Rank tracking and the &num=100 crisis

We talk a lot about shifting landscapes, algorithm updates, AI overviews and now AI mode. But sometimes the biggest shake-up isn’t a new algorithm. A month has passed since Google quietly removed a low-level technical door we’ve all relied upon for tracking. The &num=100 parameter went away overnight, surprising us all, and providing some jump scares for SEOs the world over.

Now, most SERP scrapers or rank trackers must fall back to paginated requests (10 results per page), multiplying work and cost, which is likely to end up giving us SEOs less accurate reporting at a higher cost.

After waiting a few weeks to see how it played out, it’s been clear that the fallout has been messy:

• Google Search Console (GSC) impressions plummeted for many sites, in some reports, by 40–70 %. Search Engine Land reported that 77% of sites lost keyword visibility immediately after the change.
  
  
• Average position improved (on paper), but mostly because the lower-position impressions (from tools scraping deep pages) were dropped.
  
  
• Many keywords that used to show on page 3+ vanished from GSC and 3rd party tool reports.
  
  
• Rank trackers like Semrush, Ahrefs, Moz, etc., scrambled, acknowledging data gaps, slower report generation, and reengineering their scraping logic, after already being hit by Google’s Anti-Scraping update earlier this year.

So this isn’t a “ranking update.” It’s a measurement update. The algorithm didn’t flip a switch for once; this time, our tools did. And it feels bad…

How are rank trackers adapting?

We don’t yet have full transparency from all tool vendors, but from public signals and early reporting, here’s how things are playing out:

Pagination is back

Tools must now perform multiple requests (for pages 1–10, 11–20, 21–30, etc.) to reconstruct the top 100. That multiplies infrastructure, rate-limiting exposure, and error surface. Most SEOs are concerned that this will mean increased costs and less accuracy for our tools as time goes on.

Reduced default depth

Some vendors are capping their “free” or default tracking to the top 20 (or even top 10), pushing deeper rank data into premium tiers (if offered at all).

Slower updates, more gaps

More HTTP calls, more processing, and more risk of partial fails mean delays in data refresh and occasional “no data” or “not ranked” placeholders.

Rethinking impressions and bots

Tools are acknowledging that a lot of what we saw in GSC previously was inflated by bot / scraper traffic (thanks to &num=100). Some of the “drops” in impression volume now are just the scrapers being excluded, which is good for accuracy but bad for our year-on-year comparisons.

Communications blur

Some vendors have released statements, others remain silent. Some are pushing users toward “visibility scores” or alternative internal metrics rather than raw positions. Currently, we have statements from Semrush and Accuranker, with others likely to come once tools have had time to ship some fixes.

What This Means for SEO Reporting & Strategy

Overnight, the foundations of how we measure organic visibility have shifted. As an SEO lead, here’s how I’m thinking about adapting:

1. Treat mid-September 2025 as a data break

You cannot directly compare pre- vs post-September ranks, impressions or query counts without caution. Any trendline spanning that date likely hides a discontinuity from measurement changes. Call it out in reports. Use annotations. Convey the caveat to stakeholders so that everyone knows why those lines are going up and down.

2. Lean harder on behaviour + conversion metrics

If rank and impression data are now noisier, then metrics tied to actual user value become more critical:

• Organic sessions (from analytics)
  
  
• Clicks & CTR (in GSC)
  
  
• Engagement metrics (bounce, time on page, scroll depth)
  
  
• Conversion rate, lead volume, revenue
  
  
• Assisted conversions or downstream metrics

These are harder to “break” via scrapers and ultimately more meaningful for a client’s bottom line.

3. Focus even more on the top 10 / top 20

The lowest-hanging fruit (from an organic lift perspective) is already in those ranges. If you’re in positions 11–20, your chance to push into page one may now be more clearly visible; waves beyond that probably matter less from a traffic ROI view. Is it annoying not to be sure of what’s working for deeper rankings? Yes. Can we do much about it right now? No.

4. Shift to visibility / share metrics instead of absolute ranks

Instead of obsessing over “Keyword A is #57 today,” consider:

• What fraction of tracked keywords are in the top 3 / top 10 / top 20?
  
  
• How is your visibility or share of voice changing vs competitors?
  
  
• How many new keywords cross into the top 10 or top 20 over time?

These aggregate signals feel more robust when day-to-day ranks are shaky at best.

5. Educate stakeholders now

This is a subtle but high-impact shift. Your managers, clients or internal stakeholders may panic at a sudden 50% drop in “impressions.” The narrative must be consistent: this is measurement-related, not performance collapse. Transparently communicating this will preserve trust.

Final thoughts

The deprecation of &num=100 has been less of a crisis and more of a forced introspection for SEO. We’re being pushed to rely less on brittle rank reporting and more on meaningful user-centred metrics, which, with a little hindsight, I think is a good thing for the industry. The tools will adapt. The scrapers will adapt. But the advantage goes to those who can tell a clearer, more honest story with their SEO data.