Revolutionizing Small Business: Understanding the AI Employee (GoHighLevel)
Today I want to write about AI agents and give an overview with a practical use case anybody can do. For a lot of us, the concept of implementing AI into our business practically seems like a distant and overly complex technology to setup.
I try to keep things as practical as possible with my writings. So today I want to bring up something that everybody can do.
One of the most practical use cases for AI is with AI agents (or AI employees, as I also see them called). There are a ton of really great platforms out there. Personally, I'll endorse GoHighLevel, Lindy, and N8N. These three I feel confident you can get started with quickly and easily to solve real world business problems.
FYI, I'm going to focus on GoHighLevel just because that's the platform I'm using with most clients right now to build out these use cases. But you can seriously follow along and do this on any platform.
By the way, I see "AI Agent" and "AI employee" used interchangeably on the internet. They really mean the same thing to me.
Q: What exactly is the AI Employee in GoHighLevel?
A: The AI Employee is basically a bunch of AI tools designed to help businesses automate and streamline operations. It includes features like Voice AI (AI Agents), Conversation AI, Reviews AI, Funnel & Website AI, Content AI, and Workflow AI Assistant.
These tools connect with email, social media accounts, your website, SMS, WhatsApp, etc… Just about anything you can think of.
Its purpose is to automate and streamline operations, enhance communications, and automate away a lot of the routine interactions that our staff are doing.
Q: Why Do I need to Know this Technology Exists?
A: Here are a few reasons off the top of my head…
Extremely Practical: Yes, practical use cases exist. AI agents can fix some of the most common low level business problems that we all have. Here are a couple that come to mind:
Speed to lead
Following up & nurturing leads indefinitely
Missed call text backs
Responding to chats
Asking for online reviews (Google, etc.)
Creating ads
SEO content
etc…
You can automate so many routine interactions that are taking over your staffs time…
It's Constantly Improving: These tools are the worst they will ever be. I hope you can feel that statement. The thing with tech is that it only goes up from here. Companies are always learning and making things better based on feedback. AI is being sprinkled into almost every single platform I know of, and updates are happening basically on a weekly basis.
It's Incredibly Easy to Implement: This I cannot reiterate enough. Pick one problem you want to solve and just focus on that one problem. The setup process for AI features is ridiculously simple. Everything I'm talking about is something I can have my team set up for you in literally a day.
Automation & Efficiency: Whether you implement it today or next year, AI agents are here to stay. They will reduce manual workloads, improve your conversion rates, and keep humans in the business focused on the higher-level tasks that literally require a human. Like I said before, so many of these routine interactions that take over our day can be automated.
Cost-Effective: The pricing on these tools ranges from just a couple of cents to around $150-$250 per month depending on your use case and pricing plan. Seriously, it's low stakes to test these things out. Plus, you might just discover that you don't need another full-time staff member. The savings alone outweigh the costs.
24/7 Availability & Consistency: And then you have the obvious use case that AI never sleeps, never gets sick, never needs vacation. So it's a wonderful backup for when your human staff aren't available or there's call overflow. It's always on, never off, and never forgets. It's a great option for leads if they choose to use it.
Bottom line is that they are really effective at solving common business problems. Missed calls, unresponded chats, lack of reviews, and inefficient lead follow-up.
For instance, a voice AI agent can answer calls after hours to give your team a chance at capturing critical lead information that might otherwise be lost.
There's truly no reason not to dive into this. The benefits are substantial and the barrier to entry is remarkably low.
Q: Where do companies go wrong with AI?
Love this question. A few areas:
Believing They Can Remove the Human Element: Let's be real. You need humans. This technology is still in its infancy. Every customer and every lead should always have the option to speak to a real human. There are many times where the AI will struggle with accents or typos, or not be able to answer a question. Sometimes you really just need a human, and that's a fact of life given where we are.
Not Being Upfront: I'm a firm believer that if you're going to use these tools, you should outline upfront that they're speaking to an AI that's pre-trained in all the company's services and internal documentation. ALWAYS give your leads and customers the option to speak with a human. There's a time and place for AI. For example, if somebody calls in, they obviously want to speak to a human… Hence why they called in. Otherwise they would have chatted with your bot online or via text.
Overwhelm: It's far too easy to get overwhelmed with what to do. Literally just focus on one use case inside of a real business problem. That's it. I always recommend starting off with speed to lead or lead nurturing. That's it.
Q: Can an AI Employee Learn and Adapt?
A: Yes, AI agents are designed to learn and improve continuously. First off, they're connected to a knowledge base. It's on you to provide the AI with all relevant information about your business, such as hours of operation, services offered, FAQs, and common concerns. If you have a public URL with all this info, it's as simple as telling the AI to look at those websites for information. So my point is that yes, the AI will only get better as your SOPs & documentation improve. They'll learn quicker than any human.
Plus, as you use the AI Employees, you will provide them with feedback (thumbs up/down for responses) to train them. This allows them to continuously evolve and deliver better results over time based on real interactions.
Where I see a lot of our VAs' roles going in the future is to essentially act as the "manager of the AI agents," overseeing the AI's performance and providing necessary tweaks.
Q: What Can an AI Employee Actually Do for My Business? (Practical Example)
A: Let me go over a simple use case of setting up conversation AI to reactivate old/new leads.
Step 1: Identify and Segment Leads:
Everybody has access to lead data. You can connect to tools like Apollo.io, Lusha, and so many more for scraping lead data. If you're in real estate, there are tons of skip-tracing services to find residential leads. You can even have a VA do this for you. And for those of you with a ton of existing leads, you can take that data as well and segment it.
Bottom line is that you need the data. Once you have the data, you can follow up and nurture them forever.
Step 2: Build an Outreach Messaging Strategy:
AI agent tools can communicate across different channels like SMS, WhatsApp, Instagram, Facebook Messenger, Email, and website chat widgets.
Super versatile. Personally, I've found SMS the best for reactivating old leads.
Step 3: Implement Workflow Automation:
In any AI agent platform, you can set up workflows that will automatically send messages to your leads.
For example, when a lead is created in your database, an AI agent can be triggered to send an SMS. This SMS conversation can then be managed by your AI agent in perpetuity.
The system can send an initial message, wait for a reply, and then send different follow-up messages based on whether the reply is positive or negative, or if there's no reply at all.
Step 4: Train the Conversation AI Bot for Reactivation:
Four critical things here:
The Prompt: The prompt areas (personality, intent, additional information) are important for defining how the bot should converse. For those of us who aren't experts at prompt engineering, you can always use other AI tools like Claude or ChatGPT to refine your prompts. Ask them to create a better prompt. It's genius.
Knowledge Base: The bot needs a knowledge base to handle common questions or concerns that might arise from old leads. This can be built by inputting Q&As directly or using the web crawler to scrape your business website, Notion docs, or anything else that has a public URL for information.
Guardrails: Always set guardrails for these tools, such as maximum messages a bot can send to prevent "tire kickers" from wasting resources. Remember, with AI agents, you're mostly charged for usage. So you need to protect yourself.
Define the AI Agent's Goals: You need to choose the primary goal, which can be to "get the customers' information" or "have a customer fill out a calendar link." Ask it to collect primary address, first name, last name, phone number, email. This is how we'll know when to stop. By completing this part, you've effectively replaced a call center and admins in about five minutes.
Step 5: Monitor and Optimize:
These tools allow you to test your bot by conversing with it and proactively training it by giving thumbs up/down feedback or adjusting its prompt and knowledge base.
You can monitor real conversations and messages sent by these bots. If you are not comfortable letting it go rogue, they all have safety options where you can make sure the bot submits messaging for approval prior to it sending anything.
Overview
I hope this post is helpful and thought-provoking at the very least.
AI agents are far less complex and scary than most people make them out to be.
Literally just focus on one single business problem. ONE use case. I guarantee you that you can get something up and running in a day.