Daily Edition · Page 9

THE DAILY INDEX

Technology, AI, software, security, crypto and business

NEWS & TRENDS

TLDR Marketing

WHY LINKEDIN MIGHT HAVE 2 ALGORITHMS & WHY MARKETERS SHOULD CARE

LinkedIn likely runs two separate systems that work together to determine what shows up in the feed. Retrieval decides if a post is shown at all by matching it to a viewer using signals like profile clarity and past positive engagement. If a post never enters this pool, it will not be seen. Ranking then orders those posts based on predicted interest, factoring in relationship strength, dwell time, early engagement, and content relevance. This also explains why some posts seem random when they take off. The system may need time or new signals to identify the right audience, allowing a post to gain traction later or outperform larger accounts if it is highly relevant to a specific group.

THE ATROPHY PARADOX

Aviation research found that overreliance on automation weakens pilots' manual skills, prompting calls to keep practicing them. A similar pattern is emerging with AI. While it boosts productivity, it can erode writing and domain expertise, the same skills needed to use it well. Studies show reduced cognitive engagement and weaker thinking with heavy AI use. As more people offload work, output starts to converge, creating sameness. The key distinction is between routine tasks, which can be automated, and judgment, which must be actively practiced to avoid skill atrophy.

I STOPPED GUESSING WHAT TO WRITE AND STARTED EAVESDROPPING INSTEAD

Instead of relying on keyword research, gather real audience conversations from communities, calls, or support data, and analyze them systematically. Extract verbatim signals like pains and objections, then cluster them into recurring themes and prioritize based on frequency and emotional intensity. Build detailed theme profiles with motivators and anti-patterns, then layer in insights on emotion and competitors. Finally, turn themes into content ideas mapped to funnel stages, ensuring each topic is grounded in real demand and uses the audience's own words.

HOW OPENAI ADS AFFECT ORGANIC STRATEGY

The launch of ChatGPT ads in February 2026 reinforces the importance of organic strategy rather than replacing it. Ads appear below the AI's answer and only on Free and Go tier accounts, meaning paying subscribers see brands exclusively through organic citations. With zero-click searches now at 60%, the priority has shifted from driving site visits to being the source AI models repeatedly reference. Brands with thin third-party coverage or poor technical accessibility for LLMs are especially vulnerable, as competitors can place ads directly beneath unfavorable AI-generated mentions.

PAYWALLS: EXPERT TIPS FOR TECHNICAL AND EDITORIAL SEO

Publisher paywalls require both editorial and technical alignment to avoid damaging SEO performance. Engagement signals like Navboost penalize pages with high bounce rates, and paywalled stories tend to have shorter visibility in Top Stories as a result. A strong editorial approach links free top-of-funnel content to locked premium pieces, creating a second click that moves readers closer to converting. Technically, the most common issue is missing or incorrect structured data, particularly the "isAccessibleForFree" attribute and CSS selectors that help Google properly crawl gated content.

BRINGING THE POWER OF PERSONAL INTELLIGENCE TO MORE PEOPLE

Google is expanding Personal Intelligence across AI Mode in Search, the Gemini app, and Gemini in Chrome for free-tier users in the US. The feature connects Google apps like Gmail and Google Photos to deliver personalized responses without requiring users to provide extra context. Use cases include tailored shopping recommendations based on past purchases, travel itineraries drawn from booking confirmations, and schedule-aware suggestions during layovers. Users control which apps are connected and can turn connections on or off at any time.

UTILITY-WRITING: ADD STRUCTURE TO LANGUAGE; NOT THE OTHER WAY AROUND

Writing should embed structure directly in language so machines can extract meaning without relying on technical markup. Tests show about 99% of sentences fail to stand alone, and systems like Gemini may use only about 380 words per page or 15.5% of content, so unclear or buried information becomes invisible. Content performs better when it includes named entities, explicit relationships, clear conditions, and self contained statements that answer specific questions. Dense and precise passages outperform longer pages because only 12% of very long content may be used, so adding more text can reduce visibility. The most effective approach is to write sentences that machines can quote directly while still giving humans clear and useful information.

EMAIL PERSONALIZATION HAS AN OVERUSE PROBLEM

Hyper-personalized email campaigns are creating subscriber fatigue as recipients grow numb to tactics like cart reminders and name-filled subject lines that thousands of brands use simultaneously. The argument is not to abandon personalization but to deploy it more selectively, reserving it for moments where it genuinely adds value, such as early access for loyal customers or contextually relevant offers. Not every campaign needs personalization. Product launches, for example, often work better as broad sends. The key distinction is between making subscribers feel understood versus reminding them how much data a brand holds on them.

TLDR IndexPage 9