AI Slop Is Losing the Feed; Original Expertise Becomes Commerce Infrastructure

Last updated on July 23, 2026 at 12:25 PM

Google says traditional links are not disappearing because people still want expert opinions, original reporting, and useful perspectives beyond an AI summary. TikTok is testing systems to suppress AI-generated spam, LinkedIn is filling with machine-written posts, and a new audit says 70% of leading retailers are not ready for agentic commerce. As content and commerce become easier to automate, authentic expertise and reliable product data are becoming the scarce growth assets.

Today’s theme is clear: AI can make almost every marketing output cheaper, but it cannot make every output worth seeing.

Search engines, professional networks, social platforms, and commerce agents are all learning to filter an expanding supply of synthetic content. At the same time, retailers are discovering that being visible to an AI shopping agent depends less on publishing more pages and more on exposing complete, accurate, and current product information.

For ecommerce and marketing operators, the advantage is shifting from production volume to information quality. The brands that win will give people a reason to pay attention and give machines enough trustworthy data to act.

Let’s get into it.

THE BIG STORY

The Content Flood Is Making Original Expertise More Valuable

Google VP of Search Liz Reid told Tech Brew that the familiar list of links is not expected to disappear as Search becomes more AI-driven. Her argument is that users do not want to choose between AI and the web. AI can help them begin a task, but people still seek expertise, opinions, taste, videos, newsletters, podcasts, and deeper reporting when they want to understand a subject or make a decision.

That distinction matters because it changes the value of brand content. A generic article that only restates common information is easy for an AI system to compress. A useful expert explanation, original dataset, product demonstration, customer case study, or strong point of view gives the user a reason to continue beyond the summary.

Reid also said Google’s AI search experiences draw on the company’s underlying ranking systems to surface higher-quality sources. That does not guarantee publishers will keep the same level of referral traffic, but it suggests that the old foundations of credibility, relevance, and usefulness still feed the new interface.

TikTok is reaching a similar conclusion from the social side. The company says it is testing improved systems to identify accounts dedicated to AI-generated spam because that content can crowd out original creators. It is also preparing an in-app educational hub to help users recognize AI-generated media and use AI tools more responsibly.

The strategic signal is more important than the individual moderation feature. A platform that spent years making content creation easier is now investing in ways to distinguish useful creation from automated clutter. If low-quality synthetic content weakens engagement, trust, or advertising performance, platforms have a commercial reason to reduce its distribution.

EMARKETER’s analysis of LinkedIn shows the same tension. A Pangram analysis cited by EMARKETER found that LinkedIn accounted for 62% of the AI-generated social content it flagged, despite representing roughly one-third of the material analyzed. Across LinkedIn, Medium, Reddit, Substack, and X, 13.8% of posts were classified as fully AI-generated, rising to 25.7% for posts longer than 250 words.

Yet LinkedIn remains valuable precisely because it can distribute real professional expertise. EMARKETER notes that the platform is also an important source for AI search systems, making original executive perspectives, proprietary research, customer evidence, and industry analysis useful for both human discovery and AI visibility.

This creates a simple but demanding content standard. AI can help with research, structure, editing, and repurposing. The brand still needs to contribute the part that cannot be generated from an average of everything already published: experience, evidence, judgment, and a recognizable voice.

THE SHIFT

Agentic Commerce Rewards Product Data, Not Content Volume

A Search Engine Journal audit suggests many large retailers are not prepared for commerce journeys in which an AI agent can discover, compare, and potentially purchase a product without sending the shopper through the traditional website funnel.

The audit reviewed 207 high-traffic product-detail URLs from 29 retailers and evaluated 141 readable product pages against a readiness framework based on Google’s Universal Commerce Protocol. Most pages included basic product schema such as price, availability, brand, and SKU. The larger gaps appeared in the fields an agent needs to complete a transaction with confidence.

According to the audit, only 18% of the pages included a price-validity date, 13% exposed delivery-time information, and 11% provided a machine-readable return window. Seventy percent missed all three fields, while 65% did not include a GTIN that would help an agent recognize and compare the same product across retailers.

The article also found that some product pages could not be read because anti-bot protections returned access errors. That may be an intentional defense against scraping, but it can also block legitimate commerce agents. The business decision is no longer simply whether to allow bots. It is which agents receive access, which data they can retrieve, and how reliably the retailer can keep that data synchronized.

This is a different discipline from conventional content SEO. Category pages can rank well and still contribute little to an agentic transaction. Commerce agents rely on merchant feeds, product-level schema, inventory systems, shipping data, return policies, and stable identifiers. The challenge is not producing more copy. It is connecting the ERP, PIM, ecommerce platform, and merchant feed so the same product truth reaches every channel.

THE OPERATIONS LAYER

Amazon Grocery Shows How Ecommerce Becomes A Routine, Not An Event

EMARKETER reports that 51% of US digital buyers surveyed had purchased food or beverages through Amazon, Amazon Fresh, or Whole Foods during the previous 30 days. That makes Amazon grocery shopping a mainstream online behavior rather than an occasional experiment.

The result does not mean physical grocery retail has been displaced. EMARKETER notes that roughly 87% of grocery purchases still occur in stores, while Amazon’s physical grocery network remains much smaller than the footprint of major supermarket chains.

The more useful conclusion is that Amazon is becoming part of the routine replenishment journey. A shopper may still visit a supermarket for a large weekly trip while using Amazon for packaged food, beverages, household essentials, forgotten items, or products discovered through retail media.

For consumer brands, that changes marketplace strategy. Amazon is not only a search and conversion channel for considered purchases. It increasingly participates in high-frequency categories where availability, price, reviews, delivery speed, Subscribe & Save, and sponsored placement influence repeat behavior.

A MarketScale summary of a July 2026 market forecast estimates that global ecommerce could grow from about $7.65 trillion in 2025 to $19.83 trillion by 2035. As with any long-range forecast, the exact figure is uncertain, but the operational direction is clear: mobile commerce, digital payments, AI-assisted merchandising, cross-border fulfillment, and direct-to-customer models will place more pressure on the systems behind the storefront.

For B2B sellers, that pressure appears in account-specific pricing, procurement approvals, contract terms, tax handling, product availability, technical documentation, and repeat ordering. Growth will not come simply from adding an online checkout. It will come from making complex purchasing easier without losing the controls business customers require.

3 Things in Social, Ads & Trust

1. TikTok Is Connecting Creator Access, Sports Viewing, And Premium Ad Inventory

The NBA and WNBA have expanded their TikTok partnership through a multiyear agreement that combines creator access to major events, live-game discovery through TikTok GamePlan, and premium advertiser inventory through Pulse Premiere.

Accredited TikTok creators will be able to capture content around marquee league moments, while highlights can point viewers toward options for watching the game. For advertisers, this creates a loop between short-form discovery, creator storytelling, official league content, and larger-screen viewing.

The lesson extends beyond sports. Creator content works best when it does not sit beside the customer journey as a separate awareness tactic. It should help users discover the event or product, understand why it matters, and move toward the next action without breaking the experience.

2. AI Can Produce More Ads, But It Cannot Rescue Weak Creative

JumpFly argues that creative affects every stage of a social campaign: whether someone stops scrolling, understands the message, visits the site, and eventually converts. Those interactions also supply the platform with signals that influence delivery and optimization.

This is an important correction to the current AI creative narrative. Generative tools can produce more hooks, images, videos, and variations, but higher volume does not solve a weak proposition or an undifferentiated visual system. A campaign still needs a clear product truth, an audience-relevant idea, and enough variation to learn without losing brand consistency.

Creative testing should therefore begin with distinct hypotheses rather than cosmetic volume. Test different customer problems, demonstrations, proof points, formats, and offers. Then use AI to accelerate production inside those strategic directions.

3. Scam Ads Are Turning Platform Trust Into A Legal Question

The Consumer Federation of America is pursuing a case against Meta that alleges the company misleads Facebook and Instagram users about scam risks while profiting from fraudulent advertising.

Meta has argued that legal protections for platforms, including Section 230, support dismissal of the case. The consumer group counters that its allegations concern Meta’s own statements about platform safety, not simply content created by third-party advertisers.

The case may take time to resolve, but it highlights a growing tension in automated advertising. Platforms want to make buying easier and expand the number of advertisers, while users and legitimate businesses expect stronger controls against impersonation, fake offers, and fraudulent storefronts.

OFF THE CLOCK

Visibility Without Trust Is Just More Exposure

Today’s stories point to the same larger shift. Google says people still want links because they want to engage with human expertise. TikTok says AI spam can crowd out original creators. LinkedIn’s synthetic-content boom is making recognizable professional insight more valuable. Commerce agents are showing that even a strong brand can become invisible when product data is incomplete.

This means AI visibility cannot be managed as a single score. A brand may appear frequently in generated answers while failing to earn qualified visits, branded demand, assisted conversions, or purchases. Traditional SEO metrics such as rankings, click-through rate, and organic sessions still matter, but AI Overviews and zero-click behavior can change what those metrics mean.

A recent Forbes Agency Council article argues that marketers need to reconsider how they interpret familiar SEO indicators in the AI Overview era. The practical response is not to discard those metrics, but to pair them with stronger evidence such as branded-search growth, citation quality, qualified traffic, conversion, and revenue.

ADWEEK’s sponsored AI visibility resource frames search as a conversation rather than a simple quest for a ranked link. That shift makes inclusion important, but inclusion is only useful when the model represents the brand accurately and the user can verify the answer.

The scarcity is no longer content. The internet has more content than any person can process, and generative tools can expand that supply almost without limit.

The new scarcity is reliable distinction: a source worth citing, a creator worth watching, a product feed worth trusting, and a brand worth choosing.

For ecommerce and marketing operators, the operating system behind that distinction includes original expertise, platform-native creative, product data governance, transparent customer policies, and measurement tied to business outcomes.

AI can make the organization faster. It cannot decide what the organization should be known for.

The Beecommercer Team

Further Reading

1. Google exec says link lists aren’t going away as Search transforms – Tech Brew
2. TikTok says AI-generated spam crowds out original creators – TechRadar
3. Tired of AI slop clogging your feed? – Yahoo Tech
4. NBA and WNBA deepen TikTok ties with creator access and ad inventory – EMARKETER
5. 3 Mindset Shifts Marketers Need to Maximize AI Visibility – ADWEEK
6. LinkedIn’s AI content boom makes original brand expertise more valuable – EMARKETER
7. Consumer Federation Battles Meta Over Scam Ads – Yahoo Finance
8. 3 SEO Metrics Impacted by AI Overviews and What to Check Instead – Forbes
9. Why Creative Matters in Social Media Ads – JumpFly
10. Global ecommerce market forecast and B2B operational stakes – MarketScale
11. 70% of Top Retailers Are Invisible to Agentic Commerce – Search Engine Journal
12. Amazon Is Quietly Becoming a Grocery Run for Half of Online Shoppers – EMARKETER