AI has made publishing cheaper, while making attention harder to earn

If you run a small business, the pressure is familiar. An AI tool can produce a service-page draft, five social posts and a newsletter before you have finished your first coffee. The temptation is to publish all of it, because an empty content calendar feels like a problem that volume can solve.

Usually, it cannot.

The practical answer now is to publish fewer, deeper content pieces, then distribute each one more intelligently. “Deep” does not mean producing long articles stuffed with background information. It means giving the reader something generic AI advice cannot easily provide: a real example, a local observation, a trade-off you have seen firsthand, a price or process explanation, a mistake you can help someone avoid, or evidence from your own work.

AI is useful in that system. It can help collect questions, organize notes, identify gaps, create first drafts and turn one substantial article into email and social formats. It should reduce the mechanical work around publishing. It should not become the source of your point of view.

That distinction matters because the web is filling with content that is structurally competent and commercially interchangeable. HubSpot reports that 56% of marketers say the internet is now so saturated with AI-generated content that it is harder for high-quality work to stand out. The same report says 65% of marketing teams believe consumers are becoming better at identifying AI-generated content. (blog.hubspot.com)

Depth beats volume when the buyer needs confidence

The numbers support a sharper workflow, not a publishing arms race

The strongest case for more content is obvious: AI increases production capacity. Ahrefs surveyed 879 marketers and found that 87% use AI to create or help create content. Companies using AI published a median of 17 articles per month, compared with 12 among companies that did not use it. Ahrefs also found that 97% of respondents edit or review AI-assisted content, while only 4% publish primarily pure AI content. (ahrefs.com)

That is not evidence that small businesses should publish 42% more articles. It is evidence that the competitive baseline is moving. More businesses can now produce acceptable drafts. Acceptable drafts therefore have less value.

The more useful finding comes from the 2026 B2B research by the Content Marketing Institute. Among marketers using AI for content creation, 87% reported improved productivity and 80% reported improved operational efficiency. But the figures fell as the question moved from production to substance: 58% said content quality improved, 39% said content performance improved, and 12% said quality decreased. (contentmarketinginstitute.com)

That gap is the business case for depth. AI is very good at helping a team move faster. It is less reliable at deciding what deserves to exist, what a skeptical buyer needs to believe, or what your business knows that a competitor does not.

Practitioners are reporting the same pattern in plainer language. CMI’s research describes teams producing content faster while still struggling with differentiation and strategic clarity. Its top performers attribute success primarily to understanding their audience, producing high-quality content and possessing industry expertise. Those are not volume advantages. They are context advantages. (contentmarketinginstitute.com)

Search is becoming less generous to generic informational content

Search visibility is another reason to stop treating article count as the main output. Google does not ban AI-assisted writing. Its current guidance says generative AI can be useful for research and for adding structure to original content. The warning is aimed at generating many pages without adding value, which may fall under scaled content abuse. Google tells publishers to focus on accuracy, quality and relevance. (developers.google.com)

That wording is important. The risk is not that a machine touched the draft. The risk is that the finished page gives the reader no reason to choose it over the pages already available.

Search behavior is also changing. Pew Research Center analyzed 68,879 Google searches from browsing data collected in March 2025. About 18% produced an AI-generated summary. When a summary appeared, users clicked a traditional result in 8% of visits, compared with 15% when no summary appeared. They clicked a link inside the summary in just 1% of visits. (pewresearch.org)

For a small business, this does not mean informational content is useless. It means a generic answer may be consumed without a visit, while content that helps someone evaluate a provider still has a reason to earn the click. A comparison, diagnostic checklist, pricing explanation, project walkthrough, local case study or original dataset is more defensible than another article explaining a familiar concept in familiar language.

This is also why your content should connect to an actual commercial decision. If the piece could be summarized by an AI answer without losing anything important, it probably needs a stronger point of view, better evidence or a more specific audience.

A practical depth-first publishing cycle

1. Choose a decision, not a topic.

   Start with a question a real customer asks before hiring, buying or switching. “How do I choose a commercial photographer?” is more useful than “Photography trends.”

2. Collect firsthand material.

   Use customer questions, sales-call notes, support emails, project records, reviews and the owner’s experience. This is the raw material AI cannot supply honestly.

3. Take a position.

   State what you recommend, what you would avoid and what changes the answer. A useful article should make the reader’s next step clearer.

4. Add proof and boundaries.

   Show an example, explain the trade-off and say when your advice does not apply. Specific limits make advice more credible.

5. Use AI around the argument.

   Ask it to organize notes, challenge assumptions, suggest missing objections, create a concise summary or adapt the finished piece into other formats.

6. Distribute the source piece.

   Send it to your email list, give the sales team a useful excerpt, publish a short social version and link to it from relevant service pages. Do not create five shallow articles to imitate one strong one.

7. Review business signals.

   Look beyond pageviews. Track qualified inquiries, assisted conversions, replies, branded searches, sales-team usage and the questions prospects stop asking.

For most small businesses, the right cadence is one strong source piece plus consistent distribution

A small business rarely loses because it failed to publish enough blog posts. It loses because the owner spends scarce marketing time producing material that does not change a buyer’s understanding or behavior.

Enji’s 2025 survey of 245 small business owners puts that constraint in context: 57% said they spend just 1–5 hours per week on marketing, and 51% said they do not outsource any marketing tasks. The same research found that 80% rated their marketing as no better than somewhat effective. (enji.co)

Under those conditions, a high-frequency editorial program is usually the wrong goal. A better operating model is to choose one meaningful customer decision, build one genuinely useful source piece, and give it several jobs. It can answer a sales objection, support a service page, become a newsletter, generate social excerpts and give an employee or partner something worth sharing.

The important word is source. Distribution can still be frequent. The business does not need to publish one deep article and then disappear for a month. It needs to stop confusing repeated production with repeated usefulness.

There are exceptions. A publisher, media business or company with a large content team may benefit from a broader portfolio. A retailer may need regular product and seasonal content. A local business may need frequent updates tied to events, inventory or availability. But even those businesses should separate routine updates from opinion-led content. AI can handle more of the routine layer. Human attention should go to the pieces that build trust and preference.

What the current evidence says

The standard is not “human-written.” It is worth choosing

The useful question is not whether your next article contains AI-assisted sentences. It is whether the finished piece contains knowledge, judgement or proof that your ideal customer would struggle to get elsewhere.

That might be a contractor explaining why the cheapest material creates a repair problem later. It might be an accountant showing how a particular owner decision changed a cash-flow forecast. It might be a design studio documenting the moment a client’s attractive idea failed a real accessibility test. It might be a niche retailer explaining which product is a poor fit for a particular use case.

Those examples work because they contain selection. Someone with experience has decided what matters, what does not and what to do next. AI can help package that experience. It cannot be allowed to flatten it into agreeable advice.

So yes: as AI increases the volume of mediocre content, most small businesses should publish fewer, deeper pieces. Keep the useful frequency through distribution, updates and customer-facing reuse. Put the scarce human time into original evidence, clear recommendations and uncomfortable specifics. That is where a small business can still be recognized, cited, remembered and chosen.