- AI for Ecommerce and Amazon Sellers
- Posts
- Keywords Won't Matter For SEO In 2025?
Keywords Won't Matter For SEO In 2025?
Plus new from Anthropic & Meta

From Our Sponsor:
The smarter way for DTC brands to drive repeat revenue
What if you could drive 15% more repeat purchases, without the headache of building customized flows?
Black Crow integrates with your existing email & SMS tool, identifies customers that need a nudge to repurchase and determines the precise timing to trigger a pre-built flow.
Drive traffic to custom branded storefronts personalized to every shopper, showcasing the products they are most likely to buy.
Black Crow is so confident you’ll get results that they’re offering a $100 gift card to Shopify brands with $2M+ in GMV to get a demo.
Keywords Won't Matter For SEO In 2025?

The SEO keyword is dead, though its passing has gone largely unnoticed, leading to suboptimal marketing decisions and waning trust in SEO as a channel. This silent death has significant implications for e-commerce businesses and marketplace sellers who have long relied on keyword-based strategies for visibility and growth.
The transformation of search has been gradual but profound. Since 2013, when Google stopped sharing keyword referrers, marketers have been limited to Google Search Console data to understand user search patterns. The landscape grew more complex in 2014 with the introduction of Featured Snippets, creating winner-takes-all scenarios at the top of search results. The subsequent proliferation of SERP features – from People Also Asked boxes to video carousels – has fundamentally altered how users interact with search results.
Recent developments have accelerated this shift. Since January 2025, Google now requires JavaScript for Search, a technical change that has made rank tracking more expensive and complicated. This requirement represents the latest development in an ongoing tension between SEOs and Google, where rank tracking tools operate in a tolerated but officially unsanctioned space.
The implications are significant: a single webpage can now rank for thousands of keywords, provided it addresses the same underlying intent and comprehensively answers implied questions. However, this potential for broader ranking is offset by diminishing returns. As Sparktoro's research reveals, over 37% of searches now end without a click, while 21% lead to another search. The traditional correlation between ranking and traffic has broken down.
The challenge is exemplified in real-world cases. As Kevin Indig notes, his guide to in-house SEO achieved a position-one ranking but failed to generate any clicks, despite over 200 monthly searches for the target keyword. The traffic instead appears to flow to SERP features like People Also Asked boxes, which appear directly below the top result.
For e-commerce businesses, this shift demands a fundamental rethinking of SEO strategy. Rather than tracking individual keywords, success metrics should focus on aggregate organic traffic at both the page and domain level. This approach better aligns with how modern search engines process queries and how users interact with search results.
The rise of AI adds another crucial dimension. Google's AI Overviews, now available in more than 100 countries, provide in-depth answers that often make clicking through to search results redundant. This development coincides with the growing influence of AI-powered search alternatives. ChatGPT has already surpassed Bing's traffic volumes, while Perplexity continues to gain ground.

Google's JavaScript requirement appears strategically aimed at Gen AI competitors, whose LLM crawlers typically can't execute JavaScript, potentially limiting their ability to crawl Google's search results for grounding their answers. This technical barrier might temporarily affect LLM quality, but alternative data acquisition methods remain available. More significantly, this could push LLM developers to build independent models and web indices, potentially creating a new frontier in search optimization.
The implications for e-commerce and marketplace sellers are clear: success in search now requires understanding and optimizing for user intent rather than specific keywords. This means:
Target the most engaging SERP elements, including video carousels or AI answers, potentially expanding SEO efforts to platforms like YouTube
Focus on aggregate traffic metrics rather than individual keyword rankings
Consider seasonal patterns and distinguish between new and existing page performance
Utilize customer conversations and platform engagement data (from Reddit, YouTube) to inform content strategy
Leverage paid search data, which often better reflects user intent patterns
The data challenge remains significant. Google filters approximately 50% of query data for "privacy reasons," according to Ahrefs' analysis of 150,000 websites. Search volume metrics have become increasingly unreliable predictors of actual traffic potential.
For marketplace sellers and e-commerce businesses, this evolution requires a strategic pivot. Success in modern search demands understanding the flow of user intent rather than counting individual keyword performance. The future of search optimization lies not in tracking keyword positions but in comprehending and serving user intentions – a shift that particularly suits the emerging era of AI-driven search.
The Quick Read:
Anthropic is set to release a “hybrid” AI model that balances deep reasoning with fast responses. Expected soon, it could outperform OpenAI in coding tasks.
Despite VP J.D. Vance’s pushback on AI laws, multiple U.S. states are adopting regulations similar to the EU’s AI Act.
Meta’s Yann LeCun says today’s AI is far from human-like intelligence but expects a major breakthrough within 3–5 years, paving the way for smarter robotics.
A new partnership embeds AI agents into Snowflake’s platform, enabling businesses to analyze vast datasets while maintaining strict security controls.
The EU will invest €200B in AI, including €20B for “gigafactories” to boost model training. Aiming for AI leadership, Europe matches U.S. and China’s efforts.
Anthropic’s new Economic Index analyzes real-world AI usage across industries, revealing trends in automation, augmentation, and workforce transformation.
At the AI Action Summit, leaders warned AI could reshape economies and security by 2027.
Past chats can now be taken into consideration by Gemini AI
Today’s Content Spotlight:
The Tools List:
📹 PlayPlay - AI-powered video creation at scale. High-quality videos in just 1 click (link)*
🤖 Delphi Adapt - Create multiple versions of your clone, each suited for its intended audience.
🪶 SciSpace - Do hours worth of reading in minutes. Data extraction, AI Copilot, and paraphraser in one tool.
🎧 Save time by My Ask AI - Less time on customer support, more time on customer success.
🛠️ Altern - A website to find tools, products, resources, and more related to AI.
🗎 Guidde* - Magically create video documentation with AI.