How I Profitably Use AI in My Consulting Business

How I Profitably Use AI in My Consulting Business

Artificial intelligence is a form of leverage. And if you’re not leveraging systems, media and technology, then your growth will stay linear.

Today we’re going to be talking about how I profitably use AI in my marketing, sales and operations.

You might be trying to figure this out on your own. You might be trying to figure out how to ROI the tech stacks that you’re implementing. How do you even think about the tech stack? And how do you discern what is actually necessary for you to execute versus what is just an AI wrapper?

People are selling and building wrappers on top of OpenAI and selling that as a solution. When in reality, you can build it yourself by going straight to the source.

And I’m getting a lot of that. Even just talking to my growth plan clients. They don’t want complicated. They don’t even have enough time to tinker with all the tools. They just want something simple that works in execution.

So my hope with this is to give you one, the frameworks of how to use artificial intelligence. Number two, where to use artificial intelligence. And three, the workflows that I personally create and use to leverage this technology in my business.

The Framework To Think About Applying AI

I think it’s purely important that we start here. Because if you just go doing the tactics and executing, you’re going to fail to do the thinking.

And when we talk about thinking, I actually need you to know what good looks like.

Whatever you’re optimizing. Whatever you’re building out with artificial intelligence, with technology, with automations, synchronizing all that with your team. You need to know what good looks like.

The only reason some of the things I’m about to show you work. The only reason I can automate and streamline so much is because for each one I’ve put in the time to do it badly and poorly as a human. Executing and learning upon that experience. And then developing something that is good or acceptable to the market. Or acceptable to me, frankly, which is a little bit higher than market experience. But shipping something that is good.

And I know through my experience of testing and iterating what good looks like. Either through strategy or through copywriting or through just knowing how things are supposed to work.

If you jump into the deep end of using the technology not knowing what good looks like, you will produce crap.

And that’s not just an input output thing. It’s an awareness thing. If you’re unaware of what actually good looks like, you will be at risk of developing something dangerous.

I actually did this when I did a private training for a team. Where I showed them specifically the dangers of just not thinking and executing based on what the technology tells you. The large language models.

What I did is they wanted to know how do I improve my workflows or my actual personal work as an individual contributor using AI.

All I did is I went to ChatGPT. I asked that same question. It gave me some answers. I went to Gamma and then I exported that as a slide deck because it auto created a slide deck for me. I presented that slide deck and I showed them real proof.

This is the danger. If you just create with AI and don’t think, it’ll produce standard or substandard information.

Now, some of the information that’s gonna come out in the future will be better. But that’s not the lesson here.

The lesson here is if you’re implementing things for yourself or your own business where you’re putting your own money in. Or for clients that are paying you to do an extraordinary job. You need to know what good looks like. You need to have gone down the path of experience and know what the end outcome is supposed to be like. For you to then work backwards and build the right workflows and the right execution of the right technology for them to get the results.

And the whole purpose of artificial intelligence, AI, technology workflows is for you to enhance. For you to evolve and further deepen your impact through one input by being able to do more with significantly less.

And that’s not a hyperbole. That is just the state of how things are now.

Where To Use Artificial Intelligence

Right now as of this time of recording, this is January 2025. Where do you use this?

There are two major points.

Number one is data. Managing data sets. Researching datasets. Handling large data sets. Being able to discern through large data sets that help you create better decisions from large data sets.

Number two is creation. Accelerating the creation of copy. Image. Now video.

Those are the two that I primarily focus on to create leverage with AI effectively and profitably.

Today there might be more with leveraging agents and building that. I’m not there yet. And I don’t teach what I don’t know.

So here’s what I’m doing now.

The Actual Workflows I Use

Now let’s dive into the actual workflows that I use in my marketing, sales and some of the operations. The technology. The workflows that I’m building out.

I’m going to just tell you exactly what I’m using here.

The first workflow is my podcast.

I film this podcast directly in Descript. Automatically transcribed again using the technology. And then from there I can search unique clips that I want to pull from the podcast to create social posts.

And those social posts go across specifically on my LinkedIn, but they go across the entire social network. Whatever the networks are. I primarily like to focus on LinkedIn.

Then from there, those clips are transcribed and I have a team member leverage a bot. I created a GPT. And all I did on that GPT, so I am paying for it, but what I’m doing specifically is I trained a bot off of 94 different pieces of content that I’ve personally written. 94 inputs for it to know how I talk. How I don’t use sentences correctly. How I sometimes chop up my sentences. How my grammar is off. As well as how I don’t use periods in most of my sentences.

So I trained the bot to feel like me and to have a tonality like me. I had to manually prompt it and manually input it.

However, now all the team does is they take the transcript from the clips. They send it to the bot. Obviously you can automate this with Zapier or Make. But they send it to the bot. The bot reviews it and then outputs a review for ClickUp for me to review.

Then all I do is copy edit and review what I’ve said. Because it’s a voice transcript. And then the bot is trained off of my actual voice. The output is actually pretty good. All I have to do is copy edit so that I can think about my own thinking or enhance my thinking or completely change it. So that everything that I produce is still organic. But I have a head start because I don’t have to start from a blank screen.

Then from there, it goes to my publishing syndication for the social platforms. Newsletter. The whole nine.

The second workflow is leveraging that content into my ad system.

Some of the content that I produce for ads are video content. Some of them are actual text content or some PDFs or infographics. Some of them are actual trainings. Like 30 minute trainings that I create. I don’t automate that.

But the content that I produce from the podcast and the clips can inspire new content for ads. Or I can even use those clips into the ad system for inbound.

And then what I do there specifically fuels my inbound system.

Content and ads are almost like synonymous twins in my opinion. Where you’re creating something of value and you’re offering it to an audience. And that audience reciprocates either by visiting your website or visiting your LinkedIn profile.

Now this is where the inbound system works.

My personal inbound right now is on LinkedIn, website, as well as newsletters. People opt in for the newsletter. People opt in to visit my content via LinkedIn. And LinkedIn tells me who visits. And people go to my website or the trainings to review the stuff that I give away for free.

Now with that, I leverage tools to add workflows to follow up with them.

The reason I create workflows, and these are not just automated bots, is because there was a one time where I was running ads and I got frustrated because I had 97 people to follow up with.

I personally did not have the time or the space to follow up with 97 individuals. Because the followup is not just, hey, how are you doing? Or hey, it’s a simple whatever.

The followup is personable. And then it’s also three to four different touches to make sure that they have value. And that’s through LinkedIn and email and just social selling.

I didn’t have time to do that.

So I created a workflow using La Growth Machine, which I’ll link out below. And when I do that, I’m able to then maximize my input, which is the content and the ads. And then people reciprocate. Either they go to the website or they’re interested or they actually opt in. And then I have systems.

The newsletter is pretty simple. That’s like the oldest system that’s been in existence. They just subscribe and then they get access to the rest of your newsletters. They get hit up in an email every time you send an email.

But the interesting part. I’m using a tool called The Vectors. And that is able to identify who hits on your website. And I’m also filtering who hits very specific pages of the website. So if they watch a training and then they visit back, I get a notification for that. They go to my CRM on ClickUp. I use everything on ClickUp.

And then from there, all I have to do is reach out to that person. Because I’m manually doing the sales myself. If I was having a sales team or had a larger machine, I might use a different tool. A larger tool. Maybe like HubSpot.

Now with that being said, I also have different ways that I’m segmenting people who go to the website. If some people go to the newsletter and the opt in, I sync them with Zapier or Make and I create a Clay table so that I can enrich their data. Because I want to make sure that I’m sending the emails and the messages to the right people. And that I’m not going to get a bounce because I don’t want to risk my domain.

And then from there, I send them to La Growth Machine where I thank people for being in the newsletter.

And I do this because I personally would do this. The copy that I write is pretty simple. I’m not using AI. But I would personally do this with them in the past. Now that I have technology to do that, it can leverage my time so that I can be here filming more content for you versus having to take 90 minutes to two hours to do this every single day.

Let Me Recap Those Workflows

That’s a lot that I just like digested or just sent your way. Hopefully you followed that.

The podcast creates clips. Those clips, I create copy drafts for me to review. Those drafts go to the publishing and syndication on the socials and LinkedIn. Some of those pieces of content I enhance with ads. Some of those create custom content just for ads.

That goes to more content where people go on my website or on LinkedIn. From there, I have followup automated so that I can build a relationship or have an excuse to build a relationship.

Between You and I…

I don’t care about leads.

I had a conversation just now with another agency owner. And he mentioned, hey, leads, people want leads and whatever. My clients want consistent leads as well.

Leads don’t matter. What matters is relationships.

It’s not that hard to get leads. If you want leads, it’s probably the easiest thing in the world to get a lead. Multiple leads every single day.

The hardest thing is conversion. The best way to convert today is a great relationship. Because at the end of the day, you’re working with people you want to work with. And you’re adding asymmetrical value to them.

And that’s really the name of the game. That’s the name of the game of business. Creating asymmetrical value where they feel tremendous value for the dollars that they’re giving you. And you’re finding a way through your crafty ingenuity to maximize the value without maximizing your expenses to create a profit.

That’s essentially the name of business. Asymmetrical relationship building.

Anyways, that aside. Because all this technology might feel like I’m just doing this to blast people. It’s not. It’s for me to have excuses to reach out to people and create a real relationship.

And the call I got was literally from one of these systems. The guy that I was just talking to about this exact same topic was a guy that we built a relationship from these systems. If I didn’t build the system, I probably would never go on a call with him.

That’s just how it is nowadays. Because I don’t have all that time in the world.

The Last Workflow Is Streamlining Operations

Some of the deliverables. I don’t do any new thinking with AI. Just keep that in mind. I do a lot of research using artificial intelligence.

I do tons of automations for onboarding. I have 19 to 21 proprietary questions that clients go through. But in order for them to get access to that. For me to fulfill on that. To get them all the tech access. It’s all automated. 90% of it is automated.

The 5 to 10% that isn’t automated is because I actually ship a box to clients. I manually do that because I like doing that. It’s fun.

So anyways. I do deep research with AI. Right now I’m testing. I use Clay. I use GPT. Perplexity. I am testing deep research from Google. Gemini. Again, these are all different players that we don’t know who the winner is going to be. Just test them all.

But I really do like the research approach.

Warning: Don’t Outsource Your Own Thinking

You’re not outsourcing your thinking. Do not outsource your thinking to the technology.

Outsource the actions that you would normally do. That’s not a task that’s not worthy of your time. That you just don’t have time for. And that you have to discern.

If I’m doing 20 things in a day, what are the three that are going to move the needle for everyone involved? And the rest. Can I delegate to someone else? Can I delegate to a machine? Or can I create an automation?

And the machine part, that’s really where this comes in.

When I work with clients, most of it I can automate so that I can focus on the most important piece. What is that? Creating offers. Offer creation. Followup. Strategy. Nuance. Sales approach. Listening to sales calls. Being able to help them convert. Creating campaigns and iterating campaigns. Meeting with the team to review the data. Even using AI to repeat the data. I don’t care.

The whole goal is like, what are we doing that’s new that we can create?

Only You Can Create NET NEW

We don’t want to use AI for outsourcing your own thinking. I want to think about my own thinking. That’s compound thinking. That is exponential thinking.

When you do that, you create new workflows. New automations. New hires. New technology. Whatever you’re building out. New offers. New lines of business.

When you’re shaping your own point of view and thinking about your own thinking, you have to really sit hard and say, what is different here? How can I improve what exists? How can I create something that doesn’t yet exist?

Now this is a philosophical point. But AI cannot create net new because it doesn’t have the conscious awareness of what exists. It has the data of what is the past. But it’s not all the data because not all data has been transcribed into digital. There’s a lot of data that’s still non digital.

And when you think about that. It can create based on what exists. You have the capability to create what doesn’t yet exist. AI can predict what doesn’t yet exist, but it can’t create net new.

And I’m not going to die on that hill because I don’t want to die on a hill. But that is something that I will argue because it just doesn’t make sense that AI can create net new. It can create new things that look net new to us because sometimes we’re blind. But it can’t create true net new.

Only you can create true net new.

Think about that.

So don’t use AI to outsource your thinking. Don’t use this to make important decisions. Especially gut reaction decisions.

And the third. A lot of people are using it for selling. I don’t believe in that. I still believe that right now, if you look at the technology companies, they’re hiring humans to sell the technology. Let’s just put it that way. They’re hiring humans to sell technology.

Why? Because human to human sales is relationship building. And I’m not sure what the future looks like. But as of today, you’re not going to build a significant relationship with a bot when you’re making significant or important purchases.

And I could be wrong on that one. But that’s just my take for that specifically.

Your Experience Stills Triumph

This is less about how I use AI. Obviously. This is more about you need to know what good looks like.

You need to have the experience. Have gone through the suck. Been punched in the face plenty of times to figure out what works. What doesn’t work. And how do I automate or streamline with technology what works?

Remember, think about the frameworks. Think about that first specifically. What is the workflow you’re trying to do? What does good look like? What does the outcome look like?

Then you identify how am I going to leverage technology for this. Artificial intelligence for data sets or for creation. Not net new creation. But creation of the content or getting you a baseline of creating visual copy, copy, or visual content.

And the third is figuring out what are the workflows that you create that are in alignment with your team, your business and your needs.

Artificial intelligence is a new form of leverage. Some would argue that it’s a form of leverage under the tech and systems leverage piece. I actually would argue that it’s combining both time and labor and technology and systems together to create something forward. Especially now with the rise of agents.

This essay originally appeared in the Do Good Work newsletter. Read on Substack and subscribe →

Frequently Asked Questions

What framework should I use before applying AI to my business?

You need to know what good looks like. During a private training for a team, Raul went to ChatGPT, asked “how do I improve my workflows or my actual personal work as an individual contributor using AI,” exported the answers as a slide deck through Gamma, and presented it to the team as real proof of the danger: if you just create with AI and do not think, it produces standard or substandard information. If you are unaware of what good actually looks like, you are at risk of developing something dangerous.

Where does AI actually create leverage in a consulting business?

Two places. The first is data: managing, researching, and handling large data sets, and being able to discern through them to make better decisions. The second is creation: speeding up the production of copy, images, and now video. There might be more with leveraging agents and building that, but Raul is not there yet and does not teach what he does not know.

How does Raul use AI for his podcast and content workflow?

The podcast is filmed and transcribed automatically, and clips get pulled from it for social posts. Those clip transcripts go to a custom bot trained on ninety four pieces of his own writing, built to match his tone, grammar quirks, and the way he sometimes chops up sentences. The team sends the clip transcript to the bot, the bot reviews it and outputs a review to ClickUp, and all that is left is for him to copy edit and review what he said, before it goes to publishing and syndication.

Can AI create genuinely new ideas for a business?

No. This is a philosophical point: AI cannot create net new because it does not have the conscious awareness of what exists. It has the data of the past, and not even all of that, since a lot of data is still not digitized. It can predict what does not yet exist, and it can create things that look new to people because sometimes people are blind, but it cannot create true net new. Only you can create true net new.

What should you never outsource to AI?

Your own thinking. You are not outsourcing your thinking, and you should not outsource your thinking to the technology. Outsource the actions you would normally do, the tasks that are not worth your time or that you just don’t have time for, and that you have to discern by asking what are the three things, out of everything you are doing in a day, that are going to move the needle. The rest, delegate to someone else, a machine, or an automation.

Building an AI ops layer in your consulting or agency practice?

Book a conversation with Raul: calendly.com/rherocho/performance →