TL;DR: Google's recent removal of the "num=100" URL parameter was not a minor technical update; it was a calculated strategic coup. With this single move, Google has achieved three critical objectives: 1) It dramatically raised the data-scraping costs for its AI rivals (like ChatGPT and Claude), reinforcing its formidable AI moat. 2) It burst the "ghost impression" data bubble that the SEO industry has long relied on, forcing a painful but necessary "great data correction." 3) It signals the definitive end of an era—the age of relying on unofficial shortcuts for data is over. For leaders, the message is clear: future digital authority will belong not to those who can scrape data, but to those who build genuine, verifiable "Trust Layers."
I am James, CEO of Mercury Technology Solutions.
Over the past several weeks, the global digital marketing landscape has been thrown into chaos. Brands worldwide have watched in confusion as their website's search impressions have plummeted in Google Search Console (GSC), while, paradoxically, their average rankings have miraculously improved.
The epicenter of this storm is Google's unannounced removal of a single, unofficial feature: the "num=100" URL parameter. To many, this is just an irrelevant line of code. But for the entire SEO data industry, this code was the very foundation upon which their business models were built.
However, to view this event as a mere technical adjustment is to make a grave strategic error. This was not an accident. This was a precisely calculated, far-reaching strategic maneuver aimed not just at the SEO world, but at the future balance of power in the entire AI ecosystem.
The Primary Strategic Objective: Fortify the AI Moat, Starve the Competition
We must first understand a fundamental truth: for AI challengers like ChatGPT, Claude, and the host of emerging AI search engines, Google's index is their "digital breadbasket." These Large Language Models (LLMs) depend on large-scale, systematic scraping of Google's search results to acquire the data needed to train their models and provide factual grounding for their real-time answers.
This has created an "AI supply chain" that is deeply unfavorable to Google: its competitors are using its most valuable asset—an index built over two decades—to create products that could ultimately unseat Google itself.
Before the removal of "num=100," an AI crawler could easily fetch the top 100 search results in a single request. This meant they could analyze a deep and diverse set of content, synthesizing more comprehensive and nuanced answers.
By deprecating this one line of code, Google has effectively increased the difficulty and cost of this data scraping by a factor of ten. This move significantly hinders its competitors' ability to access the depth and breadth of its index, effectively reinforcing Google's data moat. It is a textbook defensive strategy, achieving maximum strategic impact with a minimal technical change.
The Shockwave: A Long-Overdue "Great Data Correction"
The most fascinating aspect of this event is the data paradox it created in GSC. Why would impressions plummet while average rankings rise?
The answer reveals one of the SEO industry's long-standing "open secrets": GSC data has been heavily polluted by bot traffic from SEO tools.
Previously, when a crawler used the "num=100" parameter, it loaded a page with 100 results. Per Google's own definition, any link that appears on the current page counts as one "impression." Therefore, even if a site ranked at position 99—a position no real human user would ever see—the bot's single crawl would generate a valid "impression" record for that keyword in GSC.
Over time, these bot-generated "ghost impressions"—views that were never seen by a human eye—accumulated to a staggering scale. When Google removed "num=100," this artificially inflated impression data vanished overnight, causing the total number of impressions to collapse. What remained were the keywords with higher rankings, closer to the view of actual users, which naturally pulled the "average ranking" up.
This is more than a data cleanup. This is a "great data correction." It's a brutal exposé revealing that many brands may have been making strategic decisions based on a dangerously inflated "illusion of visibility." This once again validates the importance of Pillar 4: Strategic Intelligence from our Mercury 4 Pillars framework. Without a clear compass to distinguish real signals from noise, all of your efforts can be built on sand.
The Full Strategic Intent: A Calculated, Multi-Layered Maneuver
Beyond countering AI rivals and correcting data, Google's decision was driven by several other commercial motives:
- Reducing Infrastructure Costs: Serving 100 results for every query is a resource-intensive task. With the explosion of AI crawlers, this move directly cuts Google's operational overhead.
- Strengthening Data Control: This action subtly pushes high-volume data users (especially tool providers) away from the free, unofficial channel and toward Google's official, controllable, and typically paid API channels, like the Custom Search JSON API.
- A Defensible User Experience Argument: Google can reasonably claim that the new, cleaner reporting data is more aligned with "real user behavior," as the vast majority of human searches never go past the first page.
Conclusion: The Only Survival Tactic in the New Game
The removal of "num=100" was not the product of a single motive but a masterfully calculated decision. With one minor technical change, Google has shifted the balance of power in the AI ecosystem while simultaneously reaping benefits in strategy, cost control, business modeling, and data governance.
For those of us competing on the digital battlefield, this sends an unmistakably clear signal: The era of relying on unofficial shortcuts, data scraping, and chasing superficial metrics is definitively over.
In this new game, the only sustainable strategy is to build an asset that AI cannot easily replicate and that Google is eager to feature: a truly authoritative, verifiable brand. This means you must double down on creating deep "Answer Assets" (GAIO) and a pervasive "Trust Layer" (SEVO) across the entire ecosystem.
When the illusion of data shatters, the only thing that remains is your genuine authority.
