221. “How to Trust and Embrace AI for your Business” with Jenna Nelson

Dawn Kennedy talks with Jenna Nelson, founder of HerAIgency, about adopting AI in business in a supportive, non-threatening way, especially for female founders and women-led companies. Jenna outlines practical AI use cases in marketing, operations, data analytics, and market research, emphasizing that results depend on the quality of inputs (“slop in, slop out”) and that AI must be trained with a robust brand knowledge base such as briefs, past content, and transcripts. She recommends choosing one platform to train deeply, using paid features like custom GPTs, and reinforcing learning by feeding edits back into the tool. They discuss privacy and trade-secret concerns, advising careful platform settings and guardrails, and stress a human-in-the-loop review process. The conversation also highlights AI for lead intake, pre-qualification, and faster response to improve conversion and client experience.

Who is Jenna Nelson, and whom does she serve? (01:06)

AI Use Cases Overview (01:36)

Brand Voice Concerns (03:23)

Align Before Automate (05:22)

How to Train AI (07:32)

Tools and Paid Plans (09:39)

Market Research Faster (12:25)

Privacy and Trade Secrets (16:14)

Human in the Loop (19:56)

Trust and Disclosure (23:23)

Content Volume Reality (27:04)

Lead Intake Automation (31:09)

Where to Find Jenna (37:35)

Henna Nelson’s Information:

Website: https://heraigency.com

YouTube:https://www.youtube.com/@heraigency

LinkedIn:https://www.linkedin.com/in/jennalnelson/

Instagram: https://www.instagram.com/heraigency

Facebook: https://www.facebook.com/herAIgency/

Transcript

Dawn Kennedy: 0:03

Welcome to Business Mastery with Dawn Kennedy, your quick under forty five minute dose of expert insights and strategies to make a positive impact on your business and life. Let's get started. Hello and welcome to this episode of the Business Mastery Podcast. We're gonna be talking about AI in your business, but we're going to be talking about embracing it in a way that doesn't feel threatening to maybe how you're doing things today. I know that there's some resistance about AI, and everything you read in the news is bad. So we're going to actually change that conversation and change that perspective a bit so that we can leverage it then in a way that's helpful and supportive. And we don't have the same amount of fear, I think is the best word around what we should be using it for and what we can be using it for and looking to the future. So I'm here to tell us all about We're going to work through all So thank you so much for joining

Jenna Nelson: 1:03

Thank you. I'm excited to be here.

Dawn Kennedy: 1:05

Can you tell everyone who you serve?

Jenna Nelson: 1:08

So my name is Jenna Nelson and And I primarily work with female businesses to strategically like you talked about. That doesn't feel threatening, doesn't disconnect you from We're not trying to let AI take We're trying to strategically implement it in ways that actually makes.

Dawn Kennedy: 1:35

So some of the most common ways somebody who even doesn't have a What would you say? So I've seen some stuff about What else should we be

Jenna Nelson: 1:49

Certainly marketing is a great I think we all feel like we more better marketing, or we're media, getting emails out. So that's always something that up, but also look at it in terms How can we better improve our client experience by incorporating AI? Can we get back to them faster? Can we move forward in a process go through maybe a long initial that employee doing something helpful to the client? So certainly there's some And then certainly data analytics reporting, that is something that before you may have needed a CFO to really help you understand your cash flow and where stuff is going and what you're spending the most on. And now you can get to those answers and that level of data analytics with AI, and not have to have a full C suite to do it, which I think is a big benefit, particularly for smaller businesses that are running lean.

Dawn Kennedy: 3:04

Yeah. And the fractional CFO and that world probably could be that additional data collection And then of course, humans still have to make those decisions, right and weigh those costs and risks. Absolutely. Let's get into just some of the might give people pause. Let's start with Mark. I have seen so many posts that written by AI. It's AI everywhere. It's always AI. So in the non AI, non-digital world, I know that was probably a thing with copywriters for years. It was like they have my brand Do they understand what I'm And so we have humans that became copywriters and worked with clients to get all that down. Let's talk about AI and brand presence and voice and the to sound, look, feel, taste, Yeah.

Jenna Nelson: 4:09

One thing I talk a lot about is the two sided responsibility model. AI does not think for you and it input you give it and the the parameters that you give it. So you have just as much give it the context and the output that you actually want. Slop in, slop out. Right. And that's a lot of what you're talking about with what we see online is. Yeah, absolutely. I can tell with somebody who's ChatGPT and said, make me a And that's all the context They've never trained this AI they it doesn't know anything So yes, it's going to look It's going to feel very generic. It's not going to feel like their brand, but that's because they haven't given it enough to work with. And that's no different than And if you hired a brand new executive assistant and gave them exactly those instructions and nothing else, we wouldn't expect to get great results either. And AI is no different there. So what if things that I focus first step is aligning. And I've got three step framework align, automate, appear that I take my clients through. And that alignment step cannot I, I start everyone there because I can't train an AI tool to understand your business if you don't understand your business yet, if you can't verbalize. I find that all the time, right? So much of our business lives in we haven't written it down. Maybe we haven't really thought What is what are their pain So I find when I go through this process with business owners, there's just a lot of brand strategy gaps that they haven't filled yet from a human perspective. And so it does two things. One, I think it helps align because they get a chance to questions and those answers and brand strategy, but then I can to understand exactly what that who they serve, why they do it. The first step to anybody using AI is you have to train it, and then you have to give it context and parameters. It's something I always like to a lot of it is, okay, do this. We're prompting, we're saying, And I want people to shift their thinking on that a little bit and think more about the outcome that you want and work backwards. And if you give AI that this is need to land. Here's the background, here's Here's my brand freeze. It's going to get you there without you having to do the work to figure out how to instruct it. So you can really get away from prompts in that manner. So that's really where I start We've got to understand your And once you do that, you're going to really elevate what your AI tools are giving back to you.

Dawn Kennedy: 7:31

So when you say train your AI.

Jenna Nelson: 7:33

Let's figure out what.

Dawn Kennedy: 7:34

That means to train an AI, Because a lot of people use it, I'm going to say more like Google, and they'll put pumps in and maybe they'll put parameters around it, like location parameters or price points or something, especially in market research, but training an AI to support you. What does that look like? Because what I'm hearing is you But you can also train the AI to You can actually ask it a question and say, what am I missing? What am I not considered? So how do we train an AI? What does that even mean?

Jenna Nelson: 8:14

Yeah, it's a little different platform you're using. But the basic thing is you need to create a very robust knowledge base. For me, that's a full brand That's everything from tone things like that. You want to give it if you've got good writing, if you've written a blog you're really proud of, if you've written a book, if you've got any kind of materials like that, even transcripts. I take every transcript of every podcast I've done, every webinar I've done, and I feed it back to my AI tools because it learns from the repetition of what I say and how I say it, that reinforcement learning is really important. So take all of the things that or your business feel like you. Great pieces of content, things And give it to the AI to learn and say, hey, read through this, ask me question, ask follow up questions. What's unclear? What does it feel documented? What processes do you still have So ask it like you said, use it back, bounce back, feed back and say, okay, what am I still missing? What do you not have? And by then you've got a really great knowledge base already baked in. On ChatGPT, that might look like That might look like a custom I still really love custom GPT because they are so focused on exactly what you train it for, which I think can be really valuable for certain business aspects. But you could go into cloud and So you could do this in cloud You could do this in Gemini. Gemini certainly has all of the access to your Google Drive or Gmail. It has far more information than So you can get really well trained tools without a lot of effort. And the more you use it, the And that's something I caution people about too, is don't try to use every tool that's out there. Pick one and get really good at that one and get that one really well trained on you, your brand, your offers. And then there are certain better than others. And certainly you can take something and say, make Claude tends to be a bit more of a narrative writer. So maybe you take something over polish on it. But if you're always using a different tool or you're going to get really disconnected because it's not learning and it's not noticing what you're changing. So my other two about training you edit it and you paste it a ton of effort edits, copy that tool and say, this is what I Because otherwise it's going to It gave you exactly what you We move on. So that reinforcement learning, you would do differently. If it tells you something wrong, get you up to speed faster.

Dawn Kennedy: 11:26

I do want to point out, I agents, you have to have a paid dollars or fifty, whatever it's investment, I think, to have the But the free versions, I don't

Jenna Nelson: 11:45

It has very limited memory. You certainly can't create custom gpts with the free version. You can use them, but you. If somebody else has given it to But yes, paid versions, they're. They're there for a reason and You don't need to pay for all of And you don't need. I personally use the twenty I don't need the two hundred And I work in this all day, So it is a pretty minimal And I guarantee you you can get benefit out of it in an hour a It's absolutely worth it.

Dawn Kennedy: 12:23

Amazing. All right, so let's move off of Let's get into market research, a place where just honestly, business owners don't frequently simply because it So you had to go and you had to There were a million places to search to find the industry information. There may have been something industry, but you had to know to So how do you see AI supporting We've been using it now for about a year to help us guide, I don't know, all kinds of decisions. But again, it's a trust, but But how do you like the AI and the things for market research? And how does your client base

Jenna Nelson: 13:18

Yeah, absolutely. I think there's a couple of The first is just the data You can take one spreadsheet and to summarise and categorise and pieces that we used to have to Excel formulas, or try to out what was happening. So you can get that analysis ever used to be able to. Right. And so this is no longer spending hours poring over spreadsheets. This is you can have it done in click through rate is really So just the data organization, I something that we haven't been The other thing too, is the competitors, the deep research Google and just typing in our what's out there, right. You can go into I like ChatGPT deep research because it sites better personally. The clod does. But you go in and you say, here I want to do some research on trying to come up with and let Let it save you that four hours And what's great too, is it may find stuff that you hadn't thought of. Maybe there's a competitor you're not even aware of, somebody that's coming up that is on your tail and you don't know yet. So it can do a great job with And then I think the third and this is moving into a more advanced use case of it, but with Agentic AI, you can get these agents to run this market research for you weekly, daily, every six hours, if that's what you need. And you don't have to touch it, You set that up once you get it It formats a report and it sends you a report every Monday morning of here's all the data on your competitors for the past week. Depending on how much data you need, how frequently you need it. Agents can be a really great way time, and I think find things otherwise, which is a huge

Dawn Kennedy: 15:46

Yeah, I do want to point out, because this is something I'm incredibly fascinated with, is that business intelligence in that small to medium sized businesses don't have floors of analysts. They don't have people who are research and those things. But they I can actually go in and be like, oh, we'll take the report from Statista or whoever else, and we can distill it down to the relevant data for smaller businesses.

Jenna Nelson: 16:11

Absolutely.

Dawn Kennedy: 16:12

That's what I really like about But I know there's going to be I'm going to ask the question, business up into AI, does that And now my competitors also see And if I use AI, can I trust it We know that there's been some stuff on the legal side where if you go in and you put things in, it doesn't have attorney client privilege. So that's been working its way But what about trade secrets. What about what my great ideas Right? So I do want to address that because I know someone's driving around right now, or running or biking and going, yeah, wait a minute. Can we trust this to be private?

Jenna Nelson: 16:55

Yeah, absolutely. And the answer is you do. You absolutely do need to be You shouldn't blindly hand it And the answer is really Open claw is really scary. If you don't know what Open claw does, please don't put it on your machine. That can be a really scary one. On the other hand, using ChatGPT putting those privacy filters in You can keep you there. There are options. There are settings to allow or to train models, so you can turn So that's potentially a way to But I would say, and it's obviously it depends on the business and what their trade secrets are and how secretive those are. But you want to treat it like Frankly, if you wouldn't tell somebody in your organization that you didn't trust, then probably don't broadly put it into AI. Now, again, there are ways and toggles and things like that you can work around, but you do need to be careful because yes, absolutely. The majority of it is retraining itself with the information we give it. And to some degree we want that. But for trade secrets, for please don't go put your Social chat, GPT or clipboard. But what is your.

Dawn Kennedy: 18:31

Customer list or your customer

Jenna Nelson: 18:33

Or your customer? Yeah, any of that kind of stuff. Credit card numbers. Don't get me wrong, but I think I think the fear is reasonable, but for the most part, it can be worked around. You just need to do your using, what the proper protocols And that counts too for not only what information you're giving it, but then also for if you're producing a tool that you're putting out there to let your clients use. You need to protect that IP as One of the things I do with my frameworks and guardrails built ask it like how it's trained, It won't do it right. It's going to shut down on them and be like, I'm sorry, I can't answer that. So that's an important thing something that is your IP, make in there and just reverse information right back. We don't want that either. So yeah, the security thing is I am not the top level expert on enough to be dangerous. Probably. But certainly my biggest tip is to be very careful with the agentic ones. In particular open claw. It's still. It can be a little scary for

Dawn Kennedy: 19:54

Okay. So training the AI is going to take a little time making sure that you've got the guardrails kind of set as you're using it. But let's talk about something We're so responsible to review it, to make edit, to make corrections. So I want to talk about that as Because let's say you do train this agent and now it does sound like you like a human copywriter. It is able to source things for you as far as graphic ideas and things like that, just on the marketing side. And you still have to check and make sure that those aren't copyrighted pictures. You need to make sure that the edits and things need to be done with the advancements that you have seen over, I don't know, the last six months or so though, it's. What are your thoughts and work that would come out of let's train the AI. Then I have to go back and do it It defeats our conversation. So let's talk about kind of what we would need to maybe just consider because as we're getting to that automated piece, right. Because I'm assuming some of that automation is scheduling posts or scheduling articles or being able to take and put things out into the world for your clients. You still have to review them. What is protocol for things like

Jenna Nelson: 21:13

Yeah, absolutely. I am a an absolute supporter of there should be a human in the loop at every phase of anything that's AI generated, AI published. Yes. I want to use it to expedite the process, to make it more efficient, to maybe even get a better result. But a human still needs to lay make judgment calls. Now, what's funny about that is that we have seen as recently as this week. People create AI images and But the judgment call was maybe So there is always going to be that element of should you post that? Is that appropriate? We've seen AI used to create images of celebrities that they haven't endorsed that aren't acceptable. So there there, there is a human as the AI failure point. So just because a human is reviewing it doesn't mean that it's still not doesn't make it right. But certainly I want somebody content that goes out if it's medically sensitive, financially a higher touch point to ensure someone down the wrong legal And then I think you put your see it all the time. This was generated by AI. There might be mistakes. So put some of those disclaimers in the way. This was produced in conjunction But there may be put those just in case. But yes, certainly checking not copyrighted, making sure I haven't really seen any strong examples of AI specifically plagiarizing anything, not thinking that's really a big issue, but it's not really just taking stuff and regurgitating it. So that one's probably not a big But imagery for sure has been a Yeah.

Dawn Kennedy: 23:22

I recently was watching a case study nerd, but it, it phrase, the opening frames, it or written, filmed, edited And I thought, oh, that's Are we almost to that point where there's that much of a trust issue? Because as we are encouraging more people to use it for the efficiency and we are encouraging people to train them. Do you think that there's a lot of hesitation where those disclaimers are going to make you more or less trustworthy in the marketplace?

Jenna Nelson: 24:02

I think my feeling on it is that but where that fear is coming it's exactly where we started. This conversation was with the aren't valuable out there. And unfortunately, the more people see bad AI stuff, the And the least, the less trustworthy it's going to come off. So this is my point where I'm like, this is our opportunity to make sure that we're using AI in a good way so that we can overcome that. My feeling on it is if the piece of content that you're putting out there is valuable, it's entertaining, it's educational, it helps someone in some way, then I don't personally care if AI was used partially or entirely. Right. If it makes the goal and Google In terms of how content is complete AI filter. I know a lot of people think that's AI written. That's not true. If it is still within, they call content, then you can use AI. It's just about it being good So how you come about that is not as important as the actual product itself. So I would encourage people to. I fear cautiously. Certainly. Don't believe everything. Don't trust everything that's That's always really been the But don't assume that because something is AI, that it's not valuable. I think is the big takeaway If you saw a YouTube video where it said this has been written and produced by AI, don't immediately know I'm not watching. Check it out first and see. And if it is bad AI, then tell they'll stop it. Yeah.

Dawn Kennedy: 26:03

So I think that the backlash and things that we're seeing is with the AI stuff that doesn't have human in the loop just feels ingenuous. I didn't put the effort in. You didn't put the time in, I So using AI to, let's say, the market research, right? Like what if we used to do this If you remember, there's a fun and you could research what the you could see the questions that And you could write posts, or you can do education, or you do videos based on what the top Google searches were on the topic. So I'm not sure that it's a generate five topics based on considered, maybe I doing been, at least recently, some behind the use of AI. So I do want to talk about that that as business owners, we need the customer experience and our to be able to meet with the is no longer a couple of posts, of whatever you people need to don't make a decision to follow latest research and thirty times

Jenna Nelson: 27:34

And it used to be seven and now That was like ten years ago that it was seven, and then it became twelve. And now it's thirty. Yeah. It's the amount of content that it takes to be visible is insane. And, and because there's so much Everybody has an ability to have a YouTube channel now, good or bad. And so you do, you have to be And that sometimes that's where Yeah.

Dawn Kennedy: 28:06

What it used to be, like you into the marketplace versus now If you think about AI stepping where you can meet that market spending another six, eight myself who are not graphic am not not somebody who can what social media would accept. I think that I just want to put behind this and the use of it But if you don't embrace this and you don't get to that thirty or whatever, you're actually likely going to be less competitive, at least for visibility. Even if you had greatest product or service, I think it's becoming easier and easier to become the world's best secret, which I don't know how to feel about that. And I don't know how other people feel about it, but I think that's just a market reality. And I think this is the place to tool to help you not become everybody else is just running

Jenna Nelson: 29:22

Absolutely, absolutely. Yeah. And there is a necessary need more content, but it also needs to be valuable. But that's again, where I go Like you said, you're not a That's not your skill set. But if you're really good at writing the content, if you have something valuable to say and you're using AI to create the graphic to get it on Instagram with your caption, that is helpful. I don't, I don't see a problem I think that is great if it So I always encourage people to you're not good at. Use it for the things you hate doing or the bottlenecks in the process, right? If it's graphics, let AI do Now. It's very capable. We've come a long way in the It's actually spelling things eighteen fingers and we've come Use your brain to do the big valuable Instead of burning trying to create a social media you're anything like me, you be like, and I still hate this. That's not a valuable use of And it's not benefiting your

Dawn Kennedy: 30:46

And I let you keep saying if it's valuable and useful for the client, the training of the AI is probably really, if we're honest, less about the business owner converting in sales and much more about going out in the marketplace and just being very consistent. We would have had a human copywriter and a video producer and all these things a few years ago. Where else do you see AI maybe needing consideration, if not full, embracing for businesses over the next eighteen to twenty four months?

Jenna Nelson: 31:20

Certainly, I think one of them its speed to lead, right? You can lose a lead in Two the phone and your competitor a text back right away or Ultimately, somebody calling And if you are able to answer them a text back or an email that shouldn't necessarily lose you're not good at what you do. I see this a lot. I work with a lot of small don't have the manpower, right. The attorneys are in court, the back end stuff. And maybe they had a receptionist or an admin, but the volume is just like, she can't answer every single phone call. So I've helped several of these agents in place that collect their intake link and say, this Please fill out this form and you in twenty four to forty this is the case we can support. And that sucks them from missing the call and then picking up the phone to call the next attorney, right? That client feels like they've They've been helped. They have the initial step that they need to take to, to move forward. I think that's a great use case, Instead of hiring somebody to sit there and be on the phone all day long, let the voice AI handle that. I think that's or a chat widget on your website, use it the same way. So that has been an excellent I also think lead filtering is I personally have gotten on a lifetime that they weren't the that was a waste of my time. But more importantly, it was a waste of their time to write, and it's awkward to get on those client consult calls and be like, yeah, you don't actually have enough of a budget for what? And that's awkward. So I love using initial lead intake filtering because it pre qualifies them. And if they're not the right fit, it politely excuses them or directs them to a different resource. We find that a lot the law firms I can't tell you the amount of law firm will get for things something that they handle. And then they leave a message, don't call them back. And you weren't even calling the So now we can say, hey, we don't we recommend or something like So the lead intake filtering, it I run a lot of digital ads, and I can put people through my five consultative questions that tell me whether you're ready for ads or not. If you're coming to me without a a funnel yet, uh, you think dollars a day on ads? Whatever that is, I can save you you in the right direction. And I like to make them just yes, no, yes, it's not a It's not. Just check the boxes. Okay, here's your score. It's. Thank you for sharing. A really good starting point for Google ads is at least fifty dollars a day, whatever that might be. So that I think is another gives your client the immediate Instead of leaving a voicemail instead of waiting for a call back, you get that touchpoint immediately. And when you get to helping them all want at the end of the day.

Dawn Kennedy: 35:18

Yeah, the convenience and the speed and the lack of friction that we experience now in every transaction. I actually wrote an article on this few days ago from this recording. But yeah, it's the post pandemic in two days, or I shouldn't have reorder or there's.

Jenna Nelson: 35:39

Like.

Dawn Kennedy: 35:40

Yeah, there's a huge I don't know that the use cases been accepted the same way five

Jenna Nelson: 35:50

No, no. Probably not.

Dawn Kennedy: 35:51

And people still wanted to talk to a person on the phone about their legal matter six years ago. And today they're more form without talking to a human. I think for small and medium we have to look at the change in they're not going to they're not information, but they're maybe, I'm willing to give them a is before we take the next step. And I definitely can see this, just meeting that client leads to more brand loyalty just

Jenna Nelson: 36:32

I think it does, and I think it as well, because you're also time they do get to a human, You've already taken them through some of the preliminary stuff. They feel supported or helped, And it's going to be easier to make that sale when you actually get them on the phone or get them into your office, which I think is really valuable as well.

Dawn Kennedy: 36:57

I don't think I've talked to integration space that talks you, but I think this convinced me that it's yeah, stuff about the do loops on chat service, but lead intake and completely different areas in I think I'm convinced that AI could warm a lead from using it correctly and in the brand voice and things. So fantastic. This conversation has been How can everyone find you and social, or do business with you? Where can.

Jenna Nelson: 37:40

I? Certainly. So I do a lot on I do a lot of really just short, These aren't like two hour These are like fifteen minutes So definitely check me out on It's at her agency on YouTube. So check me out there. That's probably the most education I could possibly give you. I'm certainly on on LinkedIn and And then I also on my website, just talking about. I have a growth strategist on my you through where you are and which I think is a really work with me just to know, like lever I should pull? So certainly give that a try And if you are thinking about implementing a lead intake system, it's a good kind of demo process to see how it might work. Obviously, it would be customized to your business, but see how it works because I get a lot of really great feedback on it. So I don't feel like I'm just getting on a call and not knowing what we're talking about. Now they're getting on a call know I need this. And we can just, we're off to We can go and I don't have to So I think it's been really

Dawn Kennedy: 38:58

Amazing. All right. So we're gonna put all that down So you're going to get the We're going to put the website. If you find us on the day this everything will be there. And I appreciate your time and And thank you so much for really walking us through where AI should be embraced in your business. And this is a lot of this counter to what the media says So I think it's important to But more often than not, the media is looking for those clicks too. So they want to tell the worst

Jenna Nelson: 39:34

Sure.

Dawn Kennedy: 39:35

And here we really just kind of to think about things. So thank you so much for being

Jenna Nelson: 39:41

Thank you, I appreciate it.

Dawn Kennedy: 39:43

All right. We'll talk to you all next time Business Mastery Podcast. Take care. Thank you so much for listening to this episode of the Business Mastery Podcast. If you want to learn more about dot com and you can now check us course, any of your favorite Take care.