm curious to what degree you might do any of this stuff. But like, so for example, there's the whole industry of customer support, like a lot of telephone kind of stuff, and with a lot of the new generative AI voice things. One thing that I think is just such a fascinating concept is taking whatever your area code is or wherever someone lives and you essentially. It sounds funny, but like you can do, like, a voice mask over someone that's talking. And you essentially have, like, call centers in the Philippines or whatever that are doing customer support. But it changes their accent when they're talking to. Not just like an American accent, perhaps if you're talking to someone in America, but you can go localize to, like, their exact area. So if I'm in Alabama, I'm gonna have a different accent than if I'm in Arizona or if I'm in North Carolina or if I'm in New York. And to me, that is just like, the most fascinating concept. Now, of course, I know there's a lot of people out there that are like, oh, for like, spammers and scammers, like, this is bad. And like, yes, this is definitely something to, like, I think, put a pin in and be aware of. That's kind of, I feel like on the. On the, like, AI provider side, they're working to, like, you know, ensure that there's a way to report or flag bad actors, all that kind of stuff. But honestly, like, I think even from, like, a healthcare provider, I would. I think it'd be super cool, like, if I'm getting a call from someone and I'm trying to work through, like, a medical bill or some random thing. I don't know, like, to me, it's kind of cool if they. They have, like, the same accent as you. Um, I don't know, it's just such a fascinating concept. But, like, I think you apply that to so many areas. Yes, that's like, one way where it's like voices, but like, you mentioned, I guess. Yeah, tell us. Tell me about some of the, like, customization things. Are you looking at, like, imaging, like, when you're targeting people, it's like, oh, this is like a Hispanic male. So we're going to have ads where we use generative AI to, like, target them that way. Like, how. What does that look like?
- [06:02]
B
The customization aspect, it really depends on what type of customer and what type of product. Because obviously when you're talking about specific companies and healthcare and tech and space, there's lots more regulations. So it depends, I think, to hit on your point earlier with the calls, one use case that was particularly helpful was implementing. And there's a couple tools, so I won't name them, because depending on where you sit and what your regulations are, maybe you can't even use the tools to begin.
- [06:27]
A
Right.
- [06:27]
B
But as far as, like, phone calls go, instead of optimizing the customer support, optimizing the experience for the customer themselves. So every time that you call in, you have to usually, like, verify who you are by providing, like, your address or verifying your Social Security. And that's already really risky. And that doesn't even have AI involved in it. So instead of continuing on that from an operational standpoint, because it takes time to have these conversations and calls having an AI functionality built in so that within 10 seconds of somebody speaking and saying their name, it automatically picks up on whether or not that person is who they are. So that you don't have to ask those questions. You just go through it. They opt into it because it's their data. And then they verify who they are the first time, and then after that, it will pick up whether they are who they say they are or not based on their voice.
- [07:17]
A
That is. That's so fascinating. You know, I. I swear I've been on, like, calls with my bank before where they're like, hey, you know, like, like everything that you call, they're like, hey, this is recorded for whatever, quality assurance. I have recently heard a couple that are like, we're recording this for identity verification purposes, if you're cool with that. And I'm like, whatever, like, just give me to the person faster. I record my voice. I don't care. Right, Yeah. I mean, I'm sure there's. There's still, I'm sure an aspect of it where it's like, if you use a tool like 11 labs to, like, clone your voice, I wonder if people could try to, like, abuse that system. I don't know how it's built, but anyways, it's very. That's very, very fascinating, fascinating area. So tell me a little bit about what happens if a client comes to you. Like, so you're. You're consulting some of these big companies. What does that process look like? A client comes to you and is like, what. What do they say? Are they just like, hey, we need to use more AI, what do we do? Or do they usually come and say, hey, we have a specific problem. How do we solve with AI? Or like, what does that conversation look like with you, typically?
- [08:16]
B
So all of my leads have been inbound, I think, purely because I have not done sales, so I've kind of just been stalling, putting together an outreach sequence I haven't had to do outbound. And so that's a little bit of a muscle memory thing to me that I need to work on so I can confidently say that all the clients that, that have come to me have always said, we know we want to use AI, we know that there's room for improvement. That's really the commonality. They don't know what they need or what they want. They just know that there's more out there that could be solving the problems that they need solved. And the tools in their tech stack do not solve that. So a lot of times it's something like, you know, they're, they're transitioning CRMs and they don't know which one to go to or that process from taking the data into one and putting into another is really, really time consuming and will manually slow down operations. So like, is there an AI tool that will check for duplicates and migrate that data for them so that they can kind of get back to doing their work in strategic. That's really, I would say, like what the commonality is between everybody is they don't necessarily know what they need or want. They just know that there's something out there that they could be doing to help the business. And that's a great starting point, is just being curious and wanting the business to do better and wanting your tech stack to be better. So I would say after that first conversation and me kind of hand holding and diagnosing the business problem, that's when they kind of get that confidence back. Because as executives they've seen these problems, they've solved these problems. But as you get more strategic, the responsibilities are less about what you're doing on a day to day basis and like managing outcomes. And so if you lose that muscle memory of experimentation, it kind of gives you like this fear paralysis of I don't know what's going on, but like something needs to happen. We need a solution here.
- [09:58]
A
Okay, yeah, that sounds, that sounds amazing. I think that sounds really useful. If a company like listening today wants to work with you. What are the kinds of things that you typically help with? What are the kinds of things? But yeah, you could, you know, I guess like what's your, what's your pitch to, to companies that are like, we know we need more AI, we know there's things we can automate. You know, what, what are things that you would be able to help them with?
- [10:22]
B
Here's the three business problems that I solve for. Because I always say, especially with technology, you've got to start with the problem and then find the technology, find the consultants and the resources that are here to solve that problem. So the three problems I solve for are lack of clean, enriched data. So if your CRM is buggy, that is the nucleus of your organization, I can help you get that back on track without having your whole entire workforce be stalled for a week. Lack of insight into the customer journey. So if you have digital touch points and marketing's bringing in leads, but sales doesn't know how they got the leads so they can't close them, that's something that I can help slow sales cycles, that's another one. I get really nerdy into these topics, but at the end of the day it's really just going back to the basics and do you have a strategy? Do you have a roadmap? If not, let's create you one. Are you married to your tech stack? Is it doing what you need it to do? And then let's fill in the pieces with some generative AI tools, build you some automations and then even some of my clients are using agents to manage those workflows of automations that are connecting the generative AI tools. So it's really just audit your business. If you have one of those three problems or you're starting to hear some rumors that could be connected to one of those three problems, then those are what you should reach out to me for. And if you just start with creating a roadmap and that's what you want access to and as you kind of have a relationship with me where we build a roadmap and perhaps I help you vet some vendors, create your IT policy and your usage policy for your company, then you decide to go in and actually use technology to solve one of those problems. But it doesn't have to be an all in step. This is very much phases and a la carte because I don't think anybody should be rushed to make a decision on everything all at once. They can tiptoe their way in.
- [12:09]
A
Totally.
- [12:10]
B
Yeah.
- [12:10]
A
No, I would definitely agree with that. So you mentioned agents and I'm curious what you're seeing there. Well, I guess first of all what are you seeing there? And then let's talk about where I think this is going to go in the future because I think that's a really fascinating conversation. But like you mentioned, some of your people are, you know, some of your customers are using agents to help manage some of these workflows and stuff. What does that look like? What are these agents doing like on an actual like operational level? Like. Yeah, yeah, tell us about that.
- [12:34]
B
So I have agents running pretty much everything for me on a weekly basis for my own company. So the agents that I'm setting up for myself versus customers are a little bit different I would say. But some of the agents that I'M setting up for customers, for example, are compiling leads and the automation is they take the leads from, let's just say a Google spreadsheet and they put those leads in through an enrichment, like a clear bit and then they put them in their CRM and then the agent actually starts to email out those leads based on the priority and the enrichments. And so you have a lead, you have all this extra information about them. You know what their problem is? Maybe you've scanned G2 and Capterra to figure out like what they don't like about their current tech stack. And now this email goes out from an agent every Tuesday, Wednesday, Thursday morning at 8am and then every Friday I get a report of here's all the emails I sent, here's open rates, click throughs and responses. That's one great example.
- [13:32]
A
That's incredible. Yeah, that's amazing. What, like, what tools are you using to create these agents? Is it like make or like zapier, like automations or like, like how are you building? Is there an agent platform you're using?
- [13:46]
B
I love Agent AI only because I'm very bullish on HubSpot and Dharmesh is the creator of HubSpot. So I like Agent AI because the user experience is mimicked. I really go with whatever. My customers love languages. I have a lot that are very pro Microsoft and so whenever that's the case, I'll use Power Automate and I'll code my own agent so that everything can stay in that interface for them. I like a zapier. I don't mind that. I'll use Make, Coda, N8N. It just kind of depends. I've used them all. I think that each one serves a good use case for certain areas, but it's not like a one size fits all. It kind of depends on like how much coding is needed, who's going to maintain it. Because that's like a non glamorous talk that doesn't come up a whole lot. Whoever puts these in place should probably be staying at the company for a really long time and have a high enough level of clearance because it is a lot of superpowers that you have.
- [14:39]
A
Okay, so something that I think a lot of people talk about interchangeably is kind of like automations and agents. Let's talk about where the future of this goes because I think right now even what we call agents probably won't seem like agents in a few years. Right? Because a lot of these things, they're, they're AI automations. You set them up they, they could go look at data, they can grab things like you mentioned and put them in a CRM and do these things. But like when it comes to like a real true pure agent that we're hoping, you know, OpenAI says they're working on next, for example, or even Anthropic kind of came out with their, their Computer View tool where it can actually like go take control of your computer and do tasks. How do you think those tools becoming more prolific is going to impact marketing, sales, all of this? How do you envision those being used, right, where you actually have like a piece of software that's like beyond just following a set like line of automations that you've set up, but it's kind of like making a little bit more decision making and actually going and doing things like what does that look like?
- [15:38]
B
I mean it's already happening. Zapier is a great example, right? They are an automation company, but they have central agents that they just rolled out and by the way, they're in beta, but they're amazing. I have them set up everywhere for myself and a couple clients. And so I think that you're going to see more of the companies that have automations in place adding AI agents because it just makes sense for the users that are creating automations that they want their agents to run to do that on the same company platform. It's the same user experience. You don't have to go learn a new website or how to create a new agent if you're using automations and zaps, because even though it's two different things, it's still the same platform, it still looks the same, feels the same. So it makes the onboarding experience easier. So I see a lot more companies like automation companies and the Generalists like Anthropic OpenAI creating their own. But I also see in the near future, you know, there's lots of folks who, I mean marketers are a great example who are not super tech savvy by nature, but they have very specific workflows that they need the agent to do and to operate on. So I feel like we're going to have an E commerce situation where you can buy this front end and this back end. Because if I like OpenAI's back end, but I don't like their front end experience, perhaps I'll use a Jasper. Right? And it's the same thing for agents. And as far as how it's actually going to disrupt the workforce, I think that we're going to start to see a lot of conversations happening around licenses. Because when you get a license for a tool you, they usually issue one right per plan and then you have to pay for another license. So that means that if you had an agent running your workflow, then you as yourself could be on that tool from 8:00am to 8:00pm Your agent would have to start at 8:01pm and stop at 7:59am So I think we're going to start to run into like these technical issues and, and disruptors and there's going to be almost like a subscription based license fee for lots of these agentic tools so that you don't have to have a, you know, two license issued for one tool. You can have a subscription and in that way you can have as many licenses as you want for X amount of tools. I see it more being a playground, for instance.
- [17:52]
A
Yeah, okay, I love that. That is such an interesting concept because yes, you're right, that is an issue where like let's say you have 10 agents running different tools. Like technically OpenAI wants you to have 10 seats on their team's plan and you got to buy one for each of your agents. But you're like, hey, but I only use this agent like so so often. I'm currently building a platform called AI Box, which is a pretty much a playground. It lets you access all the different AI tools and you can build, we're working on the building workflows and automations and agents, all that kind of stuff in the future. The way ours works, which I think you kind of touched on and probably is the way to go, is you essentially just buy credits. You have a whole bunch of credits, you can use as many tools simultaneously. Agents, everything can be running at the same time and it's just based on a credit system. Once you use your credits, you refill the credit. So you're right. I think there's gonna be, there's gonna have to be some pivots like that away from just like seats when things kind of run at different times and there's different amounts. Right, that's very interesting. Yeah, it's a very interesting concept. So when you're talking about the agents from Zapier, were you talking about how they, they did their like 3000 agent integrations or what was the platform you're taught? What was the tool or platform? You're talking about this in beta.
- [18:59]
B
Yeah, it's zapier, but it's, pardon me, zapier. I, it is just, it's central.zapier.com box and you build agents Right on it. And it works perfectly because you set it up, you set it up differently, but it looks the exact same. So when you're setting up a zap and you're setting up an agent, it remembers those automations and it's nice because if you have your automation set up there and then create your agents, when there's an issue with your automations, it will tell you how to fix it because you're still within its own website and it's still the same company. So then you don't run into the issue of you got to go over here to this automation platform, fix it, come back to the agent, troubleshoot that. It's like self, self cleaning, right?
- [19:42]
A
Yeah, no, I love that. I think, I think they've definitely nailed that. I'm trying to build something similar. So I like how they're doing that and I think the big problem that they're solving is we have, for example, someone like Anthropic came out with, I believe, their computer vision or their computer view where like you, you give it a, you know, a giant list of things to do and it goes and executes those tasks. And I think the problem with that is you, you want the agent to go and execute something, but you actually kind of want it to do it in a very specific way. So I think using something like Zapier, where you actually, you've built certain automations, you tell it to go accomplish the task, but you tell it to go use your automations to do that, you know, you're going to get a very predictable outcome because the certain segments or chunks, you, you, you figured them out and maybe you even have some of your own personal company data in there, some ip, some things that you know, aren't publicly available on these AI models. Um, so yeah, I think that's, that's phenomenal. That's such a cool tool. Oh my gosh, Ashley, this has been amazing having you on the podcast today, sharing your insights. Uh, I've learned so much. I, I'm, I have a whole bunch of new tools to look into. I'm sure there's a bunch of listeners that are feeling the same way, I guess. Before we wrap this up, one thing I would love to ask you is what is one big prediction you can make about or that you think is going to happen with AI and marketing, AI and sales over the next one to three years? Some big things that you think people should be looking out, coming down the pipe or things or you know, I guess big direction shifts you'd see in the industry.
- [21:07]
B
Lots more jobs. Lots more jobs around just what we, what we both were talking about just now. For instance, the ability to create and deploy and, and manage and iterate these agents is not easy. It takes a very long time to, to optimize generative AI tools in order to put them on an automation or in order to create an AI agent that manages all of that. That takes a special set of skills. And as marketers, we already come with that domain expertise. So like we know the workflow for market research. What comes after market research, you probably want to create content based off of that research. So there's certain skill sets that we already have. But I fear that because it's new technology, we feel like we don't know as much as we do. So I think my biggest prediction is there's going to be a lot more jobs. People just need to figure out a way to pull what they know about marketing and roles and, and creating really good experiences and use the technology to do all of those things, because that is going to be where the demand is and it's only going to continue to grow from there. If you can write really good copy, that's amazing. But if you can use AI to take the copy that you wrote yourself and iterate it according to the best practices and optimize it per channel, that is where I see 2025 and beyond going is the ability to use technology to do multiple things from one perspective and manage it.
- [22:35]
A
Amazing. All right, well, I'm excited. I will definitely have to have you on again. Listen, if you guys enjoyed the podcast today, make sure to go check out Ashley gross over on LinkedIn. I'll leave a link in the description to her LinkedIn to reach out. If you have a company that wants to work with her, I'm assuming Ashley is at the best place for them to reach out.
- [22:51]
B
Yes, please.
- [22:53]
A
Fantastic on LinkedIn. All right, fantastic. Thanks so much for tuning in, everyone. Make sure to leave a review to the podcast if you haven't already, and I hope you all have a fantastic rest of your day.