Tangled: Tools, Traps, and the Human Touch | Over the Bull®

What does Greek mythology have to do with AI? The story of Icarus has endured for centuries because it is more than myth—it is metaphor. A young man, fitted with wings of wax, soared to freedom only to fall because…

Published September 3, 2025|, , , ,
Square podcast cover titled “Over The Bull” with an old-fashioned circular illustration of a man reaching for a floating object above water. Episode #25 is titled “Tangled: Tools, Traps, and the Human Touch.”.

What does Greek mythology have to do with AI? The story of Icarus has endured for centuries because it is more than myth—it is metaphor. A young man, fitted with wings of wax, soared to freedom only to fall because he flew too close to the sun. The very tool that gave him liberation became the source of his destruction. This story is an apt illustration of where we stand today with artificial intelligence. The tools promise freedom, efficiency, even transformation. But when misapplied or misunderstood, those same tools threaten to entangle us in loops of inefficiency, frustration, and lost trust. For small businesses, the stakes are even higher. Without the cushion of billion-dollar budgets or global market dominance, one wrong turn with AI could unravel hard-earned progress.

Trust, Friction, and the Customer’s Invisible Ledger

Every interaction writes a line in a customer’s invisible ledger: was this easier than I expected or harder, did I feel respected or handled, did I leave with more confidence or less. AI can help you write the right lines—routing simple requests, finding the right record, surfacing context for a rep, filling out forms a customer shouldn’t have to touch. But the moment a system traps someone in an unhelpful loop, misses an obvious emotional cue, or turns a practical problem into a scavenger hunt, that ledger shows a debt. Small businesses cannot afford that debt. A tech giant can absorb churn with sheer volume and familiarity; you cannot. When I encounter a bot that won’t get out of the way, I don’t think, “what progressive efficiency.” I think, “this company doesn’t want to spend money helping me.” That’s not a software issue; that’s a brand issue. And brands are built or broken by what people feel after they finish doing business with you. AI isn’t a way around that. It’s an amplifier of either your respect for a customer’s time or your disregard for it.

The hard question isn’t whether AI works in the abstract; it’s whether it works here, with your clientele, on this step of their journey. A contractor’s client who just approved a five-figure project doesn’t want a bot to triage their nervous question about a timeline slip. A parent who needs a replacement product before the weekend tournament doesn’t want to be asked again for the order number they already typed. AI earns its keep when it collapses small frictions without flattening human dignity. That’s the standard.

The Momentum You Can’t Ignore—and the Floor That Keeps Rising

There is a second reality to hold alongside the first: the baseline is moving up whether we like it or not. People are finding answers in conversational systems. Discovery is tilting into social commerce. Direct messaging converts when it is timely, relevant, and respectful. None of that eliminates the work of building an owned presence; it just raises the minimum expectations around speed, clarity, and personalization. I often hear claims that AI “levels the playing field.” It does not. It raises the floor. If you’re below the floor, you aren’t even in the game; if you meet the floor, you’re merely qualified to play. Advantage still belongs to the teams who get past competence into craft.

Think about that in practical terms. Yesterday, “good enough” content might have ranked if you pointed enough links at it and tuned the metadata. Today, generic output gets ignored by humans and discounted by systems trained to recognize sameness. Yesterday, responding by the next business day felt acceptable. Today, a competitor with a lightweight AI-assisted service pipeline can follow up in an hour with all the details already in front of a real rep. Yesterday, a tidy site map signaled organization. Today, structured data, clear topical clusters, and conversationally framed answers signal understanding. Progress is not optional. Slowness has a cost.

From SEO to GEO: Designing for Humans and Machines at Once

Search Engine Optimization still matters, but the nature of “search” is expanding beyond lists of blue links. Generative systems pull from multiple sources, compress context, and present a synthesized answer. That changes how we design pages and plan content. You still need an authoritative home base—your site remains the place you control, the archive of your expertise, the hub that feeds every other surface—but the pages must now be understandable to models and welcoming to people. That means careful use of schema so that products, services, reviews, locations, FAQs, and policies are unambiguous. It means writing in a voice that reads like a conversation with a competent human, not a term paper stitched together from clichés. It means building pages around jobs-to-be-done, where a question is asked and decisively answered with the next, natural action in reach.

Here’s the nuance: “conversational” does not mean “short.” Generative systems prefer crisp, complete answers, but crisp is not the same as shallow. A page can be both approachable and deep if its structure respects how questions actually unfold. Lead with a direct answer. Follow with the context that matters. Add the technical specifics someone making a purchase would care about. Support it with proof—photos, case studies, policies, numbers that a buyer would bring to their finance person. Then make it scannable without turning it into a checklist of empty subheads. This is the art of writing for dual audiences: a human who wants to understand and a model that needs to parse.

The Allure of Easy and the Cost of Herd Thinking

A thriving market exists to sell you shortcuts. Most of them are expensive detours. “Build a site in an afternoon.” “Rank by next week.” “Automate your entire pipeline with one tool.” Easy sells because it flatters the part of us that wants to be clever and frugal at the same time. But commerce punishes shortcuts that ignore complexity. Your business is not average; your audience is not generic; your offer is not interchangeable. The more your tools treat you as a template, the more you will be forced to behave like everyone else. Herd behavior feels safe until you realize most of the herd is underperforming.

This is not an argument against efficiency. It’s an argument for precision. Use tools to eliminate waste, not to erase judgment. Use automation where it removes drudgery, not where it blocks empathy. And when a vendor insists that their platform solves strategy, that’s your cue to step back. Tools are accelerants. They make a good plan faster and a bad plan worse. Before you adopt a capability, write down the outcome you need, the constraint it must respect, and the interaction you refuse to compromise. If a tool can’t fit inside those lines, it isn’t for you.

Human Where It Counts, Automated Where It Helps

The most resilient service models put automation behind the scenes and people at the hinge points. Let AI assemble the dossier before a rep picks up the phone so no one has to repeat their order number. Let it draft the follow-up email that a human then trims and sends in their own voice. Let it spot the pattern in support tickets that hints at a product defect so you fix the root cause. But when someone is anxious, confused, or making a meaningful commitment, get a human there fast. That balance looks different in every business, but the principle travels: apply automation to process, not relationship.

There’s a second layer to this balance—timing. Consumers move through a series of mental postures: exploration, evaluation, commitment, reassurance. AI can be immensely helpful at the beginning, where people want ideas, comparisons, and light guidance without pressure. It is counterproductive at the moment of commitment, where a buyer wants accountability and nuance. And it is damaging when used in a reassurance moment after a problem, where tone and improvisation matter. Mapping tools to posture keeps you from making the wrong kind of speed in the wrong moment.

Content That Sounds Like a Person Who Knows

The web is bloated with copy that reads like it was generated by committee or machine. People can feel it. So can the systems that prioritize engagement. The antidote is authorship—writing that sounds like a person who knows and cares. That doesn’t require theatrics. It requires specificity, narrative, and proof. Explain how a service actually works. Show what to expect in week one versus week four. Name the tradeoffs you routinely help clients navigate. Share the three mistakes you see most often and how you prevent them. This kind of writing does more than persuade a prospect; it seeds the signals that generative systems look for when deciding whose perspective to include. Authority is not a posture; it’s a trail of useful, consistent, verifiable contributions.

Structure matters too. Organize content by real tasks instead of internal departments. If a customer is wondering how to size a pump for a backyard water feature, they should find a plain-English explanation, a calculator or table, a recommendation guideline, and a path to talk to a human—on one page. If they’re comparing service tiers, show the difference that actually impacts outcomes, not just feature lists. If you operate in multiple cities, don’t stuff a hundred place names in a footer. Build location pages that reflect local proof, local constraints, and local outcomes. Good information architecture is invisible hospitality.

Planning Before Purchasing

Strategy isn’t a luxury; it’s the only margin of safety left. Begin with goals with numbers attached: time-to-first-response, qualified lead volume, customer lifetime value, repeat purchase rate, average resolution time, cost per acquisition by channel. Name the constraints you respect: service capacity, brand voice, regulatory boundaries, seasonal demand. Choose a small set of leading indicators to watch weekly and a set of outcome metrics to watch monthly. Only then do you pick tools. If a purchase doesn’t ladder into a metric you’re committed to, you’ve bought a distraction.

The same discipline applies to content. Decide which journeys matter most in the next quarter. Outline the key questions in those journeys, and assign pages, videos, or guides to answer them at the appropriate depth. Identify the proof you’ll need—photos, quotes, data—and collect it before you write. If AI assists in drafting, treat its output as scaffolding, not structure. Your team’s experience is the structure. Edit until the piece sounds like someone on your team actually speaking, with the edges left on. Then mark it up with the technical polish that helps machines help people find it.

Measuring What Matters and Iterating Without Panic

Once the plan is live, measure like an operator, not a hobbyist. Vanity metrics inflate confidence and starve insight. Track response times, first-contact resolution, funnel dropout moments, pages that get visitors stuck, phrases that precede refunds, and touches required to close. Watch for places where automation escalates the wrong things or buries the right ones. Pay attention to qualitative signals in support transcripts and sales notes. The point of measurement is not to create a dashboard; it’s to find one constraint each month that you can shrink. That steady reduction is where compounding advantage comes from.

Iteration should be continuous but not frantic. Change one variable per page or process at a time. Give new content enough time to gather signals before you rewrite it. When a system is underperforming, don’t swap the tool before you inspect the workflow and the handoffs. Most failures are misalignments, not malfunctions. And when you do replace something, document why in a sentence that would make sense to a new hire. Discipline creates culture; culture creates resilience; resilience keeps you out of the panic that shortcuts promise to relieve and usually worsen.

Service as a Strategy

If there’s a durable advantage available to small businesses, it is service. Not service as a slogan, but as an operating system. Answer quickly, in your own voice, with the context already at hand. Call when an email will create needless friction. Give the customer the sensation of dealing with a company that remembers them, because you do. Automate the memory. Humanize the response. In a market saturated with artificial polish, competence with kindness is not quaint; it’s premium. When you build around that, AI becomes a set of backstage tools that make your team faster and clearer without putting a robot on the stage where a person belongs.

There is also an identity component here. Customers choose you not only for what you sell but for how you approach the work. A business that respects attention, reduces effort, and tells the truth about tradeoffs is rare enough to be magnetic. When AI supports that identity, it earns its place. When it undermines that identity, it has to go—no matter how slick the demo looked.

Avoiding the Melt

The Icarus lesson isn’t “never fly.” It’s “know your materials, your limits, your weather, and your objective.” AI is powerful material. It carries heat. Used wisely, it saves time you can reinvest in craft, clarity, and care. Used carelessly, it convinces you that speed is the same as progress and that sameness is the same as scale. The way through is narrower than the hype suggests, but it is navigable: define the outcomes that matter, design experiences for the moments that matter, give machines the jobs they’re good at, put people where meaning is made, and keep the authority to say no. If you do that consistently, you’ll find the altitude that fits your wings—and you’ll stay there long enough to build something that lasts.

RESOURCES:

MARKETING TREND ARTICLES:

Customer Backlash, Emotional Intelligence & AI Complaints

  • ServiceNow (UK AI Customer Perception Study):

https://www.servicenow.com/uk/blogs/2025/how-uk-customers-feel-about-ai

LISTEN TO THE FULL EPISODE:

Tangled: Tools, Traps, and the Human Touch | Over the Bull®

What does Greek mythology have to do with AI? The story of Icarus has endured for centuries because it is more than myth—it is metaphor. A young man, fitted with wings of wax, soared to freedom only to fall because he flew too close to the sun. The very tool that gave him liberation became…

Published September 3, 2025|, , , ,
Square podcast cover titled “Over The Bull” with an old-fashioned circular illustration of a man reaching for a floating object above water. Episode #25 is titled “Tangled: Tools, Traps, and the Human Touch.”.

What does Greek mythology have to do with AI? The story of Icarus has endured for centuries because it is more than myth—it is metaphor. A young man, fitted with wings of wax, soared to freedom only to fall because he flew too close to the sun. The very tool that gave him liberation became the source of his destruction. This story is an apt illustration of where we stand today with artificial intelligence. The tools promise freedom, efficiency, even transformation. But when misapplied or misunderstood, those same tools threaten to entangle us in loops of inefficiency, frustration, and lost trust. For small businesses, the stakes are even higher. Without the cushion of billion-dollar budgets or global market dominance, one wrong turn with AI could unravel hard-earned progress.

Trust, Friction, and the Customer’s Invisible Ledger

Every interaction writes a line in a customer’s invisible ledger: was this easier than I expected or harder, did I feel respected or handled, did I leave with more confidence or less. AI can help you write the right lines—routing simple requests, finding the right record, surfacing context for a rep, filling out forms a customer shouldn’t have to touch. But the moment a system traps someone in an unhelpful loop, misses an obvious emotional cue, or turns a practical problem into a scavenger hunt, that ledger shows a debt. Small businesses cannot afford that debt. A tech giant can absorb churn with sheer volume and familiarity; you cannot. When I encounter a bot that won’t get out of the way, I don’t think, “what progressive efficiency.” I think, “this company doesn’t want to spend money helping me.” That’s not a software issue; that’s a brand issue. And brands are built or broken by what people feel after they finish doing business with you. AI isn’t a way around that. It’s an amplifier of either your respect for a customer’s time or your disregard for it.

The hard question isn’t whether AI works in the abstract; it’s whether it works here, with your clientele, on this step of their journey. A contractor’s client who just approved a five-figure project doesn’t want a bot to triage their nervous question about a timeline slip. A parent who needs a replacement product before the weekend tournament doesn’t want to be asked again for the order number they already typed. AI earns its keep when it collapses small frictions without flattening human dignity. That’s the standard.

The Momentum You Can’t Ignore—and the Floor That Keeps Rising

There is a second reality to hold alongside the first: the baseline is moving up whether we like it or not. People are finding answers in conversational systems. Discovery is tilting into social commerce. Direct messaging converts when it is timely, relevant, and respectful. None of that eliminates the work of building an owned presence; it just raises the minimum expectations around speed, clarity, and personalization. I often hear claims that AI “levels the playing field.” It does not. It raises the floor. If you’re below the floor, you aren’t even in the game; if you meet the floor, you’re merely qualified to play. Advantage still belongs to the teams who get past competence into craft.

Think about that in practical terms. Yesterday, “good enough” content might have ranked if you pointed enough links at it and tuned the metadata. Today, generic output gets ignored by humans and discounted by systems trained to recognize sameness. Yesterday, responding by the next business day felt acceptable. Today, a competitor with a lightweight AI-assisted service pipeline can follow up in an hour with all the details already in front of a real rep. Yesterday, a tidy site map signaled organization. Today, structured data, clear topical clusters, and conversationally framed answers signal understanding. Progress is not optional. Slowness has a cost.

From SEO to GEO: Designing for Humans and Machines at Once

Search Engine Optimization still matters, but the nature of “search” is expanding beyond lists of blue links. Generative systems pull from multiple sources, compress context, and present a synthesized answer. That changes how we design pages and plan content. You still need an authoritative home base—your site remains the place you control, the archive of your expertise, the hub that feeds every other surface—but the pages must now be understandable to models and welcoming to people. That means careful use of schema so that products, services, reviews, locations, FAQs, and policies are unambiguous. It means writing in a voice that reads like a conversation with a competent human, not a term paper stitched together from clichés. It means building pages around jobs-to-be-done, where a question is asked and decisively answered with the next, natural action in reach.

Here’s the nuance: “conversational” does not mean “short.” Generative systems prefer crisp, complete answers, but crisp is not the same as shallow. A page can be both approachable and deep if its structure respects how questions actually unfold. Lead with a direct answer. Follow with the context that matters. Add the technical specifics someone making a purchase would care about. Support it with proof—photos, case studies, policies, numbers that a buyer would bring to their finance person. Then make it scannable without turning it into a checklist of empty subheads. This is the art of writing for dual audiences: a human who wants to understand and a model that needs to parse.

The Allure of Easy and the Cost of Herd Thinking

A thriving market exists to sell you shortcuts. Most of them are expensive detours. “Build a site in an afternoon.” “Rank by next week.” “Automate your entire pipeline with one tool.” Easy sells because it flatters the part of us that wants to be clever and frugal at the same time. But commerce punishes shortcuts that ignore complexity. Your business is not average; your audience is not generic; your offer is not interchangeable. The more your tools treat you as a template, the more you will be forced to behave like everyone else. Herd behavior feels safe until you realize most of the herd is underperforming.

This is not an argument against efficiency. It’s an argument for precision. Use tools to eliminate waste, not to erase judgment. Use automation where it removes drudgery, not where it blocks empathy. And when a vendor insists that their platform solves strategy, that’s your cue to step back. Tools are accelerants. They make a good plan faster and a bad plan worse. Before you adopt a capability, write down the outcome you need, the constraint it must respect, and the interaction you refuse to compromise. If a tool can’t fit inside those lines, it isn’t for you.

Human Where It Counts, Automated Where It Helps

The most resilient service models put automation behind the scenes and people at the hinge points. Let AI assemble the dossier before a rep picks up the phone so no one has to repeat their order number. Let it draft the follow-up email that a human then trims and sends in their own voice. Let it spot the pattern in support tickets that hints at a product defect so you fix the root cause. But when someone is anxious, confused, or making a meaningful commitment, get a human there fast. That balance looks different in every business, but the principle travels: apply automation to process, not relationship.

There’s a second layer to this balance—timing. Consumers move through a series of mental postures: exploration, evaluation, commitment, reassurance. AI can be immensely helpful at the beginning, where people want ideas, comparisons, and light guidance without pressure. It is counterproductive at the moment of commitment, where a buyer wants accountability and nuance. And it is damaging when used in a reassurance moment after a problem, where tone and improvisation matter. Mapping tools to posture keeps you from making the wrong kind of speed in the wrong moment.

Content That Sounds Like a Person Who Knows

The web is bloated with copy that reads like it was generated by committee or machine. People can feel it. So can the systems that prioritize engagement. The antidote is authorship—writing that sounds like a person who knows and cares. That doesn’t require theatrics. It requires specificity, narrative, and proof. Explain how a service actually works. Show what to expect in week one versus week four. Name the tradeoffs you routinely help clients navigate. Share the three mistakes you see most often and how you prevent them. This kind of writing does more than persuade a prospect; it seeds the signals that generative systems look for when deciding whose perspective to include. Authority is not a posture; it’s a trail of useful, consistent, verifiable contributions.

Structure matters too. Organize content by real tasks instead of internal departments. If a customer is wondering how to size a pump for a backyard water feature, they should find a plain-English explanation, a calculator or table, a recommendation guideline, and a path to talk to a human—on one page. If they’re comparing service tiers, show the difference that actually impacts outcomes, not just feature lists. If you operate in multiple cities, don’t stuff a hundred place names in a footer. Build location pages that reflect local proof, local constraints, and local outcomes. Good information architecture is invisible hospitality.

Planning Before Purchasing

Strategy isn’t a luxury; it’s the only margin of safety left. Begin with goals with numbers attached: time-to-first-response, qualified lead volume, customer lifetime value, repeat purchase rate, average resolution time, cost per acquisition by channel. Name the constraints you respect: service capacity, brand voice, regulatory boundaries, seasonal demand. Choose a small set of leading indicators to watch weekly and a set of outcome metrics to watch monthly. Only then do you pick tools. If a purchase doesn’t ladder into a metric you’re committed to, you’ve bought a distraction.

The same discipline applies to content. Decide which journeys matter most in the next quarter. Outline the key questions in those journeys, and assign pages, videos, or guides to answer them at the appropriate depth. Identify the proof you’ll need—photos, quotes, data—and collect it before you write. If AI assists in drafting, treat its output as scaffolding, not structure. Your team’s experience is the structure. Edit until the piece sounds like someone on your team actually speaking, with the edges left on. Then mark it up with the technical polish that helps machines help people find it.

Measuring What Matters and Iterating Without Panic

Once the plan is live, measure like an operator, not a hobbyist. Vanity metrics inflate confidence and starve insight. Track response times, first-contact resolution, funnel dropout moments, pages that get visitors stuck, phrases that precede refunds, and touches required to close. Watch for places where automation escalates the wrong things or buries the right ones. Pay attention to qualitative signals in support transcripts and sales notes. The point of measurement is not to create a dashboard; it’s to find one constraint each month that you can shrink. That steady reduction is where compounding advantage comes from.

Iteration should be continuous but not frantic. Change one variable per page or process at a time. Give new content enough time to gather signals before you rewrite it. When a system is underperforming, don’t swap the tool before you inspect the workflow and the handoffs. Most failures are misalignments, not malfunctions. And when you do replace something, document why in a sentence that would make sense to a new hire. Discipline creates culture; culture creates resilience; resilience keeps you out of the panic that shortcuts promise to relieve and usually worsen.

Service as a Strategy

If there’s a durable advantage available to small businesses, it is service. Not service as a slogan, but as an operating system. Answer quickly, in your own voice, with the context already at hand. Call when an email will create needless friction. Give the customer the sensation of dealing with a company that remembers them, because you do. Automate the memory. Humanize the response. In a market saturated with artificial polish, competence with kindness is not quaint; it’s premium. When you build around that, AI becomes a set of backstage tools that make your team faster and clearer without putting a robot on the stage where a person belongs.

There is also an identity component here. Customers choose you not only for what you sell but for how you approach the work. A business that respects attention, reduces effort, and tells the truth about tradeoffs is rare enough to be magnetic. When AI supports that identity, it earns its place. When it undermines that identity, it has to go—no matter how slick the demo looked.

Avoiding the Melt

The Icarus lesson isn’t “never fly.” It’s “know your materials, your limits, your weather, and your objective.” AI is powerful material. It carries heat. Used wisely, it saves time you can reinvest in craft, clarity, and care. Used carelessly, it convinces you that speed is the same as progress and that sameness is the same as scale. The way through is narrower than the hype suggests, but it is navigable: define the outcomes that matter, design experiences for the moments that matter, give machines the jobs they’re good at, put people where meaning is made, and keep the authority to say no. If you do that consistently, you’ll find the altitude that fits your wings—and you’ll stay there long enough to build something that lasts.

RESOURCES:

MARKETING TREND ARTICLES:

Customer Backlash, Emotional Intelligence & AI Complaints

  • ServiceNow (UK AI Customer Perception Study):

https://www.servicenow.com/uk/blogs/2025/how-uk-customers-feel-about-ai

LISTEN TO THE FULL EPISODE: