Is Your Rank Tracker Lying? Google Is Feeding It Bad Data

Rank tracking tools have been unreliable for weeks as Google fights scrapers. See what is happening and how to report SEO progress without trusting one number.

If your SEO report showed a sudden ranking drop in the last few weeks, or a strange recovery a few days later, the problem may not be your website at all. On 12 October, Search Engine Roundtable reported that third-party Google rank tracking has been struggling for several weeks, with Google blocking scrapers more effectively than it has in a long time.

That matters because many agencies, in-house teams and founders still judge SEO by a weekly keyword position chart. When the chart is wrong, good work gets blamed and bad decisions get made. Here is what is going on, who it hits, and how to keep your reporting honest.

What Google changed

Rank trackers work by requesting Google results pages in bulk and recording where each site appears. Google has been making that harder. According to the same report, Google began testing "goto" redirects in search result links in July, tried more human verification steps, and changed the `num=100` parameter back in 2025. The author assumes the redirects are aimed at AI bots and Google's dispute with SerpApi rather than at SEO tools, and notes tools can handle them fairly easily.

The bigger problem is what Derek Perkins of Nozzle describes. He sees a three-part approach from Google:

• Wide IP blocking aimed at major residential proxy networks.
• A significant increase in captchas.
• Injecting bad data into suspect search results, so scrapers sometimes receive wrong answers.

Nozzle's research, as summarised in the report, says that since September about 8% of the results pages it collected came back "poisoned": they look normal, but the listings have been swapped out. Social sites and forums filled the gaps, and tracked sites sometimes dropped out of the top 10. The effect passes within days, which is exactly why it is easy to mistake for a real ranking swing.

Why the bad-data part is the real risk

A tool that fails loudly is annoying. A tool that quietly records a wrong position is dangerous. The report's author calls the bad-data approach the most concerning piece, because tools may not notice it quickly. A blocked request shows up as an error. A swapped page just becomes a number in your dashboard.

Practitioners are seeing the mess first hand. Glenn Gabe reported that Ahrefs picked up a drop and recovery, Semrush lagged, and Sistrix missed both. Lily Ray said rank trackers were too unreliable to share findings. Sistrix, DataForSEO, SE Ranking, Advanced Web Ranking and Morningscore have posted notices or explanations, and SE Ranking said most of its missing-position issue had been resolved. Many other vendors did not respond.

Who gets hurt most

Three groups feel this first.

• Businesses paying for SEO on a ranking basis. If a retainer or bonus depends on keyword positions, a bad data week can trigger an argument that has nothing to do with the work.
• Teams reporting to a board or CFO. A false drop that gets presented as fact is hard to walk back later.
• Local businesses. BrightLocal told customers in September that its local rank tracker reports were running slower because of Google's scraping changes.

A better way to read SEO performance right now

Rank tracking was never the goal. Revenue is. Use this stretch as a prompt to build a reporting stack that does not hang on one fragile source.

1. Start with Google Search Console. It is Google's own data, so it is not scraped. Look at clicks and impressions trends for groups of pages, not single keywords.
2. Confirm drops with a second source. Before you report a decline, check analytics sessions from organic search, leads and sales for the same pages and dates.
3. Spot-check important queries by hand. The report mentions checking key queries in an incognito window, and manual refreshes helping in some tools. It is slow, so reserve it for the 10 or 20 terms that matter most.
4. Wait a few days before reacting. If Nozzle is right that poisoned results pass within days, a one-week dip that reverses is likely noise.
5. Ask your tool vendor directly. Find out how they detect bad pages and whether they have flagged affected dates.

One more limit to keep in mind: the report notes that Search Console lacks click and query data for AI Overviews and AI Mode, so part of your visibility is still hard to see. That is another reason to read several signals together. Our SEO reporting and forecasting work is built around that idea, and a clean web analytics setup is what makes the cross-check possible.

What this says about where SEO measurement is heading

Google has little reason to make scraping easy, and the dispute with scrapers is unlikely to end soon. Expect third-party position data to stay noisier and more expensive. Teams that tie SEO to organic leads, pipeline and revenue will barely notice. Teams that sell and buy SEO on a rank chart will keep having confusing weeks.

This is also a good time to look at what your reports reward. If a keyword moves from position 4 to position 6 but conversions rise, nobody should panic. If positions look great and leads fall, the chart is flattering you. A technical SEO check can tell you whether the site itself changed, which helps separate real losses from measurement errors.

Questions to put to your agency or SEO lead

If someone else runs your SEO, this week is a fair moment to ask how they are handling the noise. A good answer is specific. They should tell you which sources they cross-check, which dates they treat as suspect, and which business numbers they watch alongside positions.

• Which tool produced this chart, and has that vendor reported any problems since September?
• Does the drop show up in Search Console clicks and in organic leads, or only in the tracker?
• Which 10 to 20 queries have been checked by hand, and what did we find?
• What would we have decided differently if this data were wrong?

Vague replies, or a refusal to look beyond one dashboard, tell you something about how the work is measured. It is much cheaper to find that out now than after a quarter of decisions built on a flawed chart.

Key takeaways

• Third-party Google rank data has been unreliable for several weeks, according to Search Engine Roundtable.
• Nozzle reports that about 8% of results pages it collected since September were poisoned, with swapped-in listings that look normal.
• Do not report a ranking drop until Search Console, analytics and leads agree with it.
• Spot-check your most valuable queries by hand and give odd results a few days to settle.
• Move your reporting toward traffic, leads and revenue, and treat positions as a supporting signal.

If your SEO reports have started to look wrong, we can audit the data behind them and show what is actually happening with your organic growth. You can get a free quote whenever you are ready.

Sources

• Third-Party Google Rank Tracking Struggling Now (Search Engine Roundtable)
• Slow reports and Google scraping issues (BrightLocal)