At a recent Startup Folsom AI panel, moderator Rich Foreman opened with a simple observation: last summer, people shared their favorite AI tools—and since then, several of those tools had become part of their regular workflow.
That is probably the most honest way to talk about AI right now.
Not as a magic button.
Not as a replacement for strategy.
But as a fast-moving set of tools that founders are testing, breaking, rebuilding, and slowly turning into real business systems.
Rich moderated a practical conversation with Dr. Scott Campit, Nicholas Haystings, and Reneta Jenik on how they are actually using AI in their companies. The discussion covered everything from lead generation and market research to platform rebuilding, content creation, customer support, privacy, and the limits of automation.
The big takeaway: AI can make a small team more capable, but it does not replace founder judgment.
That was the theme that kept coming up.
1. Start with the Business Problem, Not the Tool
Nicholas Haystings made one of the clearest points of the evening: if you do not know where you are going, no AI tool is going to save you.
He described using Apollo for customer discovery and lead generation. Apollo can help identify prospects based on an ideal customer profile, or ICP, which simply means the type of customer most likely to buy from you.
For example, Square Root Academy might look for K–12 education leaders in a specific region with a certain budget and staff size. Apollo can help find those people and surface buying signals, such as recent interest in STEM education.
But Nicholas was careful to point out that the tool comes after the strategy.
Before building a workflow, founders should ask:
Who exactly are we trying to reach?
What problem are we solving for them?
What does the sales process look like?
What should a human still handle?
What can safely be automated?
My suggestion is this: before buying another AI subscription, write down the process you want to improve.
If you cannot explain the workflow on a whiteboard, you probably should not automate it yet.
2. Use AI to Compress Work, Not Eliminate Thinking
Dr. Scott Campit shared how AI can dramatically reduce the time spent on repetitive work.
One example was using Claude and MCP servers. MCP stands for Model Context Protocol. In plain English, it allows an AI model to connect with other software tools and data sources, such as HubSpot, Apollo, or other business systems.
That means a founder could potentially ask Claude to review contacts, classify prospects, update a CRM, or help organize customer records.
Scott gave a practical example: a task that might have taken four hours manually could potentially be completed in a few minutes with the right AI-connected workflow.
That is the real promise.
Not “AI runs the company.”
More like: AI handles the boring handoffs so humans can focus on judgment, relationships, and strategy.
Examples include:
Moving qualified leads into a CRM
Summarizing research
Drafting outreach emails
Organizing SOPs
Turning meeting notes into next steps
Classifying customers by segment
The recommendation: use AI first on the work that is repetitive, rules-based, and time-consuming.
Do not start with your most sensitive or most strategic decision.
Start with the thing your team already hates doing every Tuesday.
3. AI Can Help Founders Survive the Messy Middle
Reneta Jenik shared one of the most founder-real stories of the session.
Her company, Foodom, connects families with chefs who prepare healthy meals at home. The company grew into healthcare, including a contract with Anthem Blue Cross to provide chef services for medical patients.
That opportunity was huge.
It was also expensive.
Reneta explained that working with a health plan created major operational and cash-flow pressure. Payments were slow, chefs needed to be paid earlier, and the infrastructure required to support the contract was intense.
The company hit a near-death moment.
She had to cut expenses, let go of the team, and rethink the business.
AI became part of the turnaround.
Foodom began rebuilding its consumer platform using AI coding tools, while also planning AI-powered workflows for chef recruitment, operations, customer recommendations, recipe development, and marketing.
That is a founder lesson worth underlining.
AI did not remove the hard decisions.
It did not magically fix cash flow.
It did not make healthcare contracts easy.
But it helped a leaner team rebuild faster and focus on the next bottleneck.
Reneta described the founder mindset well: what is going to kill the business tomorrow?
For Foodom, the answer was first the platform. Then chef recruitment. Then marketing.
That is a practical AI roadmap.
Not everything at once.
One bottleneck at a time.
4. AI Is Becoming Part of Customer Acquisition
The panel spent a lot of time on customer acquisition because, as one audience member pointed out, a business that cannot get customers is just an idea sitting on the shelf.
For B2B companies, the panel discussed tools like:
Apollo for lead generation
LinkedIn Sales Navigator for prospecting
Instantly.ai for email outreach
Claude and ChatGPT for personalization
Virtual assistants for human follow-up
Nicholas shared that Suite Fleet uses AI-assisted prospecting, but still keeps humans involved. AI can help build the list, but a real person can make the call, qualify interest, and set up the discovery conversation.
That distinction matters.
Automation can create scale.
Humans create trust.
For B2C companies, Reneta pointed to a different playbook:
SEO
Google reviews
Website content
Word of mouth
Supplier-led growth
Social media experiments
AI-assisted content creation
Supplier-led growth means your providers help bring in customers because the platform makes their lives better. In Foodom’s case, happy chefs can bring customers to the platform because it helps them grow their own business too.
That is better than a random ad campaign.
Reneta also joked that a small ad budget disappeared quickly with little return. Many founders have had that exact “well, that was educational” moment.
The lesson: AI can help with acquisition, but it cannot fix a vague customer profile or a weak offer.
Make sure you know who you serve and why they care.
5. SEO Is No Longer the Whole Game
The panel also touched on a major shift in content marketing: founders now need to think beyond traditional SEO.
SEO stands for Search Engine Optimization, which means structuring content so Google can understand and rank it.
But as more people use ChatGPT, Claude, Perplexity, Gemini, and other AI tools to answer questions, companies also need to think about how AI systems understand their brand.
This is where terms like AEO and GEO come in.
AEO means Answer Engine Optimization. GEO means Generative Engine Optimization. In plain English, both refer to making your content easier for AI answer engines to find, understand, and cite.
Rich shared that when he writes blogs now, he thinks about both SEO and AI answer engines. That often means using:
Clear question-based headings
Short, direct answers
FAQs
Plain-language explanations
Specific examples
Content that directly answers buyer questions
This is a real shift.
A blog post is no longer just competing for a Google ranking.
It may also become the source an AI assistant uses when someone asks, “Who provides this service near me?” or “What companies help with this problem?”
My suggestion is simple: keep writing useful content, but make it easier for humans and AI systems to understand.
Clear beats clever.
Specific beats generic.
6. Keep Humans in the Loop
One of the strongest themes of the panel was that AI should be used “in addition to,” not “in place of,” human judgment.
That came up in customer support, writing, sales, research, strategy, and privacy.
Reneta said AI can help make customer communication warmer and more polished. As someone who described herself as very direct, she uses AI to soften messages and make them more customer-friendly.
That is a good use case.
The founder still decides what needs to be said.
AI helps improve how it is said.
Nicholas made a similar point about thought leadership. He said writing helps founders sharpen their thinking. If you outsource all of your writing to AI, you may produce polished content, but you may not actually understand your own ideas better.
That is a dangerous trade.
There is a big difference between using AI as an editor and using AI as a substitute for having a point of view.
My recommendation: use AI to draft, organize, summarize, and challenge your thinking.
But make sure the final opinion is yours.
7. Privacy and Confidential Data Still Matter
The panel also discussed one of the most important issues for small businesses: what should not go into AI tools?
Scott brought up a practical example: an accounting firm using AI to check financial work. That sounds useful, but the data may include sensitive client information.
His suggested approach was to use synthetic data.
Synthetic data means fake data that looks structurally similar to the real data but does not expose real customer details.
For example, instead of uploading an actual client spreadsheet into an AI tool, a business could use AI to help build formulas, checks, or workflows using sample data. Then the real data can be applied inside a controlled environment.
That is a smart middle ground.
AI helps build the process.
Humans protect the sensitive information.
The panel also discussed enterprise tools like Microsoft Copilot, where organizations may have data protections and internal controls. But the core warning still applies: founders should not casually paste confidential client data, health information, financial records, or proprietary IP into public AI tools.
The practical rule: assume anything sensitive needs extra care.
Do not let convenience outrank trust.
8. The Best AI Strategy Is Still a Founder Strategy
The panel was not really about tools.
Yes, Apollo, Claude, ChatGPT, Perplexity, Blaze AI, Notion, Slack, Instantly.ai, Gemini, Gamma, and other tools came up.
But the bigger lesson was about how founders think.
The best AI users on the panel were not chasing every shiny new product.
They were asking better questions:
What is the bottleneck?
What work is repetitive?
What needs human judgment?
What can we automate safely?
What customer experience do we want?
What data should we protect?
What process should we improve first?
That is the right mindset.
AI is not the strategy.
AI is leverage.
And leverage only works when it is pointed in the right direction.
Recommendations for Founders
If you are trying to figure out where AI fits in your company, start here:
Map one workflow.
Pick something specific, such as lead generation, customer support, weekly planning, content creation, or research.Define the human role.
Decide where judgment, empathy, approval, or relationship-building still needs to happen.Test one tool at a time.
Do not rebuild your whole company around five tools you barely understand.Protect sensitive data.
Use synthetic data, internal tools, or secure environments when working with private information.Measure the result.
Did it save time? Improve quality? Increase leads? Reduce errors? If not, it may just be a toy.
Final Takeaway
The most useful founders are not asking, “What AI tool should I use?”
They are asking, “What part of my business needs leverage?”
Start there. Pick one bottleneck, build one workflow, keep a human in the loop, and let AI make your team faster without making your company less human.
