PerspectiveLast updated: July 2026

The Cost of Acting on Attention Is Dropping

Your business already knows more than it uses. Customer-facing people already see repeated friction. AI lowers the cost of turning those patterns into systems.

The practical opportunity is not "AI transformation." It is converting complaints, workarounds, explanations, and objections into pages, scripts, forms, workflows, documentation, and product decisions before the noticing gets buried again.

The knowledge is already in the business

Not in a mystical way. Not because there is some hidden genius trapped inside the org chart. The knowledge is usually sitting in plain sight.

The account manager hears the same complaint every week.

The front desk answers the same confused question every morning.

The owner keeps personally fixing the same operational mess because nobody else knows how the decision is supposed to be made.

The sales person knows which explanation finally makes the buyer understand the offer.

The support person knows where customers get lost.

Most businesses are full of this kind of knowledge. They just do not treat it as knowledge. They treat it as noise, personality, customer service, hustle, or "just how the business works."

Business literature has cleaner names for this. Voice of the customer. Tacit knowledge. Organizational routines. Absorptive capacity. Customer discovery. Continuous improvement.

The names matter less than the pattern: businesses are always producing information about what is broken, confusing, slow, expensive, or underexplained. The problem is that most of this information never becomes a system.

The AI opportunity for small businesses is not that everyone suddenly gets access to magic software. The tools matter, but they are not the deepest shift.

The deeper shift is this:

The cost of acting on attention is dropping.

A repeated complaint can become a better intake form. A confusing sales process can become a clearer page. A founder's repeated explanation can become onboarding, documentation, follow-up emails, scripts, or a lightweight internal tool. A support pattern can become a product decision.

That used to require more budget, more technical help, more coordination, and more patience than many small businesses had.

Now the distance between noticing the problem and building the first version of the fix is much shorter.

That changes what attention is worth.

The work existed before the titles

Before "product teams," businesses still had customers.

Before journey maps, customers still got confused.

Before service design, someone still noticed that the phone kept ringing about the same problem.

Before product management became a formal title, someone still had to decide what was worth building, what was worth ignoring, and what the customer was actually trying to do.

The work existed before the title.

That matters because modern businesses often confuse the name of a practice with the practice itself. UX, product management, service design, design thinking. These can all be useful names. They can help a team slow down, look at the customer, and make better decisions.

Tim Brown's 2008 Harvard Business Review piece on design thinking describes it as a discipline that uses the designer's methods to match people's needs with what is technologically feasible and what a viable business strategy can turn into customer value.1 That is a useful formulation. It explains why design thinking became attractive to organizations that needed a way to make innovation less mystical and more repeatable.

But the title did not create the attention.

Design thinking did not invent customer attention, but it did repackage it.

That is not automatically bad. Packaging matters. A framework can help non-designers participate in ambiguous work. It can give a team a shared language. It can keep people from jumping straight to a solution before they understand the problem. Nielsen Norman Group's basic definition makes this explicit: design thinking is both an ideology and a process, built around a hands-on, user-centric approach to problem solving.2

The problem starts when the process replaces the contact with reality.

The workshop happens. The sticky notes go up. The persona is written. The journey map is produced. The team says "empathy" several times. Then the business ships the same decision it was already going to ship.

That is a performance. A performance of customer attention.

This is what I mean by postmodernized thinking: the activity gets separated from the source. Thinking becomes a framework about thinking. Customer attention becomes a workshop about customer attention. The diagram survives even when the noticing disappears.

That criticism is not anti-design. It is close to a criticism already happening inside design itself. Kipum Lee's open-access critique of design thinking in organizations argues that prevailing approaches can stay trapped in a making-oriented paradigm and, ironically, maintain the status quo they claim to challenge.3

That is the danger.

The useful version of a framework sharpens attention.

The useless version lets a business pretend it paid attention.

Proximity is not insight

Customer-facing people know things, but being close to the customer is not the same as understanding the customer.

An account manager may hear the same complaint every week and still treat each instance as a one-off annoyance. A support person may solve the same problem manually for two years without anyone turning that pattern into a better process. A founder may explain the business brilliantly on calls but never convert that explanation into the website, onboarding, or sales material.

The raw material is there. The conversion step is missing.

A repeated customer complaint is a product roadmap in disguise.

A repeated internal workaround is an operations roadmap in disguise.

A repeated founder explanation is a positioning asset in disguise.

A repeated buyer objection is a sales enablement asset in disguise.

Harvard Business School Online describes voice of the customer work as a way to understand customer aspirations, expectations, and pain points well enough to guide ideation, product development, marketing, and customer experience design.4 Product teams have their own version of this: collect feedback, identify recurring themes, prioritize what matters, and avoid building from assumption alone.5

But there is a catch.

Collecting feedback is not the same as converting feedback into action.

This is where small businesses often have an advantage they do not recognize. They are closer to the work. The distance between the owner, the customer, the sales conversation, and the delivery problem can be short. The signal is often less buried than it is inside a large company.

But short distance only helps if someone turns the signal into a decision.

The business does not need to become a software company to use this. It needs to notice what is already happening and turn the pattern into a system.

For a restaurant, that might be clearer pickup instructions, a better order status message, a simplified catering form, or a staff script for the same five questions.

For a home services company, it might be quote templates, job intake triage, follow-up sequences, before-and-after proof libraries, or a better handoff from sales to operations.

For a small professional services firm, it might be a client onboarding flow, a document checklist, a repeatable proposal structure, or a knowledge base built from the explanations the founder already gives on every call.

None of this is futuristic.

That is the point.

The value is not in making the business sound more like a tech company. The value is in removing drag from the way the business already works.

Why product roles get buried

This is also why product roles are so often misunderstood.

In theory, a product person owns the problem. They understand the customer. They shape the tradeoffs. They protect the product from becoming a pile of unrelated requests.

In practice, many organizations turn the product person into the container for everything they do not know where else to put.

Delivery coordination. Stakeholder management. Roadmap therapy. Internal politics. Status reporting. Requirements translation. Occasionally, yes, product judgment. But often the judgment gets treated as the decorative part of the role, not the reason the role exists.

If the organization does not know what product judgment is worth, it will bury that judgment under coordination work.

That same pattern shows up outside software companies.

A business says it wants growth, but rewards whoever keeps the current chaos moving.

A business says it wants better systems, but treats every attempt to create one as an interruption.

A business says it wants customer insight, but ignores the people who hear from customers every day.

The issue is not that the business lacks information. The issue is that the business has no reliable mechanism for turning information into change.

Knowledge-management research has been circling this problem for decades. Nonaka's SECI model is one of the best-known ways to describe how organizations convert tacit knowledge into explicit knowledge, combine it with other knowledge, and internalize it back into practice.6 A 2019 empirical study in Frontiers in Psychology found support for measuring these knowledge-generation processes and linked different dimensions of knowledge generation to organizational outcomes such as performance, innovativeness, and collective efficacy.6

Put less academically: the business has to turn what people know from experience into something the organization can use.

That mechanism is becoming cheaper to build.

People hold on tightly to their pain

There is a psychological layer here too.

Broken systems often provide status.

The owner drowning in manual work gets to feel indispensable. The manager who knows all the exceptions gets to feel necessary. The employee who suffers through the bad process gets to feel useful. The founder who personally fixes everything gets to remain the center of the business.

A better system can look like a loss if the old system is where someone gets their identity.

This is one reason small businesses resist help even when the benefit seems obvious from the outside. Fixing the process may require someone to admit that the current way is broken. Worse, it may require them to admit that the pain they are proud of is not proof of value. It is just pain.

People hold on tightly to their pain when the pain has been paying them in status.

The psychology here is not exotic. Samuelson and Zeckhauser's classic work on status quo bias found that people disproportionately stick with the current option, even in important real decisions.7 The current way of doing things has a psychological advantage simply because it is current. Add status, identity, sunk effort, and fear of losing control, and the resistance becomes stronger.

AI does not remove that problem.

In some cases, it intensifies it. If your sense of value is attached to suffering through a task, a tool that removes the task can feel like an insult. If your authority comes from being the only person who knows how the messy process works, documenting and simplifying the process can feel like a threat.

That is why the work cannot be framed as "AI transformation" in the abstract.

Most small businesses do not need a transformation speech. They need someone to sit with the actual work and ask better questions:

Where are customers confused?

Where are employees repeating themselves?

Where is the owner doing work that should be a system?

Where does the business rely on memory instead of process?

Where is chaos being mistaken for personality?

Where is pain being mistaken for proof?

Those are not software questions first. They are attention questions.

AI does not make everyone creative

AI does not make everyone creative.

It does not make everyone strategic. It does not make everyone good at product judgment. It does not make every business capable of redesigning itself.

Some people are good at executing a process but not redesigning it. Some people can evaluate a workflow once it exists but cannot imagine a better one from scratch. Some people can talk to customers all day and still miss the pattern.

That is fine. Different work requires different aptitudes.

The important change is that attentive people now have a shorter path from observation to prototype.

A capable operator can take a repeated complaint and draft the new customer email today.

They can turn the founder's explanation into a landing page this week.

They can build the first version of an intake form, quote template, support macro, training document, checklist, or internal workflow without waiting for a six-month software project.

They can test whether the fix helps.

This is where the advantage begins. Not with a subscription. With practice.

The subscription is easy to copy. The practice is not.

A competitor can buy the same AI tools. They cannot instantly copy the accumulated judgment your business builds by turning real friction into better systems week after week.

That judgment compounds.

You learn which customer questions matter. Which outputs need review. Which processes should be automated and which should remain human. Which parts of the business are actually constrained by labor, and which are constrained by unclear decisions.

That is the work.

This also keeps the AI argument honest. The U.S. Chamber of Commerce reports that almost 60% of small businesses now say they use AI, more than double from 2023.8 That does not mean almost 60% have built an AI practice. It means access is spreading. When access spreads, the advantage shifts from having the tool to knowing what to do with it.

What to do with this on Monday

Pick one repeated friction point.

Not an AI strategy. Not a transformation initiative. One repeated point of drag.

Examples:

  • Customers ask the same question before buying.
  • New clients arrive unprepared.
  • Quotes take too long to assemble.
  • Follow-up depends on memory.
  • The owner repeats the same explanation on every sales call.
  • Staff use three different versions of the same process.
  • A recurring customer complaint gets handled manually every time.

Write down the pattern in plain English.

Then ask what form it should take:

  • a page
  • a script
  • a checklist
  • an email sequence
  • an intake form
  • a quote template
  • a support macro
  • an internal guide
  • a lightweight tool
  • a changed handoff between people

Build the smallest version that could reduce the drag.

Run it for two weeks.

Measure one thing.

Did calls decrease? Did quote time drop? Did customers arrive more prepared? Did fewer tasks get missed? Did the owner stop answering the same question? Did the handoff improve?

If nothing changed, you learned cheaply.

If something changed, you have started building a practice.

This is the same logic behind lead-user research, but scaled down to a practical operating habit. Eric von Hippel's classic work showed that users can be a source of novel product concepts when their needs and preferences are systematically identified and incorporated.9 For a small business, the first version may be less formal: listen for the repeated need, build a small fix, test whether it changes the work.

The actual AI opportunity

The businesses that benefit from AI will not be the ones with the most dramatic language about AI.

They will be the ones that can notice a real pattern, turn it into a system, test it, and repeat the cycle.

That has always been valuable. The difference now is cost.

The cost of writing the first draft is lower.

The cost of prototyping the workflow is lower.

The cost of turning tacit knowledge into documentation is lower.

The cost of testing a small operational change is lower.

The cost of acting on attention is dropping.

That is the practical opportunity for small businesses.

Your business already knows more than it uses. The question is whether it can convert that knowledge into systems before the noticing gets buried again.

The advantage is not having AI.

The advantage is acting on what your business already knows.

Sources

Footnotes

  1. Tim Brown, "Design Thinking," Harvard Business Review, 2008. The article defines design thinking as matching people's needs with technological feasibility and business viability. https://readings.design/PDF/Tim%20Brown,%20Design%20Thinking.pdf ↩

  2. Sarah Gibbons, "Design Thinking 101," Nielsen Norman Group, 2016. NN/g describes design thinking as both an ideology and a process built around a hands-on, user-centric approach. https://www.nngroup.com/articles/design-thinking/ ↩

  3. Kipum Lee, "Critique of Design Thinking in Organizations: Strongholds and Shortcomings of the Making Paradigm," She Ji: The Journal of Design, Economics, and Innovation, 2021. Open access. https://doi.org/10.1016/j.sheji.2021.10.003 ↩

  4. Tim Stobierski, "Voice of the Customer: Strategies to Listen & Act Effectively," Harvard Business School Online, 2025. The article frames VoC as understanding customer expectations, preferences, pain points, and experiences well enough to guide product, marketing, and customer experience decisions. https://online.hbs.edu/blog/post/voice-of-customer ↩

  5. Productboard, "What Is Voice of the Customer in Product Development?" 2025. Vendor source, useful for practitioner framing rather than independent empirical claims. https://www.productboard.com/blog/informing-product-with-voice-of-customer/ ↩

  6. Maria Luisa Farnese, Barbara Barbieri, Antonio Chirumbolo, and Gerardo Patriotta, "Managing Knowledge in Organizations: A Nonaka's SECI Model Operationalization," Frontiers in Psychology, 2019. The paper tests a questionnaire for SECI knowledge-conversion processes and links knowledge-generation dimensions to organizational outcomes. https://pmc.ncbi.nlm.nih.gov/articles/PMC6914727/ ↩ ↩2

  7. William Samuelson and Richard Zeckhauser, "Status quo bias in decision making," Journal of Risk and Uncertainty, 1988. The abstract reports experiments and real decision contexts showing people disproportionately stick with the status quo. https://doi.org/10.1007/BF00055564 ↩

  8. U.S. Chamber of Commerce Technology Engagement Center, "Empowering Small Business: The Impact of Technology on U.S. Small Business," 2025. The report says almost 60% of small businesses use AI for business operations, more than double from 2023. https://www.uschamber.com/technology/empowering-small-business-the-impact-of-technology-on-u-s-small-business ↩

  9. Eric von Hippel, "Lead Users: A Source of Novel Product Concepts," Management Science, 1986. The paper explores systematic identification of lead users and incorporating their perceptions and preferences into industrial and consumer product development. https://doi.org/10.1287/mnsc.32.7.791 ↩