Most marketers are celebrating AI adoption rates, and I see an even bigger opportunity unfolding.
The global AI-powered content creation market exploded from $2.15 billion in 2024 to a projected $10.59 billion by 2033. AI content writing tool adoption skyrocketed by 400% over just two years, with 75% of marketers now relying on AI for content creation.
Here’s what they’re missing: this isn’t a productivity unlock. It’s a supply shock that collapses the value of undifferentiated content.
When everyone can publish “good-enough” pages at near-zero marginal cost, the scarce resource stops being production capacity. It becomes trust, attention, and distribution.
The Slop Crisis Is Real
“Slop” was named 2025 Word of the Year by both Merriam-Webster and the American Dialect Society. The term describes low-quality AI-generated content produced with little regard for accuracy or value.
The definition is precise: “digital content made with generative AI that is perceived as lacking in effort, quality, or meaning, and produced in high volume as clickbait to gain advantage in the attention economy.”
Oxford researchers project AI content could reach 90% of internet material by next year. That’s not a trend. That’s a fundamental shift in how information gets created and consumed.
The problem shows up everywhere. Clarkesworld, an online science fiction magazine that pays contributors, stopped taking new submissions in 2024 because of the flood of AI-generated writing. During Hurricane Helene, people searching for real-time weather updates found themselves wading through a sea of AI-generated junk articles that added to the confusion.
The market is drowning in content that exists only because it’s cheap to make.
Google Is Actively Filtering the Flood
Search engines aren’t sitting idle while this happens.
Google reported a 45% reduction in low-quality, unoriginal content in search results following the March 2024 core update. In June 2025, Google initiated a major crackdown on AI-generated content, with sites receiving manual actions stating: “It appears that the pages of this site use aggressive spam techniques, such as large-scale content abuse.”
Sites affected saw complete visibility drops from Google search results.
Google’s 2025 Quality Raters Update emphasizes that AI content can receive a “Lowest” rating if it lacks originality or value. The guidance explicitly calls out “scaled content abuse” and content created with “little to no effort, little to no originality, and little to no added value.”
The policy is method-agnostic. A human writing 2,000 cookie-cutter pages violates the policy just like a fully automated script. When many pages are generated primarily to manipulate search rankings rather than help users, the method doesn’t matter.
Recovery from these penalties is slow and resource-intensive. Manual actions require reconsideration requests that can delay recovery by weeks while tanking revenue, ad impressions, and brand trust.
The Hidden Cost Shows Up Before Rankings Drop
When I audit companies pumping out AI-generated content at scale, the first damage never shows up as “rankings suddenly tanked.”
It shows up as user disengagement and erosion of trust signals.
You see rising bounce rates, shorter time on page, and fewer pages per session. Users quickly recognize the content as generic or unhelpful. Conversion metrics from content stall or decline even as the content library grows, because none of it moves a real buyer closer to a decision.
60% of marketers who use generative AI content are concerned it could harm brand reputation due to bias, plagiarism, or values misalignment. Yet the pressure to publish persists.
In B2B, approximately 90% of marketers are increasing content output, with nearly half producing 3-5x more content than last year while budgets barely move. This creates the “do more with less” mentality that feeds the slop machine.
By the time you feel the SEO hit, the trust and usefulness problem has been compounding for months.
AI Search Changes Everything About Discovery
AI Overviews now appear in over 50% of all Google searches as of October 2025, up from just 6.49% in January 2025. This represents the most disruptive change to traditional search since paid ads arrived.
The impact is measurable. AI Overviews reduce clicks to top-ranking pages by 34.5%, with click-through rates dropping from 15% to 8% when AI summaries appear. Top-ranking organic results lose up to 45% of their traffic due to AI, particularly for educational and informational queries.
Up to 60% of searches end without a click because the answer now lives inside the AI interface.
The relationship between content creation and attention capture has fundamentally changed.
Analysis of 6.8 million AI citations across ChatGPT, Gemini, and Perplexity reveals that 86% of AI citations come from brand-managed sources—websites, listings, and directories that brands already control.
Brand search volume is now the strongest predictor of AI citations, with a 0.334 correlation compared to backlinks’ 0.218. Being talked about matters more than being linked.
Traditional SEO strength showed little correlation with brand mentions in AI answers. Citation behavior is emerging as the key indicator of trust and authority in AI search.
The Unit of Competition Has Shifted
When AI adoption doubles, everyone has the same leverage. Leverage stops being differentiation and becomes table stakes.
The unit of competition is shifting from pages to proof. In a world of abundant text, what wins is what AI can’t cheaply fake at scale: demonstrated experience, proprietary data, real-world benchmarks, recognizable expertise, and a distribution footprint that earns citations and mentions.
A smaller, lower-volume brand can completely dominate when it gives AI systems something they can verify and reuse—proprietary data, clear experience, and structured signals—while the high-volume publisher gives them yet another interchangeable summary.
Picture two B2B SaaS companies selling similar analytics tools. Competitor A publishes 200+ blog posts in a year, mostly AI-assisted “what is / best practices / ultimate guide” content. Brand B publishes ~40 pieces, but each anchored in first-party data, real case studies, and transparent breakdowns of their own experiments.
Within 6-12 months, Brand B’s smaller library outranks and displaces Competitor A in both traditional SERPs and AI answers for commercial-intent queries.
First-party data beats generic summaries. Documented experience beats theoretical how-tos. Verifiable proof beats pattern-matched rewrites.
What Actually Works: The Proof Asset Model
The single asset I’d prioritize in the first 90 days is a category benchmark report built from your own first-party data.
You take data you already own—product usage, win/loss analysis, campaign performance, support patterns—and turn it into a named, annual report. This becomes the canonical answer for “what’s normal?” in your space, which is exactly the kind of structured, brand-owned information AI search tends to pull into answers.
Most companies ship 20-40 AI-assisted blog posts, each sourced from public information and lightly customized. Very little net-new information. Mostly repackaged guidance that any model could generate from the open web.
The benchmark asset looks different:
A single, deep hub page containing 5-10 clear findings: medians, ranges, and segment breakdowns that only you can know. Simple charts and tables, methodology notes, and definitions so models can parse and reuse it as a clean fact source.
Structured, machine-readable data marked up with schema and consistent entity naming, so AI systems can confidently associate those numbers with your brand and product.
Opinionated interpretation from real operators. Each finding has a short “What this means if you’re X” section, authored and attributed to named experts at your company. That attribution and interpretation layer is where experience and expertise show up.
A controlled universe of spin-outs. Instead of 30 disconnected articles, you’re building one defensible data spine and a small constellation of context around it.
AI systems favor this because it’s brand-owned, first-party data. It’s structured and consistent, which makes it safer for systems to lift directly into answers and attribute back to you. It solves high-intent queries that models struggle with using generic web text alone.
The Economic Case for Concentration
Your current model is paying to manufacture waste. A concentrated model pays to manufacture assets that keep earning.
Most teams are overextended because volume has quietly outpaced budget. A single decent blog post typically costs a few hundred dollars in internal or agency time. Multiply that by 6-12 posts a month and you’re burning thousands just to keep the calendar full.
Current model: 60-120 posts per year at $300-$600 all-in per post equals $18k-$72k per year. If most of those posts never earn links, citations, or pipeline, that spend is functionally operating expense with near-zero asset value. Plus it dilutes site authority.
You’re not adding work. You’re trading 30-50 low-impact pieces for 3-5 compounding assets.
Budget stays flat or even drops. What changes is the unit of work. Instead of 8 generic posts per month at ~$400 each, you do one flagship proof asset over 6-8 weeks with heavier research and design, plus 3-4 tightly scoped spin-outs that mine that asset.
Leadership moves when they see how this model reduces ongoing load. Fewer net-new ideas to invent each month. Distribution gets easier because every channel pulls from the same asset. Analytics gets cleaner because you can track pipeline and citations from Asset A and cut anything that doesn’t support it.
Leading Indicators That Prove It Works
In the first 90 days, I’m looking for proof that the new assets are earning attention more efficiently per unit of effort than the old calendar ever did.
Week 2: Higher scroll depth and time on page than your average post. More recirculation—people click from the asset into related pages. Email click-through rate on the launch send meaningfully above your list average. Qualitative feedback saying “this is exactly what we’re dealing with.”
Week 4: Same accounts or emails coming back to the asset and related URLs. Senior titles showing up in the logs, on the webinar, or in form fills tied to the asset. Net new newsletter signups directly attributable to the asset. Social and community posts about slices of the asset getting saves, shares, and thoughtful comments.
Week 8: New backlinks or citations where someone quotes a stat, a diagram, or a specific case and links back by name. References from partner newsletters, podcasts, or event decks that didn’t originate from your team. Your brand or asset starting to appear in a small share of AI answers for 5-10 priority prompts.
At this point, we may still be pre-pipeline in the CRM and pre-visible ranking shifts. But we already know people are engaging deeper than with your old posts, the right people are coming back and sharing it, and external surfaces including AI systems are beginning to treat it as something worth mentioning.
Holding the Line Against “Do Both”
Even when the leading indicators prove the model works, there’s enormous internal pressure to “do both”—keep the proof asset strategy running AND restart the old content calendar.
I treat “do both” as an operational risk. The minute you restart the old calendar at scale, you’re back to manufacturing noise that dilutes the very proof signals you just paid to build.
When someone waves a competitor’s blog count, I re-anchor the room on unit economics. If we copy their volume, do we copy their pipeline per post, or their content waste? One proof asset already outperforms 20+ legacy posts on depth, reuse, and assisted opportunities.
We formalize a rule: we do not optimize for output metrics like posts per month. We optimize for asset-level ROI and leading indicators that we’ve already agreed matter—engagement quality, citations, assisted revenue.
To add back 4 generic posts per month, we’d have to reassign hours from keeping our benchmark fresh or from external citation outreach. That means fewer third-party mentions and slower AI search authority, in exchange for pages we already know don’t move pipeline.
Sales usually wants “more content” because they lack usable content. I arm them with 3-5 sharp, repeatable artifacts from the proof asset and track how often they actually get used in deals. When reps can see that one report or case series is helping them advance conversations, they become allies in protecting focus.
As long as the indicators are improving, the default is to not add noise back into the system.
What Becomes Scarce Next
Five years out, when everyone’s stopped playing the volume game and is shipping some version of proof assets, the real scarcity isn’t content or even data.
It’s entity clarity and verified perspective.
Entity clarity means AI systems have a confident, non-contradictory model of who you are, who you serve, and what you’re expert in. Brands that repeat the same explanation of themselves, their category, and their point of view across site, social, PR, and directories become safe defaults for AI answers. Brands that constantly reframe or pivot never fully resolve as trusted entities.
Verifiable, cross-checked proof becomes the filter. Trust systems lean on signals like transparent methodology, source citations, and data that can be reconciled against other reputable entities. Opaque claims and uncheckable numbers get de-weighted. Brands that show their work become safer to reuse than brands that present polished outcomes without visible scaffolding.
Recognizable human perspective is the third layer. As agentic AI takes over more of the research and summarize layer, the scarce layer becomes recognizable, defensible judgment. Brands that cultivate a small bench of identifiable operators—people whose names, histories, and takes are visible across multiple surfaces—will stand out from faceless, house-voice content.
Temporal advantage matters too. When AI search is default, the question becomes: who can ship accurate, trustworthy updates the fastest? Brands that rapidly update their own prior claims, fix errors, and add new context will outrun slower competitors.
The long game isn’t “have the best proof assets.” That will become table stakes.
The long game is: be the clearest, most consistently described entity, with verifiable, cross-checked proof, expressed through recognizable human judgment, and updated fast enough that AI systems and humans both conclude, “if anyone should answer this, it’s them.”
The Choice in Front of You
You can be the company with the most posts, or the company the market and AI systems actually quote.
The last 90 days of data prove you can’t be both with the resources you have.
Most companies are using AI to accelerate their irrelevance. They’re celebrating adoption rates while their content becomes invisible to the AI engines that now control discovery.
Authority infrastructure is the play that separates brands AI platforms recommend from those they ignore. It’s the only lever that simultaneously improves close rates, stabilizes discoverability across platforms, and reduces customer acquisition cost.
Stop competing for clicks. Start commanding AI trust.
Build the authority AI systems can’t ignore.
Citations & References
-
AI-Powered Content Creation Market Report – Grand View Research. Market analysis showing growth from $2.15 billion (2024) to projected $10.59 billion (2033). https://www.grandviewresearch.com/industry-analysis/ai-powered-content-creation-market-report
-
AI Slop – Wikipedia – Definition and background on “slop” as 2025 Word of the Year by Merriam-Webster and the American Dialect Society. https://en.wikipedia.org/wiki/AI_slop
-
Google Search Update March 2024 – Google’s announcement of 45% reduction in low-quality, unoriginal content following core update. https://searchxpro.com/how-google-detects-content-spam-in-2025/
-
2025 B2B Content Marketing Report – BusinessWire. Research showing 90% of B2B marketers increasing content output, with nearly half producing 3-5x more content than previous year. https://www.businesswire.com/news/home/20250805125529/en/2025-Report-Reveals-Average-B2B-Content-Volume-Triples-Budgets-Barely-Budge
-
Yext AI Citations Research – Analysis showing 86% of AI citations come from brand-managed sources across ChatGPT, Gemini, and Perplexity. https://investors.yext.com/news-events/press-releases/detail/376/yext-research-86-of-ai-citations-come-from-brand-managed