Episode 96 - Alane Boyd

Erik Martinez: Hello, and welcome to the Digital Velocity podcast. Today we're continuing our discussion on utilizing AI in our business to drive success. Our guest is Alane Boyd, co-founder of BiggestGoal.ai, the company specializing in creating AI agents to streamline work processes. Alane has been right in the middle of helping businesses use AI to accelerate growth by helping teams focus on what really matters. Alane, welcome to the show.

Alane Boyd: Thanks Erik. I'm excited to have this conversation with you.

Erik Martinez: I've been looking forward to this one as well. We've been, [00:01:00] putting more and more research into the process side of things and the fact that your team is already doing this at scale, is, absolutely interesting to me. Before we get started, can you give us a brief sense of who you are and what you're doing?

Alane Boyd: Yeah, so I started this company after exiting my agency that I built up and just kind of got bored in retirement and wanted to help other people make their workdays easier. I grew an agency, sold it, started this one, before AI was even mainstream. And so we were helping automate processes and workflows using Make and Zapier, as AI has come into play, things like N8N where you can build agentic workflows, we've incorporated that into what we do for businesses, but still at the same mission of making workdays easier.

Erik Martinez: No, that's awesome because I think that, we hear a lot about efficiency in the news. AI make yourself more efficient and that is definitely [00:02:00] part of it. But I don't think that's the main goal. I think the main goal is, can I free up time so I can do the more important work that helps me actually drive my business.

Alane Boyd: Yeah. Think about how many, tedious little things that people do in their day, whether they're the owner, manager, or worker. And it's just so many time consuming things that are eating us up and really burn us out.

Erik Martinez: Yeah, absolutely. You mentioned a couple tools, Make and Zapier, which I think a lot of people are familiar with. But you also use that term agentic AI.

Can you briefly tell us what agentic AI is and what it really means?

Alane Boyd: Yeah, and I'll talk about it in examples. That way it has some context because I don't like getting so technical that people can't follow me. Think about when you're using like ChatGPT or Claude or something like that, it is a great tool, but it's you manually putting in a prompt and it spitting something back to you.

It is up to you to do something with that. You can edit it, but it's telling you what you [00:03:00] should do. Here, you ask me this question, here's what you should do. There's an AI agent that's like an intern that's a part of a workflow, and that AI agent, robot intern is waiting for work and it knows what it needs to do. It says when this form is filled out, I am going to do this, and I'm gonna send the information here, but there's AI written into that so that it thinks like an intern. So that's why I like to use that word, like it's a skilled intern and knows what to do with the job.

When you're doing something in Make, and Zapier, you have to be so specific. You have to pinpoint everything, you have to map every detail. An AI agent, has a little bit of a brain. It doesn't need every detail to do the job. So building them is simpler, and the end result is more like a human, and it's operating without needing you to be the trigger point, you writing the prompt, it already has that in its brain.

Erik Martinez: I think that's great. So, what [00:04:00] is the simplest example of an agentic workflow that would apply to a small business.

Alane Boyd: Let's use this podcast as an example.

Erik Martinez: Okay.

Alane Boyd: So we've gotta record this podcast and then you have to go and take it. And get it out into the world. You need a title, a description, show notes, maybe you create a blog from the transcript, social post. We are recording video right now, so you might wanna do some YouTube shorts, video clips, things like that.

Now an AI agent can be a part of that process and it decreases the workload by 75%.

So all of these things have to be done by a person normally, and there are software, there's Decipher, there's all things that help, but it still is a manual step and a manual process. Well, when you have AI agents that are a part of it from recording to production, the workload gets cut by 75%.

Erik Martinez: So agentic AI is having this [00:05:00] prepackaged AI tool to help you do things more efficiently and more effectively. Can you give us a sense of AI BiggestGoal.ai range of the core business problems that agentic AI can help solve? Because I think sometimes, you know, the, this audience is heavily skewed towards marketing, but there's so many other aspects of e-commerce or business that aren't marketing.

Can we talk about a couple of those?

Alane Boyd: Yeah, so the most common one that we see for e-commerce, all the way to manufacturing distribution is customer support and how much time consuming work it is to support customers and, of course, there's already some chats like we use Rippling and actually their AI chat is pretty good for support.

So there are some things that are able to help, but a lot of companies lack the resources to build something like a Rippling did. [00:06:00] But it can be done now with a RAG database. So what are the questions that you get asked? You can save that in a Google Drive with Google Documents. You need to save that somewhere.

And then with RAG, we put that apart of the Agentic workflow. When somebody writes a support email or a support email comes in, somebody sends a chat message, then that goes in to that RAG database, finds the answer, and drafts a response. Now I think humans need to be a part of this to make sure and verify information is correct, maybe tailored a little bit, but we have customers that we've built that for and it's fast.

If you have your document, your knowledge base together to build a system like that is very quick. We love RAG database AI agentic workflows because you already have the trigger, which is people are asking you questions, and then you have some type of documentation, and we just need to get that into an AI friendly database, which could be a Google Drive with Google Docs in it, and then it spits [00:07:00] out the answer.

We see support teams go from four people to a part-time person. That's significant in employee salaries, to go from four full-time people to one support person handling everything, and that's just them being able to review. Especially when technical support jumps in. If you're e-commerce that has, that's not just a flavored drink company, but something where it's equipment based and there needs to be some technical help.

We have clients where the support team was having to go to the engineers to get the answer. Well, they eliminated having to go to the engineering team because now they can get their support and the AI agent can read that support doc put together the answer.

Erik Martinez: I've been a big proponent over the years of my clients mining their customer service calls and trying to figure out what are the most common questions, pre-packaging that, and then putting a tool in front of their their customer service agents, even if they choose not to use a chat [00:08:00] bot.

Alane Boyd: Mm-hmm.

Erik Martinez: Putting that tool where they could query it really quickly and get the answer so that they don't have to go, I don't remember, or I don't know, or whatever. And it's amazing how many companies, even with all of them, are recording their calls don't actually compile that information.

I saw this in action the other day. I had ordered something from Amazon and of course, true to Amazon, you don't always know where the stuff is coming from. Well, delivery date came, it went, so the next day I jump on and I'm like, okay, I need to put in a support request with the seller, and they have completely put in an agentic workflow.

It's kind of frustrating, to a certain extent, It's not quite as robust as I would want it to be because it wasn't answering the specific question that I really wanted to know was, "has this even arrived in the country?" We figured out that it was [00:09:00] started in India and it theoretically shipped. It did give me the ability to message the seller, which I did, and magically the next day, everything populated, but it kept saying, well, has this helped you? And I'm like, no, because I still don't know where the product is. So there's still some things to work out in the process . But it was a better experience than I would've had trying to get to a real person on Amazon.

Alane Boyd: Yeah, that's a fantastic example. So another one that I see really common with e-commerce that we do, and, and I'm gonna use an example that's not actually a client, but I think that it would help paint the picture and it's inventory matching.

So you have products and the customer needs to know what to buy and so there, this is not a client of mine, but this one just. I think paints what we do really well. Go Ruck. It is a weighted backpack company and they need this tool. So many females go there and we don't know what size rucksack to buy. [00:10:00] We're going on Reddit, we're trying to look up what size backpack best fits a woman's frame because the details are not on the page and we like to ruck, and there's a lot of women that ruck. Well, with an agent you could have a chat interface on there that says, Hey, let us help you find the ruksack that's best for you.

There's so many options. Should I just do a day pack? It's just the weight goes in and that's it or one of the other bigger packs so I can put other stuff in. But I've got several and I'll tell you just the simple one is best for the female frame. Don't try to get these other backpacks. It's too big for you. So what we do when we build agents like that for a company is we get the information, we build that database, and we put a chat interface on top, and then the customer can use that to find the best inventory match for solving their problem.

Erik Martinez: I think that's great. We've been working with another company, [00:11:00] MicroCommerce.ai, and they have built an e-commerce based agent to do exactly that. There's two benefits, there's not only the customer service benefit of helping the customer find what they actually need, but there's also the operational benefit of having that knowledge all compiled in one and that they can use for product training and training new reps.

Even though reducing the number of reps that we actually need to service X number of contacts, they still have to have knowledge because if the system goes down they still have to be able to function and operate and answer questions for customers relatively seamlessly.

One of the questions that keeps coming up is what are the real risks of using these tools or not using these tools? What are the practical things, as you're building out these agentic tools, that you're thinking about?

Alane Boyd: Day one looked very different than day 365 than [00:12:00] it did two years later. You know, your business evolves and what you do to service clients evolves. Well, your process is gonna be different than your first day trying it. Now, I'm not saying you need to go two years doing a manual process. It might be two weeks, but do something so you know what you wanna do, and then document it, process, map, bullet it out.

Whatever it is, write it down and get the people involved that are doing it because you don't want four different ways. You don't wanna get to the end of developing that workflow. And they go, mm, no, that's not how I do it. That's not the way that we do it. And then you gotta go back and redo development in that automation.

That's time consuming. And if you're working with a third party, that adds expense to you. They didn't build it wrong. What you wanted was wrong and they built it to your. Um, frame whatever you wanted. So then they've gotta go back and change it. And so start with something, you know, and start with something simple.

I [00:13:00] see so many times that we get into a process and once they really start thinking about it, it's massive. And so what I do and it, and for a good reason, right, we do work, but let's just, what is one piece of this that we could automate? To start that way, your team isn't waiting for this to happen. They can see it in two weeks.

They can see the automation working and it can alleviate one piece. Let's just start there and then we can build on it and we can add other parts to it.

Erik Martinez: Yeah, I think there's this notion that, um. We have to have everything a hundred percent perfect. I remember early on in my career when, you know, building websites was still new and there were platforms starting to come out into the marketplace to to, to simplify that process. And if you really, if you, if you, if you experienced any of that, and if you still experience a site launch today, it is kind of a messy [00:14:00] process.

From beginning to end, from requirements gathering to, uh, determining what your, what content you need on your site, whether you have existing content or not. Right? And so if you start breaking down the whole process, there's lots of little pitfalls. And a lot of times the marketing teams or whoever the owner of the website is like, we cannot launch until we have. X might be too much. It might be too much. And the nice part about websites, unlike, you know, I grew up in the direct mail industry. I'm like, when we print that piece and it hits the postal stream, there is no turning back.

Alane Boyd: Yeah. Yeah.

Erik Martinez: It had to be as close to perfect as you could get it,

Alane Boyd: Mm-hmm.

Erik Martinez: but a website.

Should it be an evolution? Like, you know, if you're looking at your analytics and saying, Hey, I see this. I see this disconnect between my conversion rate and my traffic.

Alane Boyd: Yeah.

Erik Martinez: What's causing that? [00:15:00] Well, okay, maybe today you have 2% conversion rate. You launch a new site, it goes down to 1%. Okay, what changed?

Alane Boyd: Mm-hmm. Mm-hmm.

Erik Martinez: We can, we can keep evolving the site and get that back up to 3%, you know, no plan's. Perfect. Right, and I think, I think that people have to get out of this perfectionist mindset, which is what I think I'm hearing you say is like, focus on one simple thing. Don't try to solve the whole problem

all at once.

So.

Alane Boyd: build on anything. Just like your, your example about the websites was great because we can build on a basic structure and keep imple and, you know, adding implementation to that.

Erik Martinez: So I'm, I'm curious, you know, as you're doing your work, what I'm seeing in this space is companies that have. Lots of different processes.

Alane Boyd: Mm-hmm.

Erik Martinez: When I talk to their ownership or their leadership [00:16:00] teams, the leadership team is like, yeah, no, we have a solid process and they can give you bullet point A through bullet point B.

And then you talk to their teams. They're like , that's the basic flow, but here's all these nuances.

Alane Boyd: Yeah.

And then when they get the report and there's 450 people that cite about five things that's broken, then they're like, oh, oh, okay, well there's a lot happening here. And then we work with them to really prioritize things.

Erik Martinez: Yeah, I think that's a brilliant strategy because I think, you know, we're all guilty of it, right?

Alane Boyd: Oh yeah. I mean, I get pissed if you challenge me on my belief system.

Erik Martinez: We all have our viewpoints, that we all have our things that we want to focus on .

Alane Boyd: Mm-hmm.

Erik Martinez: And then there's the things that are actually happening in the business. I've seen it with in my own organization where we're saying, Hey, we want you to move this way. And the team's like, no, we can't move that way.

Well, yeah, you can. [00:17:00] Well, no, we really can't because there's these seven things over here that are preventing us from turning in that direction. Oh right. Okay. Let's go start knocking those out. I think sometimes we just try to plow through the problem .

Alane Boyd: I'm big on letting data talk and it can be really hard to get the right data. Why is something not working? I mean, I was looking at some MailChimp stats and at face value we are getting the highest clicks that we've ever gotten. The highest open rates, but the lowest level of conversions.

Okay. So getting that data, what changed from when we were getting higher numbers to now? Well, that's really difficult to dig into. You have hypotheses, you're digging through old campaigns trying to puzzle things out. That's hard data to figure out, but you've got to dig in and get that stuff. Some things are easier.

Well, if open rates and click [00:18:00] rates had dropped, that's a very concrete thing to look at. But if they're the same and you're conversions are lower, then you're looking at not just numbers, you're actually having to dig into content. And so I can see where some people really are challenged at trying to get to the root cause.

But I don't do things on feelings. Like I'm constantly with my exec team. Well, I feel, where are the teams? I feel like this. I feel like this. I'm like, that's bullshit. I don't wanna hear about your feelings. Don't care about your feelings about this. No, I'm like, show me data and we can make decisions, but I don't wanna do it on feelings because feelings are wrong and feelings change.

How I feel one day is different than the next day based on personal factors, based on what's in the media, based on what customer emailed us and now has me like in a different realm. So I don't like making decisions based on feelings, and then that's part of what I wanna work with, with our clients, [00:19:00] is let's look at data to tell us how to make decisions and priorities need to be looked at based on that. So you're not continuing to bandaid.

Automation is not going to solve your problems. You have to dig in and figure out where the system is broken, and then we fix it with automation. But if you don't know where it's broken, you're gonna keep on glossing over and not fixing a situation.

Erik Martinez: I find that this is one place where AI is super helpful.

Alane Boyd: Absolutely.

Erik Martinez: In digging into data and making connections that you wouldn't see. Now, that means you still have to gather the data.

Alane Boyd: Right.

Erik Martinez: You have to gather the data and throw it in. But once you do that, the AI capabilities of tools, just simple tools, I consider it relatively simple.

It's amazing. But like NotebookLM. You throw a whole bunch of documents in there or a whole bunch of conversations, [00:20:00] whatever it is, over period of the month and they say, what have we been talking about? What are we still talking about? What we were talking about a year ago?

Alane Boyd: Yeah, I love that. And it's true. I'm on a committee and we had 250 responses to something and people were skimming and looking for keywords and making anecdotal comments and I'm like, you're picking out the words that feel good to you .

We all have a personal filter. We're not going to look at it with clear lenses. We're going to look through it with our own filter. If I saw Mel Robbins' name four times, I'd be like, Mel Robbins, that's who we need to have. But that's because I like her. Using AI for this analytics piece really helps remove our personal filters and gives us a true analytical summary of the data without anybody's input.

Erik Martinez: I think what's even better about it is that we can pre-script the critical questions so that we're always asking the critical question of our data. One of the challenges we've [00:21:00] had up until this point in time, when you're looking at data and you're doing analysis, I spent 80% of my time organizing the data and I spent just a fraction, maybe 10 or 15% analyzing it. And what if I could flip that script? Or increase the percentage of time that I spend analyzing it and generating insights. With AI, we have the ability now to ask those critical questions consistently and from multiple angles. So we can really get to the root of problem. And it can be semi-automated.

I know you do this work. We've been doing some of these projects too, and one of the questions I ask people and they get a little confused the first time I ask it. I'm like, is that manual, semi- automated or fully automated? I stress to them it is their opinion on it because, I want to know if there's a disconnect between their perception of whether that is automated.

There's no wrong [00:22:00] answer to this. But it is their perception of that task. Whatever they're doing. I feel like it's semi-automated because I use Excel. Or it's semi-automated because I use InDesign.

Okay. But you're still doing all the things. You're still making all the, color corrections or whatever it is. You're still correcting all the formulas. Now if you were writing a macro in Excel and you just had to punch three numbers and it does all the rest of the magic.

Alane Boyd: I have, a episode on my podcast all about moving away from spreadsheets. Cause that is not a way to actually manage your business. But the people that think, I've got a spreadsheet for that, that this is a fully fledged out automated process.

I'm going, oh my gosh. In the 1970s, that was good to have in a spreadsheet, but we need to evolve.

Erik Martinez: I just had one of those cases this morning. Because we're relatively small, I do a fair amount of work. But, somebody embedded a bunch of URLs inside a text. What's the formula to do that? And I [00:23:00] looked it up and then I'm like, here ChatGPT, extract it. Two seconds later done.

And there's lots of little teeny tiny use cases like that. Back to your earlier point, like start with really simple things like how can I simplify time? We've been in the direct mail industry for years and some of my clients are still doing a lot of direct mail and catalogs. I don't know if you know anything about direct mail, but those catalogers go spring, summer, fall, winter collection. Then they take that information from the catalog and they put it on their website. I built a process about a month and a half ago using, ChatGPT to extract it, and I took some screenshots and showed it. Well, I got another catalog recently, so I threw it through that process. It busted it. I'm like, I can take the screenshot, I can just circle the pieces I want at the parse. All of a sudden, bam.

Alane Boyd: Done.

Erik Martinez: So sometimes you have to be in the process and then take yourself out a little bit and go, what am I doing that's making this so hard?

Alane Boyd: Definitely.

Erik Martinez: [00:24:00] Let's, move off of that and go back to the rewards because I think, we talked about lots of different opportunities and little wins that people can get with AI.

I'll give you an example. I was talking to a client a week ago and we were doing one of these, discovery sessions. Really working through, tell me about what your day looks like. What are the tasks you're doing? How often do you do them? Are they automated, semi-automated, fully automated? This was a group of three people. We got through the end, I said, I know you guys could ask these questions internally in your organization.

What do you see when you have that conversation? What are people saying to you as they look at these things and say , I could do it in house. If they do take it in house, do they continue with it?

Alane Boyd: We have some cases like that, and my favorite ones are the ones that have tried it and then realize that it's harder than they imagined. And a lot of times they might have an employee that has a knack [00:25:00] for this. You know, it wasn't their main job, but they really like dabbling in it. And so that employee did it.

That employee is not gonna stay forever. And then trying to hire somebody else, that has that expertise, to do it is going to be much more expensive because you're hiring for that role, which really needs to be some type of lightweight developer. Do you need to be a developer? Not necessarily, but when you start talking about databases, a developer understands that, and that's really important. What happens when I see a company that does it internally, which is great, we wanna help support you. We can do training, we have cohorts available to train, but maintaining it is something that most people don't wanna do. A lot of times it gets built and then errors start rolling in and they don't wanna maintain it. They don't wanna fix it. They don't know how to do the error handling. They don't have time, they don't have the commitment to it.

And so, it just continues running, but every time it does, it just throws an [00:26:00] error. Those are our favorite clients because they've tried it. They know how hard it is to do it on their own, and we'll welcome that. But we also realize that a company doesn't wanna always have to use a third party for stuff like this.

They do wanna have some control over making edits, making changes. Maybe they don't wanna build the entire AI agent workflow, but they'd be fine with doing some little things, changing the prompt, making it, instead of going to Slack, they'd want it to go to Clickup.

So we started incorporating training and having training cohorts available, we call it our AI bundle, we teach three basic parts of building agentic workflows. Maybe they don't learn all the details. We're not teaching them how to code in Python, but we are teaching them the basics they need to build a workflow and maintain a workflow.

Erik Martinez: I think that's great. I think the modern way of working with your agency partner is really working hand in hand like that. Being able to recognize, hey, you know what? Our skillset runs at this point. [00:27:00] We need you guys to help us push it over the finish line or, we can get it over the finish line, but we don't know where to start. Being really honest 'cause doing it in-house sounds wonderful. And look, I sat on the other side of the desk, I wanted to do a lot of things in-house, because it was expensive to go out.

Some of the things you're talking about have real world impact and growth potential for business. As we start to move towards the end of this conversation, can you give some examples of tangible wins? You don't have to give names, but can you tell us a little bit about some of these wins and the magnitude of impact they had on the organizations and over what period of time? Because I think some of it is, I want this now. I'm working with a startup, the conversation is, I wanna see more sales now. You have no fundamentals in place.

What do you have as an example?

Alane Boyd: Yeah. There's so many different types of quick wins. I always say clients fall into two buckets. I want more sales, or I need more [00:28:00] efficient operations. It's very clear which side the customer's on. And so if you're needing more sales, there are ways to automate sales. You can get a campaign going within a week using AI. The content you're doing and your ICP needs to change as you start selling if you are a startup. If you're looking at an AI agent, there's a chat interface you can put on top of an AI agent, and that can have access to a document.

Those can be really quick where you query the AI. You put something in there and it gives you an answer. It can be a really quick way to do it. And then the last thing is, if you're on that operation efficiency thing, standard operating procedures are a big part of what companies need to do. Even if you are a solo founder. To get something out of your head so that an admin in the Philippines can take care of something for you, you have to be able to articulate it.

And so something super simple to do is record a Loom video, have the transcript created, and then run that through [00:29:00] AI and say, create a step-by-step process. And save that in a Google Doc or something where it can be accessed. You can take that same basic process I just told you, and you can turn that into an AI agent workflow so that all you do is paste the transcript in and then it creates the SOP, saves it in the Google Drive folder, it does all the work for you. That way you always have a record of what you do so that somebody can take the work off of you. You get inundated with questions all day when you don't know how to communicate what you know.

Erik Martinez: Guilty as charged. As we move the close, what is one last piece of advice that you would like business leaders or their teams to take away from this conversation?

Alane Boyd: To take away would be to start training your teams on using AI. And if you don't, every person on your team will use AI differently and be putting your data in there, be putting your customers in there. And it's not just a security breach, but it's that other [00:30:00] companies have access to your data. Train your teams on how to write prompts, how to use it securely, and start looking at what an AI agent flow could look like for some of your workflows that are the biggest pain points.

Erik Martinez: I think that's excellent advice. I'm definitely in on the training. As we were talking pre-show, there is very little training going on right now . There's a lot of companies just dumping tech into people's labs and expecting them to figure it out.

Saw an interesting article on Fortune. Fortune seems to have taken the cautionary side of AI, but I think they made a really good point. There's a study done, that showed that 75% of workers and companies that are dumping AI tech into their laps feel like they just doubled their workload. Because in order to learn how to use these tools, you have to learn how to use these tools.

I've learned a lot by just working with the tools. Trying to solve a specific problem, [00:31:00] moving on to the next problem, and so on and so forth. But sometimes it can take hours.

And when you're under deadlines constantly and have other things on your to-do list. That starts to feel stressful.

Train your team, start investing in some automation to free them up. That will allow them to take advantage of that training and improve the outcome for your business.

That's basically the model. So Alane, if somebody wants to reach out, what's the best way to get ahold of you?

Alane Boyd: I'm very active on LinkedIn, so love when people connect with me. Shoot me a message on there. I'm Alane Boyd. My website's BiggestGoal.AI, so you can always reach out on the contact form there and shoot me a note.

Erik Martinez: Just so everybody knows, Alane is A-L-A-N-E.

Alane, thank you so much for coming on and sharing some of your insights with the audience today. I really appreciate it.

Alane Boyd: Thank you, Erik.

Erik Martinez: That's it for today's episode of the Digital Velocity Podcast. Thank you for listening and have a fantastic day.

Narrator: