AI & Small Business: What Real Founders Are Actually Doing (And What They’re Warning Against)

A topic hub pulling real quotes from Unscripted Small Business podcast guests — founders, operators, and consultants sharing what’s actually working (and what isn’t) when it comes to AI in their businesses.

---

Every business owner is being told to “use AI.” Few are being told how — or warned about what it costs them if they use it wrong. We’ve been having this conversation on the Unscripted Small Business podcast across dozens of industries, and the picture that emerges from real founders is a lot more nuanced than the hype.

Some are using AI to build entire workflows that run without them. Some are deliberately avoiding it for their most important work. Some are using AI to make their content sound less like AI. And at least one tested AI content head-to-head against human writing and found a clear winner.

Here’s what they actually said.

---

Stop driving. Start dispatching.

Mason McCumber — SEO and AI consultant at buildwithmm.com — gave us the clearest framework for how to think about the shift happening right now.

Mason’s key insight: getting a great result from AI once isn’t a win — it’s only a win if you can reproduce it. He calls this the repeatability is scalability principle. If you can’t recreate the output, you haven’t built a system. You’ve just gotten lucky.

His practical solution: require the AI to document its own steps. Whatever process it used to produce the output, have it write that down in a format you can reuse. He uses Obsidian to store these as searchable markdown files that can be fed back into future sessions.

He also makes a distinction most AI content skips: not everything repetitive belongs in an AI workflow. If every time you get an email from a specific sender you forward it somewhere — that’s a connector problem, not an AI problem. Route it through Zapier or Make, not a language model. Use AI for work that involves language, judgment, or variable inputs. Use automation for everything deterministic.

---

Using AI to differentiate from AI

Conrad Coe, Operations Manager at Exotic Vehicle Wraps, offered one of the most counterintuitive takes we’ve heard on the show.

His shop uses Claude and Lovable to build custom consultation apps and workflow tools. They use AI to validate their pricing against industry data and analyze competitor markets. But when it comes to marketing copy, they’ve identified the tells — the em-dashes, the generic phrasing, the sameness — and they scrub them out deliberately.

The underlying point is sharp: AI-generated content has created uniform mediocrity across his industry. A shop open for three months can now produce content that looks identical to a shop with 16 years of experience. That’s a threat — but it’s also an opportunity. If everyone sounds the same, sounding human is a competitive advantage.

---

AI makes us lazy. That’s the real risk.

Bill Kasko has been running Frontline Source Group for 21 years by doing the opposite of what his industry expects. On AI, he’s direct about his concern.

His specific pain point: AI-generated job applications flooding his recruiting pipeline, requiring his team to spend significant resources separating real candidates from synthetic ones. That’s a new cost of doing business that didn’t exist three years ago.

He also observed something broader happening in search: the landscape is fragmenting. His niece told him they needed to be on TikTok. He resisted — then came around. “You have to be findable across Google, Bing, TikTok, Reddit, YouTube — all of it. One channel isn’t enough anymore.” AI is accelerating that fragmentation by changing where answers come from.

---

Human-written content still wins. She tested it.

Tianna Mamalick has spent 12 years doing SEO exclusively for small businesses through SMB Marketing School. She didn’t theorize about AI content — she tested it.

Her position isn’t anti-AI — it’s precise. Use AI for outlines. Use it to get past the blank page. Use it for research and ideation. But have a human write the actual content — someone with a genuine point of view and real experience in the topic.

The content that’s winning has perspective. It has an author who’s actually done the thing being described. AI can explain what a service is. It can’t describe what it feels like to have done 200 jobs and learned what the client doesn’t tell you upfront. That gap is where ranking advantage still lives.

---

AI is a better hammer. Use it like one.

Chad T. Jenkins has built and exited over 50 companies. His current model — the CoLAB at SeedSpark — is built entirely around combining what different entrepreneurs have. He’s clear-eyed about where AI fits in that picture.

Chad’s practical example: he was asked to present to Coca-Cola on AI. It came up on him while traveling — he remembered at 3 a.m. that he was presenting to 150–200 people later that day. He engaged his AI assistant, had a conversation, and prompted it to produce a McKinsey-level 49-slide deck via Genspark. He tweaked two slides. Over 50% of the audience engaged afterward.

His framing cuts through the anxiety: the entrepreneurs who will be displaced aren’t the ones AI replaces — they’re the ones who refuse to treat AI as a tool. Someone empowered by AI is coming for the work of someone who isn’t. That’s not a threat, it’s just physics.

---

Forced to learn SEO because the paid ads were blocked

Marc Pitts co-founded DiscountVapePen.com in 2014. By 2016, Google and Facebook had cut off paid advertising in his category entirely. That forced a depth of SEO and content knowledge most business owners never develop.

Marc’s biggest SEO lesson applies directly to the AI content conversation: content written for search is not written the way you’d write for a human reader in conversation. His partner went deep on structure — proper H2s, search-intent targeting, pro-tip callouts — and the results were exponential compared to their earlier editorial approach.

The framing Jeremy offered during that conversation holds: content travels human-to-bot-to-human. A human asks a question. A bot decides whether your content surfaces. A human reads what surfaces. Miss the bot and the human never arrives. AI is changing who that “bot” is — but the principle is the same.

---

The soul problem with AI content

Lynn Colepaugh, founder of Cyber PR Army, works with creators who are building businesses around their work. Her concern about AI isn’t technical — it’s relational.

Her point isn’t that AI can’t produce content — it’s that AI can’t produce the specific creative and relational chemistry that develops between particular people with particular histories. The reason people stay loyal to a brand, a creator, or a service isn’t usually the quality of the output. It’s the person behind it. That’s not replicable.

Her practical guidance: use AI as a productivity multiplier, not a voice replacement. One channel, done with genuine human consistency, will outperform ten channels running on automated content.

---

Creating marketing strategy for people AND AI

Sara Nay has been at Duct Tape Marketing for 15 years, rising from intern to CEO. She’s watched AI reshape not just what content to create, but what it has to do structurally to be surfaced.

Her practical direction: FAQs with direct answers, clear headings, structured summaries. LLMs pull from Reddit, from your website, from everywhere — so the question isn’t just “do I rank on Google?” It’s “do I show up when someone asks ChatGPT or Perplexity about my category?”

She’s also precise about where AI falls short in strategy work: when businesses use it to generate content without first doing the foundational interviews — with leadership, with real customers — they train the AI on generic inputs and get generic outputs. The strategy has to come first. AI can execute on a good brief. It can’t write one.

---

AEO: Artificial Engine Optimization

Bruce Ashford of The Ashford Agency introduced a term that captures where visibility is heading: Artificial Engine Optimization — the practice of making your business citable to ChatGPT, Grok, Claude, and Perplexity, not just findable on Google.

His three-part AEO framework: make your website technically machine-readable (natural-language FAQs, structured data, direct answers to real questions), produce authoritative content consistently, and build social proof through Google and LinkedIn reviews that specifically name your business and what you do. The trust signals that have always mattered for SEO are the same signals that make you citable by AI systems — but the format requirements are changing.

---

What the pattern shows

Across nine different industries — logistics, automotive wrapping, staffing, SEO consulting, serial entrepreneurship, vape e-commerce, creator marketing, marketing strategy, and nonprofit consulting — the same themes surface:

• AI is a capability multiplier, not a voice. The founders winning with it are using it to do more of what they already do well — not to replace the thing that makes them distinctive.
• Repeatability matters more than the first result. Getting one great AI output is luck. Documenting the process so you can reproduce it is strategy.
• Human specificity is the differentiator. As AI raises the floor of content quality across every industry, the ceiling belongs to whoever sounds most like a real person with real experience.
• The question isn’t whether to use AI. It’s whether you’re using it to think more clearly or to think less.

---

These conversations are from the Unscripted Small Business Podcast — candid, unscripted interviews with founders, operators, and marketers. Hosted by Jeremy Rivera, Zaneta Chuniq Inpower, and Daniel Hill.

Browse all episodes: unscriptedsmallbusiness.com