Last month, Google update quietly removed the num=100 search parameter. You can no longer load 100 results at once; the effective cap is now 10. That seemingly small UX change narrows visibility across the web’s “long tail,” affects how AI systems discover and cite sources, and pushes founders to treat distribution as a first-class problem, before product.
Google Update: What changed?
For years, advanced users and scrapers could append num=100 to a query to view more results in one go. That back door is now shut. Practically, discovery collapses toward the top 10 blue links. Pages that lived (and thrived) in positions 11-100 are suddenly harder to reach.
The Significance?
- AI discovery relies on what’s indexed and surfaced. Industry watchers have long noted that many LLM-powered products lean, directly or indirectly on Google-indexed results alongside their own crawlers. If fewer results are easily surfaced beyond page one, fewer diverse sources get pulled into AI citations and answers.
- The long tail shrinks. Overnight, access to that tail positions 11-100, was effectively cut by 90% for users and systems that depended on bulk result views.
The fallout
- Impressions fell broadly. Search Engine Land reported that 88% of sites saw a drop in impressions following the change.
- Communities lost visibility. Reddit, which often ranks in positions 11-100, reportedly saw its LLM citations plunge. Its stock fell ~15%, wiping out roughly $5B in market value.
The new reality: Distribution > Product
“Distribution > Product.” The line, attributed here to Adarsh Appaiah, captures the shift. A great product without discovery is invisible. In a world where page-one real estate and AI citations concentrate attention, distribution is not a marketing afterthought; it’s an existential moat.
What this concentrates and what it leaves out
- Concentrates: Page-one incumbents, brands with established authority, and content that already wins high-intent keywords.
- Left out: High-quality niche pages, new entrants without authority, and community/forum content that historically lived beyond the first 10.
What founders and marketers can do now
- Design for page-one or bust. Pick a narrow, defensible set of queries and build true authority around them (depth, topical clusters, internal linking).
- Diversify discovery. Don’t rely on Google alone. Build owned channels (newsletter, community, SMS), strengthen direct/brand search, and cultivate referral partnerships.
- Engineer distribution loops. Make sharing and network effects native to the product (invites, embeds, user profiles that rank for their own names).
- Capture demand at the source. Be present where evaluation happens, trusted newsletters, communities, podcasts, and category directories your buyers already use.
- Structure your content for machines and humans. Clean information architecture, fast pages, rich snippets/structured data, and on-page answers that earn inclusion when AI systems summarize.
- Measure like survival depends on it. Track impressions, average position, CTR, and assisted conversions; when impressions dip, shift budget and creative toward channels with provable reach.
What to watch next
- Further SERP UX changes that might tighten (or relax) access beyond the top 10.
- How LLM search products adapt, and whether they broaden their own crawls to offset reduced mid-tail visibility.
- Category-level reshuffles as attention concentrates: some sites will stabilize on page one; many will need entirely new distribution muscles.
Google’s removal of num=100 sounds like housekeeping. In practice, it’s a distribution shock. With less surface area for discovery, attention centralizes and startups must earn visibility before they can earn evaluation. In today’s AI-shaped ecosystem, distribution isn’t a lever to pull after launch; it’s the strategy that makes launch matter.
Also Read: Apoorv Sharma & Derivate X: Rewriting SEO for the AI Age
