816 | She Built 11 AI Agents and Cut 80% of Her Workload (Here’s How) | Ashley Gross

I was actually pulling in more so 80 hours a work week and especially in small businesses, you don’t have the luxury of time. You don’t really have the luxury of getting it wrong. And so I revamped that. That’s Ashley Gross, the marketing ops leader who’s not just streamlining workflows. She’s architecting AI systems that eliminate the work no one wants to do. From building multi-agent automation stacks to reducing her own workload by 60% without sacrificing quality, Ashley’s turning AI into a daily operating system, not just a distant strategy. Recognized by Forbes as one of the highest paid AI consultants, she’s teaching agencies how to do the same one micro workflow at a time. But how does she do it?

I built 11 agents that now do 80% of my operational muscles. They are doing all of the least glamorous sexy jobs that you can imagine so that I’m not just running around like a chicken with my head cut off as a solo founder. In this episode, you’ll learn how to automate 20 to 25 hours of weekly busy work with simple AI workflows you can set up in one day. All my agents talk to each other. They are very much built into my organization charts, full-time, part-time, maybe some contractor. How to turn AI agents into your first full-time digital employee. As a small business owner, I am losing deals when I’m sleeping. So having an AI agent not only answers those questions but also qualifies those leads and then going so far as to book that on my calendar has been the craziest concept for me.

How small businesses can operate like enterprises using agentic workflows, automations and new AI augmented rules. What used to be a million-dollar advantage is now available for $75 a month. With your business, you know, I read you’ve taken, you know, what used to be tasks that you would do and jobs taking 40 hours a week are down to like 15 hours now. What is the craziest thing that you’ve seen AI automate or perform either in your business or any of your clients? This is not the craziest one but I would say the most frequent one that I have, I started with myself in my own business and now I’m recreating for my clients, is the AI agent that is online all night long because most chatbots are glorified FAQ pages.

You ask a question, you get an answer and then nothing. And so I built an AI agent and I built them now for my clients that actually executes. So not only can I sleep because as a small business owner, I am losing deals when I’m sleeping. So every minute of my time that I’m sitting sleeping, there are deals that I’m losing and that’s just the reality of being a small business owner. And it doesn’t make sense to have someone hang out all night long on the computer and potentially wait for a question or maybe not receive any questions at all. And so having an AI agent that not only answers those questions but also qualifies those leads that are asking those questions to see like, are they just digging around? What does their IP address say? What is their intent in this conversation? And then going so far as to book that on my calendar has been the craziest concept for me because the way that this was designed is all of my SOPs, all my SLAs, pretty much every question that anyone has ever asked me lives inside of my Google Drive and Notion. And so I, every time I get asked a new question, I add to that database. So I already

have a ton of questions, a ton of answers, a ton of resources. And so what I did was I connected that to AI and I created Persona. I made her incredibly friendly and I gave her the rules and parameters that she could only answer questions from my list of questions and she could only answer them in the way that I provided her with documentation for. And so if you go on my website after hours and you ask a question in the chatbot, you knowingly click chat now, you ask your question and then she appears and answers them. And then if you want to have a more fluid conversation, a synthetic human interaction, if you will, you can actually click the dial button and talk to her. So instead of just typing back and forth, which we’ve kind of become accustomed to hate, you can just jump on a video call with her and she’ll answer your questions live.

And as she’s doing this, she’s asking you questions such as, hey, I see that you’re interested in consulting courses and you had a question about how this works with Ashley’s workshops. Who is your audience that you would like this workshop to be with? And as they’re answering her in real time, she’s updating my CRM, so she’s creating a contact updating the CRM, polling that information and enriching it and looking at signals such as when is their fiscal year, if they’re public, look through all their social media platforms that are public and tell me what their interpretation of AI is and what their sentiment is and then show me three posts that they’ve posted or reposted that link to that sentiment. So she’s not only answering their questions in real time, but she’s gathering intent to see, are they just asking, is this the middle of the night crusade where they’re just on the internet or are they actually interested in something? And this happens to be the one moment as a business owner, as a CMO, as a CFO that they get to do some research and right now we need to answer them and get them on

the calendar. And if so, she offers them to book on my Calendly and not only does she say, you know, hey, Ashley’s got availability tomorrow at 2pm, would you like to book, she’ll book it for them. And that’s a much cleaner user experience than just giving a Calendly booking because if we got them in the middle of the night and they actually are one of our ideal customers, then we don’t want any time to go by and we don’t want to make their journey of getting to us any more complicated. So we don’t want to give them a link and make them click and fill out a form. We just want them to say yes. And so she’s essentially doing what a sales rep would do in the middle of the night and giving them my next availability and almost kind of saying, this is her next availability. So I hope it works for you, does it? If not, I can look at the next one, but she books that pretty quick.

Now we’re creating this FOMO, this urgency, after hours. And if they say yes, she books them. The enriched contact information, including the call transcript of that interaction is in my CRM and I have an email. So when I wake up, it’s like, I can just debrief myself on what happened and look forward to that next call. If you’re an agency that serves local businesses, this is an easy service to add. Let me show you how agencies set up an AI receptionist for clients. Everything starts in one place. You drop in the details the receptionist needs to know, services, pricing, availability, qualifications, and the questions customers ask every day.

Next, you define how that receptionist should behave on every call. You control the greeting, the questions it asks, how it qualifies leads, and how it closes. So it follows your client’s exact process. It can even book appointments directly into the client’s calendar. If a call needs a human, you can set up transfers to the right phone lines, sales, support, or the client’s front desk. Then you connect it to the tools you already managed for your clients, your CRM, scheduling links, lead forms. So every call becomes a complete workflow, not just a conversation. And because it’s fully connected, the AI can automatically follow up by text. So no leads fall through the cracks. This is how agencies deliver AI call handling at scale, whether it’s for one client or hundreds.

Build an AI receptionist for your clients with Vendasta. To look under the hood with some of the other parts of your enterprise, where do you have other little gremlins working and what types of jobs are they doing? They are very much built into my organization chart. So whenever you’re talking about pre-AI times, if you will, if we can even remember what that was like, the way you structure your employees is full-time, part-time, maybe some contractors, maybe some freelance workers. That’s no different than how you structure AI agents. Your ability as a leader to decide how many hours and what type of employee should be in which role to enhance their own abilities, that is a leadership skill set.

So I’m using AI agents to do everything that I don’t want my humans focusing on. So they’re doing things like going through every single contract that we’ve ever had, making sure that if there’s old branding or old addresses, phone numbers, that we have an audit running so that we know when to update our NDAs and our contracts. They’re doing things like cleaning up our CRM, making sure that there’s not duplicates. Anytime that we are getting ready to have an upsell with a client, like anytime there’s a contract that’s about to be up for renewal, there’s an agent specifically that the whole entire year leading up to that renewal, every email, every call transcript, every contract, every agreement, it’s got like a relationship database on every single client we have down to a person at that organization so that we know like six months before the renewal, what our talking points are, how to actually get that renewal, if we can upsell, what does that look like, how do we talk to these folks. And then also not just like how to do it, but what we’ve done in the past. And so my favorite thing about this agent

is it will cite all of the different calls where if it’s the CFO who’s the main stakeholder, it will pull the last year’s worth of calls where he had a really good experience with us, or he had really great sentiments, or he was paying us a compliment about something that we did. So that not only are we getting our own answer and we’re proactively having these conversations and managing relationships better, but we don’t have to be as reactive, right? Like there’s never a surprise that happens to this point in time where someone’s not happy, right? Because like we have a happiness meter. And that’s one of the AI agent’s jobs is to have a happiness meter and to keep a check on that, because I don’t want my humans to be stressing out about every email and reading in between the lines and saying, well, George, the CFO usually sends me these nice Friday emails, but today his was short. Does he hate us? Is he going to cancel our contract?

Like that’s just not a good productive use of my team’s time. And a lot of the times like we’re emotional. And so we’re not going to come at this logistically and factually. And so I would rather not be spending my employee’s time there. I would rather be spending their time at in-person events, nurturing those relationships and taking George golfing because that’s his favorite hobby. And we got to learn that about him. This agent is the interconnected agent that has access to every piece of my tech stack, but also it has access to every email I have. And so it can look at where I’m going to be a year before. And it can look at my schedule, what I’ve got going on, my different opportunities that aren’t necessarily about AI workforce Academy or AI workforce Alliance. And it can take all of that and it can form a comprehensive plan so that if Ashley, as herself, is going to be speaking at this conference in Vegas in February, then what are the opportunities for the academy?

What are the opportunities for the Alliance? Who are people that Ashley said she wanted to network with before? Or who are clients that we currently have that are out there that I could potentially have a visit with and check up on? So it really acts as my second brain so that I’m not just running around like a chicken with my head cut off as a solo founder. It’s really able to try to find the strategy that I talk about and add to all the time, but optimize my productivity even more because there can always be more. I saw here that you’ve got a CFO who’s an AI. And so I wanted to ask you, number one, how did they compare? How does it compare to flesh and blood CFOs?

And did you give it a happy personality? I made the CFO way more aggressive than I usually would because of that notion that most folks come into, whether it’s like one of my employees or somebody else at the organization that I’m trying to win over saying, oh, well, our CFO is tough. Good luck with him. Most of the time, it’s just this mismatch of understanding the root problem. And so I made this persona really aggressive in the fact that the three focuses I wanted to enhance by creating this persona were the pitch, the actual offering, and the closing hook. Because when we’re talking about small businesses, especially, those are often the three areas that they get wrong, one of those three, and with CFOs, you have less time to make your point. You have less time to experiment as a small business and get it right. And so the idea to experiment synthetically and get grilled and annihilated safely from the comfort of your own home so that you can take that as a learning experiment and not be doing that face to face with your real customer and prospect.

That was what I wanted to give my employees and myself is I’m going to give you the worst CFO that you could possibly be in front of. And by the way, they’re fake. We’re not watching you talk to them. This is for your benefit and your benefit only. Go wild. Try your lines. And on the back end of this, the coding is if the representative can give the offering, the pitch, and the close in terms of the least emotional, the least products and features, and the most business revenue focused, then they get nice answers. The more they explain the products and features, the more they try to push a pain point that they think is a generic CFO pain point, the more aggressive the CFO gets. Jobs taking 40 hours a week are down to like 15 hours now. What were the key tasks and processes that you started at in your automation?

I was actually pulling in more so 80 hours a work week because I just didn’t have enough time in that 40 hours to get done the work that I felt like needed to be done. So the five tasks that I immediately started to look at and optimize were content repurposing.

It was email, inbox management, researching and prep, client onboarding and delivery, and then lead qualification and the pipeline management that comes with that. In those five tasks, let’s just look at them one by one. For content repurposing, I was spending hours turning every piece of content into multiple formats. I revamped that. My entire content marketing operation is Notion, Claude, Descript, Canva, Buffer, and Opus Clip. That’s it. Your content creation marketing stack costs you about $75 a month. Yes. Notion is at the center of everything. I don’t mean in terms of task management, I mean every piece of content I have ever written lives in Notion, every LinkedIn post, every newsletter, every talk transcript, every client story. But more importantly, also in Notion is my brand identity, my archetypes, my content pillars, my voice guidelines, the words I use, what I refuse to use, my origin story. And so when I go to create content, I’m not starting from zero and neither is Claude. So if I have everything stored in Notion and organized, then when I’m using Claude in my content marketing stack,

and I hope that folks copy this because they should, when I use Claude, I’m integrating Claude to research what’s being discussed right now. What is trending? What are the questions people are asking? What debates are happening right now? That research is getting pulled directly into Notion and then it’s getting tagged by my content pillars so that it’s relevant. And then when I draft, Claude doesn’t get a prompt. It gets my brand document, my pillars, all of my information. And when I draft, it’s getting drafted the same way that it started. It’s getting drafted into one long form video content that I can then tweak and iterate and hate and revise as much as I want. And then it gets reformatted into the different posts and the different formats and the different channels. And I’ll use Descript to help me do that in terms of editing. And I’ll use Opus Clip in terms of getting the content in the right formats at the right times with the captions. And then Buffer, of course, to schedule it all because I can’t be in all the places at once. So that was just one task. And honestly, out of all five of them, I think selfishly my favorite out of all five that really,

I think showed me the optimization in ways I wasn’t used to seeing before was the lead qualification and pipeline management because I don’t think anyone has a perfect way of doing this at this point. And it is very manual as far as scoring and enriching your inbound leads and then creating the nurture sequences until they’re sales ready. And what is sales ready? Is it an MQL? What’s an MQL? If this is an MQL, when is it SQL? So like all these little nuances that do exist in every business but bubble up to even bigger visible opportunities for executives, that was really where I spent the majority of the time out of all five of those tasks because I decided what and who I actually wanted to work with based on what I was seeing in the CRM.

And then I just gave the AI that I was working with at the time the qualification and the criteria because as a marketer, it’s half formula and it’s half experimentation. And especially as small businesses, you don’t have the luxury of time. And so when it comes to lead qualification, you don’t really have the luxury of getting it wrong. Enterprises can get that wrong for a couple months. And then at some point that will catch up, right? Small businesses, they don’t have that time. And so being able to look at your CRM and see what has worked in the past, what hasn’t worked, creating the formulas, and then just using your gut when you’re actually trying to refine your strategy for leads and qualifying them, or perhaps change your nurture sequences. That’s where I decided to spend the human time. And that was ultimately out of all five of those tasks, the one that saved me the most time automating.

Email and inbox management. So just drafting responses to routine questions from multiple people. I really enjoyed automating that because I kept the relationship critical emails, anything involving negotiation or conflict or nuance. But then I automated the part of opening my email inbox and it already being in priority levels of what is most urgent. And it already had drafted responses for the ones that weren’t critical. So that was the third task that I automated. And then research and prep was the remaining task. So for every client call, speaking engagement, I was manually researching the company and attendees and industry context. And now and then I automated compiling all the documentation, all the research, the LinkedIn updates and posts, even like comments on YouTube channels that folks were actually adding. Like I wanted to know everything and I wanted it to be in one nice briefing documents that I walked into every conversation informed without doing all that upfront like work. So I just automated all of that, except for the part of deciding like what to do with the information, right? And how I was going to use it.

You work with a lot of clients from different sizes, medium-sized enterprises, larger and some smaller. So especially for smaller businesses, what have you seen them spending the most time on in terms of repetitive tasks? And if you had some advice for them, like where’s the first place that you would have them start if you’re looking at sort of the broader organization? So my short answer to this is any task that your team hates because if they hate doing it, they’re not going to pour their heart and soul into it. If it’s just something that they dread doing. For me, what it comes down to, and when I have to name those tasks, it’s back and forth emails that are a necessity to the business.

It’s things like social media posting, which can become very tedious. And also, I don’t know anyone who actually enjoys posting on social media, but you have to out of necessity, data entry. And so those three areas, I would say, would be the biggest call-outs in terms of where small businesses can be saving their money. So looking at how you can automate those portions of it to give your humans back more time to spend that time doing tasks that are connected to revenue opportunities for the business, but also fill their cup so that they’re putting more energy into them. That’s what I would say you should start with. 2026 is going to be messy. There’s still going to be a lot of information out there. Some of it’s going to be valid. Some of it’s going to be misinformation. You can’t really put too much weight on what the news is saying you should do with AI or what your favorite company is doing with AI. Because unfortunately, we can’t actually look in their books and see what’s happening. We can’t just see that they’re generating $10 billion and be like, yeah, that sounds great. We’re going to

go do it now, but we have no blueprint for how you did it. And we’re just going to take your word for it. That’s not a thing. I would say pause before you automate anything, track every task that you’re doing for two to three weeks, and then ask yourself these questions. Does this require my judgment, relationships or creativity? Does this follow a predictable pattern? And then what’s the cost of getting it wrong? The first two questions, if it’s a yes, then it’s a good candidate for automation. The third question is really just a benchmark for how much human oversight is needed. The first two questions are really your guiding principles and are these tasks things that actually are good candidates for automation? And then if they are great, but if you get these wrong, what is that going to do for you in your career? What’s that going to do for your business?

So kind of look at it as like an autonomy spectrum in terms of find your good use cases for AI, but then put them in a certain order that makes you motivated every single time that you knock one of those use cases out and you’ve got really great metrics. Don’t just jump all in all at once, trying to solve everything because you’re not going to be able to, and you are going to fail. So the idea isn’t, oh my gosh, I don’t want to use AI agents because what if they fail, they’re going to fail. They are. The idea is when you have those good use cases, put them in an order so that you can take the least scary, least risky use case and solve it and get a little bit of a habit, a little bit of a muscle memory built up and then continue to build up your use cases in terms of more risk, but more ROI as you go. So it’s not an all or one, it’s very much a spectrum. I think that a lot of the businesses who are smart are looking at AI and automations and even generative AI as foundational steps. And so they’re looking at their tech stack and their tools that they’ve been using for years and they’re learning how to play with the AI components of those tools. And then they’re going from there

to actually solo experimenting with automation tools and being able to derive real value and productivity gains for themselves. I think that that’s the expectation that we should set for 2026 because you can’t skip those two foundational elements and go right into building agents. When you’re building automations and agents, it’s the exact same thing. The only difference is agents have more autonomy. So there’s way more risk involved than automations. But if you don’t have a documentation of how you’re working and what you’re doing and what the strategy is, and it’s written down and your humans are already normalizing it and using that standardized approach, you’re not going to be able to add automations and agents to it.

So I would say going back to the basics, making sure that we’re challenging what exists and if it’s actually working for the businesses and then figuring out a way to do that faster, do it better, scale it out more, and then automate the parts of it after we do that work up front. That’s where the businesses that are smart are going in 2026.