How Real Business Owners Are Using AI: 7 Lessons From the Trenches
There is a big difference between talking about AI and actually running a business with it.
That was the idea behind our latest AI for Business panel at Frequency Coworking in Rancho Cordova. I didn't want three AI consultants on stage telling business owners what they could someday do with artificial intelligence.
I wanted people who were actually using it.
Our panel brought together three business owners with very different backgrounds:
Michelle De La Campa Martinez, owner of The Ludlow Cleaning Company, a local residential and commercial cleaning business.
Jenifer Namestka, founder of Avo Solutions, a consulting company focused on organizational change and data-driven decision-making.
Dr. Jordan Plews, scientist, entrepreneur and consultant with a background in biotechnology, regenerative medicine and healthcare.
The monthly AI for Business series is supported by the City of Rancho Cordova and Frequency Coworking. The goal is pretty simple: skip the AI hype and show local businesses how people are actually putting these tools to work.
The conversation went everywhere from ChatGPT and Claude to automated scheduling, proposals, meeting notes, data analysis and even AI-assisted drug discovery.
But I walked away with seven lessons that I think matter to almost any business owner.
1. Don't Start With AI. Start With the Work You Hate Doing.
One of the biggest mistakes I see is people asking, "How can I use AI in my business?"
That's probably the wrong question.
A better question is: What am I doing over and over again that takes too much of my time?
Michelle gave a great example.
Running a cleaning company means constantly juggling schedules. Someone calls out. Someone goes on vacation. A new job gets added. Suddenly you're texting five employees trying to figure out who can work Wednesday morning.
Michelle described the old process: text employees individually, wait for responses, update the schedule, then notify everyone.
She is now building a system using Claude Code that connects scheduling information with automated text messages. Employees can receive a message about a shift and respond with simple options such as yes, no, need more time or have a question.
That's not a flashy AI demo.
It's better.
It's a real business problem that consumes time every week.
Make a list of the repetitive things you did during the last seven days. Don't ask which AI tool you should buy yet. Find the pain first.
2. AI Is Becoming a Pretty Good Operations Assistant
Michelle's use of AI goes well beyond writing emails.
She records prospect conversations, turns them into transcripts and feeds that information into her AI workflow. From there, she can extract customer requirements, think through pricing, prepare for appointments and create proposals.
She also uses business data to examine productivity and job pricing.
In other words, AI isn't sitting off to the side as a writing tool. It's becoming part of the operating system of the company.
That distinction matters.
Most people first experience generative AI by typing something like, "Write me a social media post."
Useful? Sure.
Transformative? Probably not.
The bigger opportunity comes when AI starts touching the workflows that run the company: leads, proposals, scheduling, customer communication, reporting and financial analysis.
That's where the hours start adding up.
3. Your Company's Existing Knowledge May Be More Valuable Than the AI
Jenifer described using AI in a way that I think a lot of established businesses should pay attention to.
Avo Solutions has been around for roughly 15 years. That means the company has years of proposals, methodologies, client work, data and institutional knowledge.
Instead of asking AI to invent something from scratch, Jenifer can give it previous bid responses and compare them with a new opportunity.
The AI can help identify alignment and structure a response based on work the company has already done.
As she explained during the panel, the AI isn't really replacing her writing. It's helping organize and reuse what the company already knows.
That's a much more interesting use case.
If you've been in business for ten years, you may be sitting on thousands of emails, proposals, reports, customer questions, presentations, SOPs and project documents.
That's your company's memory.
AI can make that memory searchable and useful.
4. AI Can Scale Work Without Scaling Headcount at the Same Rate
Jenifer also described a reporting workflow that illustrates where automation starts affecting margins.
Her company collects data for clients and may need to produce reports sliced by department, geography or other variables.
Historically, producing 20 versions of a report could mean significantly more staff time than producing one.
Using automation tools such as Make and Carbone, the underlying data can be processed repeatedly to generate different reports without manually rebuilding each one.
Her business point was simple: the number of outputs can increase without labor costs increasing at the same rate.
That's where AI and automation become much more interesting to a business owner.
Saving five minutes writing an email is nice.
Changing the economics of delivering your service is something else entirely.
If your business sells expertise, reports, analysis, proposals or other knowledge work, ask yourself:
Which parts of delivering one more unit of our service could eventually cost almost nothing?
That's worth thinking about.
5. Keep Humans Where Judgment and Emotion Matter
Jordan offered one of my favorite frameworks of the evening.
Think about work in terms of IQ versus EQ.
If a task is primarily logic, analysis or information processing, AI may be able to handle a substantial part of it.
If it requires emotional intelligence, human judgment or a sensitive conversation, keep a person involved.
He also recommended mapping a process before automating it and initially putting human approval points into the workflow. As confidence grows, you can decide whether some of those checkpoints are still necessary.
That's a lot smarter than turning an AI agent loose on your company and hoping for the best.
Think of AI like a new employee.
On day one, you probably don't give the new hire the company credit card, passwords to everything and permission to email every customer.
You give them a task.
You check the work.
You correct mistakes.
Then you gradually expand responsibility.
AI should earn trust the same way.
6. AI Will Make Mistakes. Build for That.
Michelle learned this one the fun way.
While developing her employee scheduling system, something went wrong and the system began sending an employee notifications for recurring shifts throughout the year.
Suddenly the texts started flying.
Fortunately, Michelle had built a safeguard into the system that notified her whenever an employee message was sent. She saw what was happening and shut it down.
That's actually a pretty useful lesson.
AI automation isn't magic. Sometimes it behaves more like an intern who drank six Red Bulls.
You need guardrails.
Jenifer has another simple rule at Avo Solutions: AI-assisted work doesn't leave the company without human review.
Especially when you're starting, keep approval steps around anything involving:
Customers
Money
Legal or compliance issues
Public communications
Sensitive information
Actions that are difficult to reverse
Automate aggressively where mistakes are cheap. Be much more careful where mistakes are expensive.
7. You Don't Need to Become an AI Expert. You Need to Start Playing With It.
One theme kept coming up throughout the discussion: experimentation.
Jordan talked about starting with ChatGPT as what he jokingly called "super Googling," then moving into image tools and eventually more agentic AI systems.
Michelle isn't a software developer, yet she's building systems that connect AI, scheduling, QuickBooks, SMS and other tools.
Jenifer's advice was even more practical: put the AI on your phone and talk to it. Use voice. Give it something you're working on. Argue with it. Correct it.
Near the end of the event, the panel talked about learning through "quick hits" rather than believing you need 20 hours of training before doing anything. Find one useful idea, try it and learn from what happens.
I think that's good advice.
Three years ago, learning AI mostly meant learning how to write prompts.
Today, AI can research, analyze documents, work with files, write software, create presentations, interact with other applications and increasingly execute multi-step workflows.
You won't learn that by watching another webinar about AI.
You learn it by using it.
The Bigger Lesson: AI Is Moving From Tool to Coworker
The part of this discussion that struck me wasn't any particular AI product.
Products will change.
ChatGPT gets better. Claude gets better. Gemini gets better. Some startup we've never heard of will probably release something next Tuesday that makes everybody rethink their stack again.
The more important shift is how people are using AI.
We're moving from:
"Write this for me."
to:
"Help me think through this."
and increasingly:
"Handle this process for me, but bring me in when you need a decision."
Jordan described the ultimate goal as digitally cloning portions of his work: filtering emails, managing tasks and handling repetitive activities so his attention can go toward things that actually require him.
That doesn't mean eliminating humans.
It means asking a better question about every process in your company:
Where does a human actually add value?
Everything else becomes a candidate for automation.
What I Recommend Business Owners Do This Week
Don't try to automate your entire company.
Pick one annoying process.
Maybe it's:
Following up with leads
Preparing proposals
Taking meeting notes
Scheduling employees
Answering repetitive customer questions
Creating a weekly report
Sorting email
Reviewing business metrics
Write down how you currently do it.
Then ask your AI tool:
"Here is my current process. Help me identify which steps could be automated, which steps AI could assist with, and which steps should remain human."
You'll probably discover something useful.
Then build the smallest version possible.
Test it.
Break it.
Fix it.
And repeat.
That's basically how every good business system gets built anyway. AI just lets us do it a whole lot faster.
The AI for Business Panel Discussion: How Real Business Owners Use AI was held September 9 at Frequency Coworking in Rancho Cordova and featured Michelle De La Campa Martinez, Jenifer Namestka and Dr. Jordan Plews. The monthly AI for Business series is supported by the City of Rancho Cordova and Frequency Coworking.
